Of Speech and Sanctions: Toward a Penalty-Sensitive Approach to the First Amendment

Louisiana State University Law Center Louisiana State University Law Center

LSU Law Digital Commons LSU Law Digital Commons

Journal Articles Faculty Scholarship

2012

Of Speech and Sanctions: Toward a Penalty-Sensitive Approach to Of Speech and Sanctions: Toward a Penalty-Sensitive Approach to

the First Amendment the First Amendment

Michael Coenen

Louisiana State University Law Center

, [email protected]

Follow this and additional works at: https://digitalcommons.law.lsu.edu/faculty_scholarship

Part of the Constitutional Law Commons, and the First Amendment Commons

Repository Citation Repository Citation

Coenen, Michael, "Of Speech and Sanctions: Toward a Penalty-Sensitive Approach to the First

Amendment" (2012).

Journal Articles

. 122.

https://digitalcommons.law.lsu.edu/faculty_scholarship/122

This Article is brought to you for free and open access by the Faculty Scholarship at LSU Law Digital Commons. It

has been accepted for inclusion in Journal Articles by an authorized administrator of LSU Law Digital Commons.

For more information, please contact [email protected].

Electronic copy available at: http://ssrn.com/abstract=1908408

COLUMBIA LAW REVIEW

VOL. 112 JUNE 2012 NO. 5

991

ARTICLE

OF SPEECH AND SANCTIONS: TOWARD A PENALTY-

SENSITIVE APPROACH TO THE FIRST AMENDMENT

Michael Coenen*

Courts confronting First Amendment claims do not often scrutinize

the severity of a speaker’s punishment. Embracing a “penalty-neutral”

understanding of the free speech right, these courts tend to treat an indi-

vidual’s expression as either protected, in which case the government

may not punish it at all, or unprotected, in which case the government

may punish it to a very great degree. There is, however, a small but im-

portant body of “penalty-sensitive” case law that runs counter to the

penalty-neutral norm. Within this case law, the severity of a speaker’s

punishment affects the merits of her First Amendment claim, thus giving

rise to categories of expression that the government may punish, but only

to a limited extent. This Article defends penalty-sensitive free speech ad-

judication and calls for its expanded use within First Amendment law.

Pulling together existing strands of penalty-sensitive doctrine, the Article

identifies five ways in which penalty-sensitive analysis can further im-

portant constitutional objectives: (a) by increasing fairness for similarly

situated speakers; (b) by mitigating chilling effects on protected speech;

speech restrictions; (d) by rooting out improper government motives; and

(e) by promoting transparency in judicial decisionmaking. The Article

also considers and rejects potential objections to the penalty-sensitive ap-

proach, concluding that it will often generate proper results in difficult

First Amendment cases.

* Climenko Fellow and Lecturer on Law, Harvard Law School. Thanks to Akhil

Amar,

Anna Arkin-Gallagher, Douglas Berman, Anya Bernstein, Michael Burstein, Joseph

Blocher, Dan Coenen, John Coyle, Jaime Dodge, Tom Donnelly, Avlana Eisenberg, Barry

Friedman, Heather Gerken, Jeremy Golubcow-Teglasi, Zach Gubler, Rebecca Haw, Paul

Horwitz, Michael Klarman, Ashley Lott, Dina Mishra, Sonja Ralston, Chris Sherman,

Richard Tao, Susannah Tobin, Cora True-Frost, Anthony Vitarelli, Eugene Volokh, and

Daniel Winik for helpful comments and suggestions. Thanks also to Jennifer Green and

the staff of the Columbia Law Review for skillful editorial assistance. All errors are mine.

Electronic copy available at: http://ssrn.com/abstract=1908408

992 COLUMBIA LAW REVIEW [Vol. 112:5

NTRODUCTION .......................................................................................... 993

I. WHAT IS PENALTY SENSITIVITY?............................................................... 999

II. PENALTY SENSITIVITY IN OPERATION ................................................... 1002

A. Penalty Sensitivity and Free Speech Doctrine............................ 1002

1. Seditious Libel and the Abrams Dissent............................... 1002

2. Defamation........................................................................... 1005

3. Public Employment.............................................................. 1008

4. Vagueness/Overbreadth...................................................... 1010

5. Indecency. ............................................................................ 1013

6. School Speech...................................................................... 1015

7. Three Red Herrings............................................................. 1016

a. Prior Restraints ............................................................. 1017

b. Unconstitutional Conditions ....................................... 1018

c. Sentencing Enhancements........................................... 1021

B. Penalty Sensitivity in Other Constitutional Contexts ................ 1022

III. JUSTIFYING THE PENALTY-SENSITIVE APPROACH ................................. 1026

A. Fairness........................................................................................ 1027

B. Chilling Effects............................................................................ 1031

C. “Preserving” Unprotected Speech.............................................. 1035

D. Government Motives................................................................... 1039

E. Transparency............................................................................... 1042

IV. OBJECTIONS TO THE PENALTY-SENSITIVE APPROACH.......................... 1044

A. Legitimacy ................................................................................... 1045

B. Administrability........................................................................... 1047

C. Liberty.......................................................................................... 1052

CONCLUSION ............................................................................................ 1054

2012] OF SPEECH AND SANCTIONS 993

NTRODUCTION

Governments restrict speech by punishing speakers.

They incarcer-

ate inciters,

make defamers pay damages,

and put pornographers on

probation.

They fine broadcasters,

enjoin publishers,

and sanction

internet service providers.

They lengthen the sentences of hate crime

defendants

and hold loose-lipped litigants in contempt of court.

They

fire employees who criticize superiors,

discipline schoolchildren who

protest in the classroom,

and withhold subsidies from applicants who

espouse disfavored beliefs.

In these and other circumstances, state ac-

tors implicate the free speech right not just by imposing restrictions on

expressive behavior, but also by fortifying these restrictions with the

threat and exaction of punishment.

1. See, e.g., Bartnicki v. Vopper, 532 U.S. 514, 529 (2001) (“The normal method of

deterring unlawful conduct is to impose an appropriate punishment on the person who

engages in it.”).

2. See, e.g., Brandenburg v. Ohio, 395 U.S. 444, 444–45 (1969) (reviewing free

speech claim of defendant sentenced to one to ten years’ imprisonment (and fined

$1,000) under Ohio Criminal Syndicalism statute).

3. See, e.g., N.Y. Times Co. v. Sullivan, 376 U.S. 254, 256 (1964) (addressing First

Amendment issues arising from $500,000 defamation verdict).

4. See, e.g., Roth v. United States, 354 U.S. 476, 485 (1957) (rejecting First

Amendment challenge to obscenity prosecutions); see also Kingsley Books, Inc. v. Brown,

354 U.S. 436, 444 (1957) (describing sentence of one defendant in Roth as two years’ pro-

bation).

5. See, e.g., Fox Television Stations, Inc. v. FCC, 613 F.3d 317, 323 n.3 (2d Cir. 2010)

(noting FCC issued $8 million in fines for indecency violations in 2004), cert. granted, 131

S. Ct. 3065 (2011).

6. See, e.g., Near v. Minnesota, 283 U.S. 697, 722–23 (1931) (reviewing district court

order enjoining publication of allegedly defamatory periodical).

7. See, e.g., John Doe, Inc. v. Mukasey, 549 F.3d 861, 864 (2d Cir. 2008) (addressing

internet service provider’s First Amendment challenge to statutory requirements prohibit-

ing disclosure of investigative requests).

8. See, e.g., Wisconsin v. Mitchell, 508 U.S. 476, 479–81 (1993) (reviewing two-year

enhancement of aggravated battery sentence under Wisconsin hate crime statute).

9. See, e.g., Eaton v. City of Tulsa, 415 U.S. 697, 703–04 (1974) (Rehnquist, J., dis-

senting) (discussing contempt citation and resulting $50 fine issued against witness who

used term “chicken shit” in open court).

10. See, e.g., Pickering v. Bd. of Educ. of Twp. High Sch. Dist. 205, 391 U.S. 563, 565

(1968) (reviewing school district’s dismissal of teacher for criticizing school board in letter

to local newspaper).

11. See, e.g., Tinker v. Des Moines Indep. Cmty. Sch. Dist., 393 U.S. 503, 504 (1969)

(reviewing public school’s suspension of students who wore black armbands in protest of

Vietnam War).

12. See, e.g., NEA v. Finley, 524 U.S. 569, 576 (1998) (reviewing statutory require-

ment that National Endowment for the Arts “‘tak[e] into consideration general standards

of decency and respect for the diverse beliefs and values of the American public’” in re-

viewing grant applications (alteration in original) (quoting 20 U.S.C. § 954(d)(1)

(2006))).

994 COLUMBIA LAW REVIEW [Vol. 112:5

While punishments lie at the core of real-life First Amendment dis-

putes, they occupy only the peripheries of First Amendment analysis. In

particular, the severity of the penalty imposed—though of central impor-

tance to the speaker who bears it—does not normally affect the merits of

his free speech claim. Fighting words doctrine does not distinguish be-

tween ordinances that punish with fines and those that punish with im-

prisonment.

Commercial speech doctrine does not distinguish between

large and small monetary penalties.

Public employment doctrine does

not distinguish between temporary suspensions and permanent dismiss-

als.

Much free speech review follows this pattern, asking the question of

whether, and not to what extent, the government may punish expressive

conduct. Consequently, First Amendment litigation tends to proceed as a

winner-take-all affair. Speech is either protected, in which case it may not

be punished, or unprotected, in which case it may be punished to a very

great degree.

In this respect, the standard method of First Amendment

analysis is penalty-neutral.

This penalty-neutral approach to free speech adjudication is no

doubt familiar to students of First Amendment law. Less familiar is its

penalty-sensitive alternative, which treats the severity of a speaker’s pun-

ishment as relevant to the merits of his First Amendment claim. On a

penalty-sensitive understanding of the free speech right, some forms of

expression warrant neither total immunization against nor total exposure

to the threat of government-sponsored sanction. Rather, for such expres-

sion, the First Amendment permits some forms of punishment but not

others. Thus, to the penalty-sensitive analyst, it matters whether prison

sentences are long or short, whether civil damage awards are large or

small, whether public workplace dismissals are permanent or temporary,

and so on. Such an analyst asks not just, “What kinds of speech may the

government sanction?” She also asks, “What kinds of sanctions may the

government impose on speech?”

Without a doubt, penalty sensitivity is an uncommon feature of the

First Amendment landscape. Indeed, it is sufficiently uncommon that,

13. Cf. Chaplinsky v. New Hampshire, 315 U.S. 568, 571–72 (1942) (“There are cer-

tain well-defined and narrowly limited classes of speech, the prevention and punishment

of which have never been thought to raise any Constitutional problem. These include . . .

insulting or ‘fighting’ words.”).

14. Cf. 44 Liquormart, Inc. v. Rhode Island, 517 U.S. 484, 489, 493 (1996) (invalidat-

ing enforcement action resulting in collection of $400 fine).

15. See Rutan v. Republican Party of Ill., 497 U.S. 62, 75 (1990) (observing “there are

deprivations less harsh than dismissal that nevertheless press state employees and appli-

cants to conform their beliefs and associations to some state-selected orthodoxy”). But see

infra Part II.A.3 (noting limited instances of penalty-sensitivity in public employment

cases).

16. See Vincent Blasi, Toward a Theory of Prior Restraint: The Central Linkage, 66

Minn. L. Rev. 11, 27 (1981) (noting “[l]ong criminal sentences and enormous civil dam-

age awards are not uncommon, even for speech at the margins of constitutional protec-

tion”).

2012] OF SPEECH AND SANCTIONS 995

among those few scholars who have paused to consider the matter, there

is general agreement that courts reviewing free speech cases rarely, if

ever, employ the penalty-sensitive approach. Hence, Christine Jolls, Cass

Sunstein, and Richard Thaler have observed that “no one has suggested

that the First Amendment imposes limits on the severity of punishment

for speech that the government is entitled to criminalize.”

Vincent Blasi

has similarly noted that, with one “limited” exception, “the Supreme

Court has been unwilling to erect constitutional limitations on how se-

verely persons can be sanctioned for engaging in unprotected expres-

sion.”

Such conclusions comport with statements from the Court itself,

whose members have sometimes rejected, in no uncertain terms, the idea

of imposing penalty-based limits on the government’s power to restrict

speech. In declining to impose such limits, the Court has proclaimed, for

instance, that “the First Amendment . . . protects state employees . . .

from ‘even an act of retaliation as trivial as failing to hold a birthday party

for a public employee’”

and that “it is not for this Court to limit the

State in resorting to various weapons in the armory of the law.”

Under-

stood in isolation, such statements might be read to indicate that free

speech adjudication is an exclusively penalty-neutral affair.

This view, however, is too strong. Penalty-sensitive free speech analy-

sis may be less prevalent than its penalty-neutral counterpart, but it is by

no means nonexistent. Indeed, the Court, individual Justices, and some

lower courts have tinkered with penalty-sensitive analysis in a variety of

free speech settings. One finds penalty sensitivity at work, for example, in

Oliver Wendell Holmes’s famous dissent in Abrams v. United States, which

characterized lengthy prison sentences for seditious libel as a significant

First Amendment problem.

More recently, the Court assumed a

penalty-sensitive stance in FCC v. Pacifica Foundation, where, in upholding

an FCC enforcement action against First Amendment attack, it pointed

to the lenience of the sanction imposed.

Penalty sensitivity manifests

itself in the Court’s practice of reviewing First Amendment vagueness

and overbreadth claims more aggressively in the criminal context than in

17. Christine Jolls, Cass R. Sunstein & Richard Thaler, A Behavioral Approach to Law

and Economics, 50 Stan. L. Rev. 1471, 1517 (1998).

18. Blasi, supra note 16, at 26–27; see also Tom W. Bell, Treason, Technology, and

Freedom of Expression, 37 Ariz. St. L.J. 999, 1039 (2005) (“[T]he Supreme Court has at

best only suggested that excess punishments may render a restriction on speech more re-

strictive than necessary.”).

19. Rutan, 497 U.S. at 76 n.8 (quoting Rutan v. Republican Party of Ill., 868 F.2d 943,

954 n.4 (7th Cir. 1989), aff’d in part, rev’d in part, 497 U.S. 62); see also Smith v. Fruin, 28

F.3d 646, 649 n.3 (7th Cir. 1994) (“[E]ven minor forms of retaliation can support a First

Amendment claim, for they may have just as much of a chilling effect on speech as more

drastic measures.”).

20. Fort Wayne Books, Inc. v. Indiana, 489 U.S. 46, 60 (1989) (quoting Kingsley

Books, Inc. v. Brown, 354 U.S. 436, 441 (1957)).

21. 250 U.S. 616, 629 (1919) (Holmes, J., dissenting).

22. 438 U.S. 726, 730 (1978).

996 COLUMBIA LAW REVIEW [Vol. 112:5

the civil context, and in some lower courts’ refusal to extend First

Amendment protection to public employees who have not suffered suffi-

ciently “adverse” forms of retaliation. Penalty-sensitive analysis underlies

the rule of Gertz v. Robert Welch, Inc., which permits states to award “actual

damages,” but not presumed or punitive damages, in certain categories

of defamation actions.

In these and other cases, courts have signaled

that the scope of the free speech right depends on the harshness of the

penalty administered.

The shared penalty-sensitive character of these cases has gone

largely unnoticed—and hence unevaluated—by judges and scholars

alike. When it has flirted with penalty-sensitive review, the Court has pro-

ceeded in an ad hoc manner, neglecting to explain why, how, or when

the severity of a speaker’s penalty should affect the scope of his First

Amendment rights. And legal academics have left this field unexplored.

Although some scholars have discussed individual areas of doctrine in

penalty-sensitive terms,

no one has previously tied these doctrinal lines

together or evaluated in general the merits and demerits of the penalty-

sensitive approach. One thus searches the case law and legal literature in

vain for a systematic assessment of penalty-sensitive free speech analysis.

Given penalty sensitivity’s now limited—but still unmistakable—

presence within First Amendment law, it is well worth asking whether this

sort of analysis ought to be included. Should we regard existing examples

of penalty-sensitive analysis as aberrant mistakes not to be repeated? Or

should we regard them as the scattered seeds of a promising doctrinal

development? What, in short, are the promises and pitfalls of penalty-

sensitive free speech analysis? And if this approach does have merit, how

should courts pursue it?

23. 418 U.S. 323, 349 (1974).

24. See, e.g., Tom Bennigson, Nike Revisited: Can Commercial Corporations Engage

in Non-Commercial Speech?, 39 Conn. L. Rev. 379, 448 (2006) (inquiring “whether the

First Amendment limits the severity of the penalties that may be imposed for false state-

ments by corporations”); Richard A. Epstein, Was New York Times v. Sullivan Wrong?, 53 U.

Chi. L. Rev. 782, 817 (1986) (“[O]ne clear defect of the actual malice requirement of New

York Times is that by making basic liability turn on subjective intentions, it blurs the line

between actual and punitive damages and thereby unwittingly simplifies the recovery of

the latter.”); Pierre N. Leval, Commentary, The No-Money, No-Fault Libel Suit: Keeping

Sullivan in Its Proper Place, 101 Harv. L. Rev. 1287, 1288 (1988) (“I suggest that recogni-

tion of a no-damages libel suit, free of Sullivan’s actual malice requirement, would im-

prove the efficiency of the cause of action, and reduce its costs and burdens for both de-

fendants and plaintiffs.”); Rosalie Berger Levinson, Superimposing Title VII’s Adverse

Action Requirement on First Amendment Retaliation Claims: A Chilling Prospect for Gov-

ernment Employee Speech, 79 Tul. L. Rev. 669, 693–99 (2005) (discussing emergence of

penalty-sensitive “adverse action” requirement for public employees’ First Amendment

retaliation claims); William W. Van Alstyne, First Amendment Limitations on Recovery

from the Press—An Extended Comment on “The Anderson Solution,” 25 Wm. & Mary L.

Rev. 793, 802-09 (1984) (discussing penalty-sensitive aspects of Holmes’s dissent in Abrams,

as well as problem of punitive damages in defamation proceedings).

2012] OF SPEECH AND SANCTIONS 997

This Article addresses these questions. Pulling together existing

strands of penalty-sensitive doctrine, it argues that, while penalty sensitiv-

ity’s utility will vary from case to case, the approach is fully capable of fur-

thering salutary constitutional objectives. Penalty-sensitive analysis can,

for example, equalize constitutional treatment of similarly situated

speakers, root out improper motives, facilitate “efficient breaches” of

constitutionally problematic speech restrictions, and increase the trans-

parency of judicial decisionmaking. What is more, the potential prob-

lems presented by the penalty-sensitive approach are not unduly severe.

Penalty-sensitive rules and rulings comport with traditional understand-

ings of the judicial role. They are workable in practice. And they do not

threaten to vitiate the free speech guarantee. For all these reasons, courts

should seek to build upon the penalty-sensitive foundations they have al-

ready laid down.

The Article proceeds as follows. Parts I and II advance the descrip-

tive thesis of this Article, with Part I outlining the criteria of penalty-

sensitive analysis and Part II cataloging existing examples of penalty-

sensitive analysis arising within both First Amendment law and case law

involving other constitutional rights. Parts III and IV then defend penalty

sensitivity as a worthwhile form of free speech review. Part III explains

how penalty-sensitive free speech review serves beneficial purposes: pro-

moting fair judicial outcomes, mitigating chilling effects, “preserving”

unprotected speech, screening for improper government motives, and

enhancing transparency. Part IV anticipates and rebuts objections to the

practice, including the claims that it is illegitimate, that it is too difficult

to implement, and that it will undermine the free speech guarantee.

Before proceeding further, let me offer three clarifications regard-

ing the scope of the project. First, this is not an Article about how to con-

ceptualize the relationship between penalty-sensitive doctrine and the

precise meaning of the Free Speech Clause. The descriptive analysis sets

aside the question whether penalty-sensitive rules derive from a straight-

forward reading of the First Amendment’s prohibition on “abridging the

freedom of speech”

or whether penalty-sensitive doctrine is better re-

garded as embodying a set of “constitutional decision rules” designed to

facilitate indirect or prophylactic judicial enforcement of this prohibi-

tion.

And the normative analysis does not address the textual or histori-

25. U.S. Const. amend. I (emphasis added).

26. See Mitchell N. Berman, Constitutional Decision Rules, 90 Va. L. Rev. 1, 15

(2004) (distinguishing between “constitutional operative propositions (judicial statements

of what the Constitution means)” and “constitutional decision rules (judicial statements of

how courts should decide whether the operative propositions have been complied with)”

(emphasis omitted)); Richard H. Fallon, Jr., The Supreme Court, 1996 Term—Foreword:

Implementing the Constitution, 111 Harv. L. Rev. 54, 57 (1997) (arguing “the Court must

often craft doctrine that is driven by the Constitution but does not reflect the

Constitution’s meaning precisely”); Kermit Roosevelt III, Constitutional Calcification: How

the Law Becomes What the Court Does, 91 Va. L. Rev. 1649, 1655–58 (2005) (evaluating

998 COLUMBIA LAW REVIEW [Vol. 112:5

cal legitimacy of penalty-sensitive review. These issues matter, and they

may affect some people’s attitudes toward the penalty-sensitive approach.

But they lie beyond the reach of this Article, which concerns penalty sen-

sitivity’s existing presence in First Amendment doctrine and its useful-

ness as a tool of free speech analysis.

Second, in endorsing penalty sensitivity as a method of First

Amendment analysis, this Article takes no position on whether, as a gen-

eral matter, First Amendment doctrine should be more or less protective

of speech than is currently the case. As Part IV.C explains further,

penalty-sensitive free speech review is capable of both expanding and

shrinking the zone of First Amendment protection, and this Article does

not advocate for one application over the other. That is not to say that

the long-term consequences of incorporating penalty-sensitive review

into First Amendment law are irrelevant to the analysis that follows; any

appraisal of penalty-sensitive review must ask, as this one does, whether

the method will tend to strengthen or weaken the First Amendment’s

rights-vindicating effects.

In the end, however, this Article is about how

to draw First Amendment boundaries, not where those boundaries

should lie.

Finally, the Article’s primary focus on the freedom of speech should

not be taken to imply that free speech cases are uniquely amenable to

the penalty-sensitive approach. If anything, the arguments here may lead

to a different conclusion: namely, that penalty sensitivity is capable of

improving not only free speech doctrine but other areas of constitutional

doctrine as well. One might, for instance, employ penalty-sensitive analy-

sis in cases involving the free exercise right, substantive and procedural

due process rights, and the newly revitalized right to bear arms.

To be

sure, these observations are only preliminary: How, why, and to what ex-

tent penalty-sensitive analysis should occur outside the free speech set-

ting are questions more raised than answered in the discussion that fol-

lows. For now, however, it suffices to note that this Article’s largely posi-

tive story about penalty-sensitive analysis and the freedom of speech sug-

decision rules model); Lawrence Gene Sager, Fair Measure: The Legal Status of Under-

enforced Constitutional Norms, 91 Harv. L. Rev. 1212, 1213 (1978) (“[C]onstitutional

norms which are underenforced by the federal judiciary should be understood to be le-

gally valid to their full conceptual limits, and federal judicial decisions which stop short of

these limits should be understood as delineating only the boundaries of federal courts’

role in enforcing the norm.”); David A. Strauss, The Ubiquity of Prophylactic Rules, 55 U.

Chi. L. Rev. 190, 190 (1988) (claiming “‘prophylactic’ rules are not exceptional measures

of questionable legitimacy but are a central and necessary feature of constitutional law”);

see also Paul Horwitz, Three Faces of Deference, 83 Notre Dame L. Rev. 1061, 1140–46

(2008) (reviewing literature on constitutional decision rules).

27. Cf. infra Part IV.C (responding to concern that, over time, penalty sensitivity may

undermine free speech protections).

28. See District of Columbia v. Heller, 128 S. Ct. 2783, 2821–22 (2008) (holding

Second Amendment protects individual right to bear arms).

2012] OF SPEECH AND SANCTIONS 999

gests that there are similarly positive stories to tell about penalty-sensitive

analysis and other constitutional liberties.

WHAT IS PENALTY SENSITIVITY?

What is penalty-sensitive free speech analysis? And how does it differ

from other styles of constitutional review? These questions do not admit

of easy answers, and this Article does not purport to offer a rigid classifi-

cation scheme. As with most attempts to create categories, dividing lines

become hazy and contestable at the margins, sometimes making the

presence of penalty sensitivity a matter of eye-of-the-beholder judgment.

That said, some general guideposts may help to clarify what sorts of rul-

ings do—and do not—qualify as penalty-sensitive.

First, penalty-sensitive free speech review is a method of implement-

ing the free speech right; it is not a method of applying other constitutional

protections to cases that happen to involve expressive conduct. Penalty-

sensitive analysis derives punitive limits from the free speech right itself—

it does not derive these limits, at least exclusively, from the Cruel and

Unusual Punishments Clause, the Excessive Fines Clause, the Due

Process Clause (including its limits on punitive damages), or any other

punishment-restricting norm within the Constitution. This is not to say

that speech-related punishments could never violate Eighth Amendment

or due process requirements. But were such a violation to occur (as

would happen, for instance, if a trial court sentenced to death a juvenile

purveyor of obscenity

), it would implicate Eighth Amendment propor-

tionality analysis, not penalty-sensitive First Amendment analysis.

Penalty-sensitive First Amendment analysis, by contrast, asks whether a

restriction on speech violates the First Amendment in light of the severity

of the punishment attached to it.

29. See Roper v. Simmons, 543 U.S. 551, 578 (2005) (holding Eighth Amendment

prohibits application of death penalty to minors).

30. That said, penalty-sensitive First Amendment analysis need not ignore the pres-

ence of independent constitutional restrictions on punishment. Indeed, some readers

might be inclined to view the enterprise as grounded in the joint application of the

Constitution’s independent prohibitions on excessive punishment on the one hand, see

Erwin Chemerinsky, The Constitution and Punishment, 56 Stan. L. Rev. 1049, 1052–61

(2004) (comparing Court’s Eighth Amendment limits on criminal punishment with its

due process limits on punitive damages); Pamela S. Karlan, “Pricking the Lines”: The Due

Process Clause, Punitive Damages, and Criminal Punishment, 88 Minn. L. Rev. 880, 883–

914 (2004) [hereinafter Karlan, Pricking the Lines] (similar), and its protection of the

right to free speech on the other. My view is that the penalty-sensitive approach may be

conceptualized as a freestanding method of First Amendment review, untethered to

Eighth Amendment and due process doctrine. Whether this is so, however, is not a ques-

tion of great importance to this project. The important point is that penalty-sensitive re-

view involves analyzing the severity of a law’s penalties in connection with its impact on ex-

pressive conduct—not simply applying Eighth Amendment or Due Process analysis to laws

that happen to regulate expressive activity.

1000 COLUMBIA LAW REVIEW [Vol. 112:5

Second, the term “penalty” in the phrase “penalty-sensitive analysis”

encompasses a wide range of government actions. For purposes of this

Article, government action is said to constitute a penalty as long as it in-

flicts some measure of harm on a speaker as a result of his speaking.

Thus, the ensuing discussion does not confine itself to criminal punish-

ments; it also deals with other sorts of punitive measures, such as civil

penalties, money damage awards, administrative enforcement actions,

and the revocation or withholding of government-sponsored benefits.

Some of these actions are more punitive than others, and some may ac-

complish both punitive and nonpunitive aims. (Civil damages awards, for

example, compensate plaintiffs even as they burden defendants with the

obligation to pay.) But an underlying assumption of this Article is that all

of these actions share a common “penalizing” component—disadvantag-

ing a speaker for speaking—and that this shared component brings them

under the same conceptual umbrella.

Third, penalty-sensitive review posits a positive correlation between

the harshness of the governmental sanction and the strength of the

speaker’s First Amendment claim. Laws should become more, not less,

constitutionally problematic as their penalties increase in severity. Courts

can recognize this correspondence in many different ways. They can ac-

cord dispositive effect to the severity of a speaker’s penalty, commanding

that all penalties harsher than a baseline minimum automatically trigger

a finding of First Amendment invalidity. Or they can treat the severity of

the penalty as a softer variable, which functions as one of many factors

weighing on the outcome of a case. Indeed, as Part IV further discusses,

courts that promulgate penalty-sensitive doctrine can (and do) proceed

down many different paths—relying on categorical penalty-based rules,

semicategorical penalty-based presumptions, penalty-inclusive balancing

tests, and other intermediate approaches. Whatever the path taken, the

analysis counts as penalty-sensitive so long as rising levels in penalty sever-

ity correspond, even if roughly, to rising odds of a judicial finding of con-

stitutional invalidity.

31. This emphasis on penalties also illustrates how penalty-sensitive free speech analy-

sis differs from more traditional forms of time, place, and manner analysis: Time-place-

manner analysis addresses burdens that arise when individuals choose to comply with a

speech restriction, while penalty-sensitive analysis addresses burdens that arise when indi-

viduals fail to comply. If a municipal ordinance prohibited the posting of signs on public

property, see, e.g., Members of City Council of L.A. v. Taxpayers for Vincent, 466 U.S. 789,

791 (1984), time-place-manner analysis would assume full compliance with the ordinance

and then ask whether the individual’s inability to post signs in his yard unduly burdened

his right to free expression. Penalty-sensitive analysis, by contrast, would expand the in-

quiry to include an assessment of the means by which the city punished defiant sign-

posters. The penalty-sensitive analyst might care, for example, about whether the city im-

posed particularly heavy fines on violators of the ordinance or whether the city ever im-

prisoned repeat offenders.

32. Penalty-sensitive analysis is different from the sort of “proportionality analysis”

that has recently attracted the attention of comparative constitutional scholars. Propor-

2012] OF SPEECH AND SANCTIONS 1001

Finally, penalty-sensitive free speech review concerns itself with the

variable of severity. For a decision to qualify as penalty-sensitive, it cannot

lean on just any penalty-related property; rather, the analysis must have

something to do with the extent to which a penalty penalizes. This crite-

rion is less restrictive than it may initially seem. Penalty-sensitive doctrine

need not explicitly compare “severe” against “not severe,” or “more se-

vere” against “less severe”; instead, it may (and often does) accomplish

this comparison through the use of proxies. In Gertz, for example, the

tionality analysis, broadly speaking, involves an explicit weighing of the benefits and harms

created by constitutionally suspect laws, along with a case-by-case investigation into the

closeness of fit between legitimate government aims and the means by which a law

achieves them. See, e.g., David M. Beatty, The Ultimate Rule of Law 159–76 (2004) (cata-

loguing examples of proportionality-oriented rights review, through which judges “assess

the legitimacy of whatever law or regulation or ruling is before them from the perspective

of those who reap its greatest benefits and those who stand to lose the most”); Aharon

Barak, Proportionality and Principled Balancing, 4 L. & Ethics Hum. Rts. 1, 4 (2010) (de-

scribing proportionality analysis as “creat[ing] a conceptual framework in which to define

the appropriate relationship between human rights and considerations that may justify

their limitation”); Alec Stone Sweet & Jud Matthews, Proportionality Balancing and Global

Constitutionalism, 47 Colum. J. Transnat’l L. 72, 73–74 (2008) (characterizing propor-

tionality analysis as “overarching principle of constitutional adjudication, the preferred

procedure for managing disputes involving an alleged conflict between two rights claims,

or between a rights provision and a legitimate state or public interest”). In this sense, it

bears a close kinship to those areas of U.S. constitutional doctrine that eschew absolutist,

category-oriented rules in favor of flexible, balancing-oriented standards. See, e.g., T.

Alexander Aleinikoff, Constitutional Law in the Age of Balancing, 96 Yale L.J. 943, 944–45

(1987) (describing “rise and spread of ‘balancing’ as a method of constitutional interpre-

tation” in America and offering several criticisms); Joseph Blocher, Categoricalism and

Balancing in First and Second Amendment Analysis, 84 N.Y.U. L. Rev. 375, 381–98 (2009)

(summarizing debates between proponents of balancing and categoricalism within First

Amendment law); Laurent B. Frantz, Is the First Amendment Law?—A Reply to Professor

Mendelson, 51 Calif. L. Rev. 729, 729–32 (1963) (contrasting balancing and absolutist ap-

proaches within U.S. constitutional doctrine); Wallace Mendelson, On the Meaning of the

First Amendment: Absolutes in the Balance, 50 Calif. L. Rev. 821, 821 (1962) (similar). See

generally Vicki C. Jackson, Being Proportional About Proportionality, 21 Const. Comment.

803, 832 (2004) (reviewing Beatty, supra) (noting “proportionality analysis shares certain

characteristics of balancing approaches, as both may be distinguished from more categori-

cal or conceptual forms of analysis”).

In a sense, penalty sensitivity may be attractive to proponents of proportionality analy-

sis, as it moves away from a rigid distinction between “punishable” and “unpunishable”

speech and toward a more refined understanding of speech restrictions and the penalties

they employ. But it is important not to overstate the connection here. As this Article later

explains, penalty-sensitive rights review need not devolve into a case-by-case, “totality of the

circumstances”-like weighing of government interests against liberty interests. That is cer-

tainly one way of putting penalty sensitivity to work. But penalty-sensitive doctrine may also

assume a structure more reflective of categoricalism. (A penalty-sensitive rule might, for

instance, divide speech-restricting penalties into a limited number of classifications and

assign to each classification a precise rule of constitutional treatment.) Such an approach

may be less categorical than certain penalty-neutral alternatives, but it may still qualify as

categorical in an absolute sense. As a result, while proportionality analysis can certainly

accommodate penalty-sensitive analysis, one may still accept the penalty-sensitive approach

without abandoning rule-based categoricalism.

1002 COLUMBIA LAW REVIEW [Vol. 112:5

Court disallowed the recovery of punitive damages for certain types of li-

belous speech but in so doing did not grapple directly with the variable

of penalty severity.

Nevertheless, the Gertz rule qualifies as penalty-

sensitive because the in-kind distinction between compensatory and ex-

tracompensatory damages corresponds to an in-degree distinction be-

tween less severe and more severe penalties. (This is so because, all else

equal, jury awards consisting of both actual damages and punitive dam-

ages will exceed awards limited to actual damages.) In short, penalty-

sensitive rules can rely on qualitative distinctions among types of penal-

ties, provided that these distinctions map onto quantitative distinctions in

harshness.

II.

PENALTY SENSITIVITY IN OPERATION

With penalty sensitivity’s definitional criteria in place, this Article

now considers some real world illustrations of the practice. Part II.A

identifies and describes six examples of penalty-sensitive free speech

analysis. It also identifies three “red herrings”—rules and rulings in the

First Amendment area that, though perhaps suggestive of the penalty-

sensitive approach, do not in fact manifest its hallmark traits. Part II.B

then supplements the discussion with examples of penalty-sensitive re-

view from outside the free speech setting. Though not the central focus

of this Article, this case law deserves attention, as it illustrates the range

of doctrinal contexts in which penalty sensitivity has emerged and high-

lights the many different forms that such review might take. The ensuing

discussion thus demonstrates that (a) penalty-sensitive review underlies a

surprising number of free speech rules and rulings and (b) these rules

and rulings are in keeping with many other areas of rights-related doc-

trine. These observations set the stage for the Article’s normative defense

of penalty-sensitive free speech adjudication, which unfolds in Parts III

and IV.

A. Penalty Sensitivity and Free Speech Doctrine

1. Seditious Libel and the Abrams Dissent. — At the center of the pan-

theon of great First Amendment writings stands Oliver Wendell Holmes’s

dissent in Abrams v. United States.

Abrams and his colleagues had written

and distributed leaflets encouraging strikes at ammunition factories in

response to the U.S. government’s intervention in the Russian

Revolution. The trial court had convicted them of violating the 1917

Espionage Act, which prohibited, among other things, “urg[ing], in-

cit[ing] and advocat[ing] curtailment of production of things and prod-

ucts . . . necessary and essential to the prosecution of the war.”

Three

33. See Gertz v. Robert Welch, Inc., 418 U.S. 323, 349–50 (1974).

34. 250 U.S. 616, 624–31 (1919) (Holmes, J., dissenting).

35. Id. at 617 (majority opinion).

2012] OF SPEECH AND SANCTIONS 1003

defendants received sentences of twenty years’ imprisonment, one de-

fendant received fifteen years, and a fifth defendant received three

years.

Abrams marked a change of course for Holmes, who had earlier

voted to uphold Espionage Act prosecutions in Schenck v. United States,

Frohwerk v. United States,

and Debs v. United States.

Why he shifted posi-

tions remains subject to debate,

but one factor may have been the

harshness of the penalties imposed on the Abrams defendants. Holmes

had balked at the sentences in Schenck (six months),

Frohwerk (ten

years),

and Debs (ten years),

but he had voted to uphold them none-

theless.

As Professor David Bogen has observed, however, “[t]he over-

36. Richard Polenberg, Fighting Faiths: The Abrams Case, the Supreme Court, and

Free Speech 145–46 (1987); Stephen M. Feldman, Free Speech, World War I, and Repub-

lican Democracy: The Internal and External Holmes, 6 First Amendment L. Rev. 192, 234

(2008).

37. 249 U.S. 47 (1919).

38. 249 U.S. 204 (1919).

39. 249 U.S. 211 (1919).

40. Scholars disagree as to whether Holmes’s reversal in Abrams was prompted by a

material shift in his free speech philosophy, see, e.g., Geoffrey R. Stone, Perilous Times:

Free Speech in Wartime 198–211 (2004) (discussing Holmes’s correspondence with

Learned Hand about Espionage Act cases and noting “Holmes had begun to rethink the

issue of free speech” by time Abrams came before Court); Feldman, supra note 36, at 229–

44 (describing “transformation” of Holmes’s views on free expression during months pre-

ceding Abrams); Gerald Gunther, Learned Hand and the Origins of Modern First

Amendment Doctrine: Some Fragments of History, 27 Stan. L. Rev. 719, 734–44 (1975)

(similar); G. Edward White, Justice Holmes and the Modernization of Free Speech Juris-

prudence: The Human Dimension, 80 Calif. L. Rev. 391, 433–42 (1992) (similar), or

whether Holmes had simply been “biding his time until the Court should have before it a

conviction so clearly wrong as to let him speak out his deepest thoughts about the First

Amendment,” Zechariah Chafee, Free Speech in the United States 86 (1942); see also

Sheldon M. Novick, The Unrevised Holmes and Freedom of Expression, 1991 Sup. Ct.

Rev. 303, 335–50 (claiming Holmes’s views on free speech did not change between Schenck

and Abrams). See generally Kathleen M. Sullivan & Gerald Gunther, Constitutional Law

762–63 (16th ed. 2007) (summarizing debate).

41. Polenberg, supra note 36, at 212–13. Schenck’s accomplice, Dr. Elizabeth Baer,

received a sentence of ninety days. Id.

42. 249 U.S. at 206.

43. 249 U.S. at 212.

44. See, e.g., Letter from Oliver Wendell Holmes to Frederick Pollock (Apr. 27,

1919), in 2 Holmes-Pollock Letters: The Correspondence of Mr. Justice Homes and Sir

Frederick Pollock 1874–1932, at 11 (Mark DeWolfe Howe ed., 1942) (“Now I hope the

President will pardon [Debs] and some other poor devils with whom I have more sympa-

thy.”); see also David S. Bogen, The Free Speech Metamorphosis of Justice Holmes, 11

Hofstra L. Rev. 97, 183 (1983) (noting “Holmes had been upset by the prosecutions in the

earlier cases and stated to his friends that the sentences there should be sharply reduced

and commuted by the President”). Holmes’s feelings on this score were by no means

unique to him. As a roughly contemporaneous account of the cases explained, “[m]uch of

the bitter feeling which these cases have engendered in the United States is due to the

manner in which the trials were conducted . . . and to the length of the sentences.” Notes, 42

1004 COLUMBIA LAW REVIEW [Vol. 112:5

reaction of the judges and juries that . . . [Holmes had] perceived in . . .

[Schenck, Frohwerk, and Debs] was mild compared to the fervor exhibited

against the Abrams defendants in sentencing them to a twenty-year im-

prisonment.”

What is more, whereas “Debs and Frohwerk addressed

significant sympathetic groups and Schenck addressed his message to

persons who might respond from self-interest,”

the Abrams defendants

were, as Zechariah Chafee would later describe them, “five obscure and

isolated youngsters . . . who hatched their wild scheme in a garret and

carried it out in a cellar.”

In short, the punishments at issue in Abrams—

far more so than the punishments at issue in its predecessor cases—were

strikingly disproportionate to the defendants’ underlying crimes.

The heaviness of these sanctions did not escape Holmes’s notice,

and his dissent in Abrams drew attention to their severity. Early in his

opinion, Holmes had speculated that the First Amendment might not

prohibit the government from prosecuting attempts to disrupt the pro-

duction of wartime goods, provided that the government could show that

a defendant specifically intended to achieve that result.

Holmes con-

cluded that the government had not proven specific intent in this case,

but his analysis did not stop there. He explained,

In this case sentences of twenty years imprisonment have been

imposed for the publishing of two leaflets that I believe the de-

fendants had as much right to publish as the Government has to

publish the Constitution of the United States now vainly in-

voked by them. Even if I am technically wrong and enough can

be squeezed from these poor and puny anonymities to turn the

color of legal litmus paper; I will add, even if what I think the nec-

essary intent were shown; the most nominal punishment seems to me all

that possibly could be inflicted, unless the defendants are to be

made to suffer not for what the indictment alleges but for the

creed that they avow . . . .

In other words, even if Abrams and his collaborators had specifically

intended to curtail wartime production (and thereby surrendered their

right to absolute constitutional protection), the First Amendment would

L.Q. Rev. 1, 14 (1926) (emphasis added).

45. Bogen, supra note 44, at 183.

46. Id. at 182. Charles Schenck was the Secretary of the Socialist Party of America at

the time of his prosecution. Schenck v. United States, 249 U.S. 47, 49 (1919). Eugene Debs

was a founding member of the Industrial Workers of the World and frequent presidential

candidate for the Socialist Party of America. See Kermit L. Hall & John J. Patrick, The

Pursuit of Justice: Supreme Court Decisions that Shaped America 78–79 (2006); Novick,

supra note 40, at 341.

47. Zechariah Chafee, A Contemporary State Trial—The United States Versus Jacob

Abrams et al., 33 Harv. L. Rev. 747, 773 (1920); see also Bogen, supra note 44, at 182

(“Abrams and his codefendants scattered to the winds a message, in a foreign language,

that no one had a personal interest in accepting.”).

48. Abrams v. United States, 250 U.S. 616, 628 (1919) (Holmes, J., dissenting).

49. Id. at 629 (emphasis added).

2012] OF SPEECH AND SANCTIONS 1005

still have prohibited the trial court from imposing a twenty-year sentence.

Indeed, for Holmes, only “the most nominal punishment” would have

passed constitutional muster.

The critical point is not hard to see. As Thomas Reed Powell later

explained, Holmes’s dissent put forth “the suggestion that the First

Amendment limits the degree of punishment for speech concededly

punishable.”

Many years later, David Currie offered the same

interpretation, crediting this passage with the “the novel and interesting

notion that the first amendment . . . limited the punishment that could

be imposed even for speech not wholly protected.”

2. Defamation. — Penalty sensitivity was limited to a dissenting opin-

ion in Abrams, but it later made its way into several majority opinions—

and hence, operative First Amendment law. A good example comes from

the defamation context. Gertz v. Robert Welch, Inc.

followed in the after-

math of New York Times Co. v. Sullivan, the first Supreme Court case to

recognize First Amendment limits on state court defamation judg-

ments.

In New York Times, the Court held that a “public official” could

not “recover[] damages for a defamatory falsehood relating to his official

conduct unless he proves that the statement was made with ‘actual mal-

ice’—that is, with knowledge that it was false or with reckless disregard of

whether it was false or not.”

In so holding, the Court left many ques-

50. Id. Holmes’s Abrams dissent, moreover, does not present the only example of

penalty-sensitive musings from this era. Whitney v. California, decided eight years after

Abrams, involved a prosecution under California’s Criminal Syndicalism Act. 274 U.S. 357

(1927). The Court unanimously rejected the defendant’s constitutional challenges, but

Justice Brandeis, joined by Justice Holmes, concurred separately, expressing concern with

the majority’s analysis of the free speech claim. Brandeis argued for strict application of

the “clear and present danger” test, which would limit the government’s power to regulate

seditious libel to situations in which “immediate serious violence was to be expected or was

advocated, or . . . [where] the past conduct furnished reason to believe that such advocacy

was then contemplated.” Id. at 376 (Brandeis, J., concurring). But the opinion also ap-

peared to endorse Holmes’s earlier suggestion that the clear and present danger test

should apply in a penalty-sensitive manner. Specifically, Brandeis noted that “[a] police

measure may be unconstitutional merely because the remedy, although effective as means

of protection, is unduly harsh or oppressive,” id. at 377, and that, while the state could cer-

tainly criminalize trespass, “it is hardly conceivable that this Court would hold constitu-

tional a statute which punished as a felony the mere voluntary assembly with a society

formed to teach that pedestrians had the moral right to cross unenclosed, unposted, waste

lands and to advocate their doing so,” id. at 378 (emphasis added).

51. Thomas Reed Powell, Constitutional Law in 1919–1920 (pt. 3), 19 Mich. L. Rev.

283, 292 (1920). Powell went on to wonder whether “Mr. Justice Holmes has the due-

process clause of the Fifth Amendment in mind.” Id.

52. David P. Currie, The Constitution in the Supreme Court: 1910–1921, 1985 Duke

L.J. 1111, 1154–55 n.225.

53. 418 U.S. 323 (1974).

54. 376 U.S. 254, 283 (1964) (holding “the Constitution delimits a State’s power to

award damages for libel in actions brought by public officials against critics of their official

conduct”).

55. Id. at 279–80.

1006 COLUMBIA LAW REVIEW [Vol. 112:5

tions unanswered, one of which was whether the “actual malice” rule ap-

plied to plaintiffs other than public officials. A subsequent decision,

Curtis Publishing Co. v. Butts, extended the rule to cover defamatory

speech about “public figures.”

Gertz presented the question whether the

rule should also apply to defamatory speech about “private citizens”

caught up in “public issues.”

Rosenbloom v. Metromedia, Inc., decided prior to Gertz, had posed the

same question, but there a fractured Court failed to answer it.

The

Rosenbloom plurality favored furnishing full New York Times protection to

“all discussion and communication involving matters of public or general

concern, without regard to whether the persons involved are famous or

anonymous.”

But other Justices put differing options on the table. Of

particular note, Justices Marshall and Harlan, in separate dissents, out-

lined approaches that would have (a) relaxed the “actual malice” re-

quirement for defamation actions brought by private-figure plaintiffs but

(b) limited the amount of damages that private-figure plaintiffs could re-

cover.

As for how to implement this second part of the plan, the two

Justices had different ideas. Marshall favored permitting damages for “ac-

tual” injury but forbidding presumed and punitive damages across the

board.

Harlan, in contrast, would have permitted punitive damages but

only when “actual malice” was shown and the damages bore “a reason-

able and purposeful relationship to the actual harm done.”

Notwith-

standing these disagreements as to details, both Justices thus endorsed a

penalty-sensitive rule.

Three years later, revisiting the issue left open by Rosenbloom, the

Gertz majority built on Marshall’s and Harlan’s earlier proposals. The

Court held (a) that private citizens could recover damages for negli-

gently published defamatory falsehoods, but they could not recover any-

thing beyond compensation for actual injury and (b) that presumed and

punitive damages were off limits, unless the plaintiffs could satisfy the

56. 388 U.S. 130, 163 (1967) (Warren, C.J., concurring in result); see also Gertz, 418

U.S. at 336 (noting “[a]lthough Mr. Justice Harlan announced the result in [Butts], a ma-

jority of the Court agreed with Mr. Chief Justice Warren’s conclusion that the New York

Times test should apply to criticism of ‘public figures’ as well as ‘public officials’”).

57. 418 U.S. at 325 (considering “the extent of a publisher’s constitutional privilege

against liability for defamation of a private citizen”).

58. 403 U.S. 29 (1971).

59. Id. at 44 (plurality opinion).

60. See id. at 86 (Marshall, J., dissenting) (advocating approach that would “restrict

damages to actual losses”); id. at 77 (Harlan, J., dissenting) (“I would hold unconstitu-

tional, in a private libel case, jury authority to award punitive damages which is unconfined

by the requirement that these awards bear a reasonable and purposeful relationship to the

actual harm done.”).

61. See id. at 84 (Marshall, J., dissenting) (“The threats to society’s interest in free-

dom of the press that are involved in punitive and presumed damages can largely be

eliminated by restricting the award of damages to proved, actual injuries.”).

62. Id. at 77 (Harlan, J., dissenting).

2012] OF SPEECH AND SANCTIONS 1007

strict actual-malice standard set forth in New York Times.

Writing for the

majority in Gertz, Justice Powell defended this rule in penalty-sensitive

terms.

He reaffirmed the Court’s recognition of the “legitimate state in-

terest in compensating private individuals for injury to reputation,” but

he also stipulated that “this countervailing state interest extends no fur-

ther than compensation for actual injury.”

Anything beyond this, he

reasoned, would pose a special threat to free expression. With presumed

damages, juries could award “substantial sums as compensation for sup-

posed damage to reputation without any proof that such harm actually

occurred.”

This in turn would “compound[] the potential of any system

of liability for defamatory falsehood to inhibit the vigorous exercise of

First Amendment freedoms.”

And with punitive damages, “jury discre-

tion over the amounts awarded is limited only by the gentle rule that they

not be excessive.”

Consequently, juries could award these damages “in

wholly unpredictable amounts,” which could “unnecessarily exacerbat[e]

the danger of media self-censorship.”

The bottom line for the Gertz majority was that neither presumed

nor punitive damages substantially furthered the state’s interest in com-

pensating injury and both types of damages threatened unpopular ex-

pression with severe forms of punishment. Hence the need for a penalty-

sensitive rule—that is, a rule that could accommodate the state’s interest

in making victims whole, while staving off the specter of sky-high damage

awards.

The penalty-sensitive nature of the rule is thus apparent. The

Court in Gertz did not limit the state’s power to impose liability; it did,

however, limit the amount of damages that the state could attach to this

liability. And although the scope of the rule has been refined in later

cases,

its penalty-sensitive framework remains firmly intact. When the

rule applies, speech may be penalized but only to a limited degree.

63. Gertz v. Robert Welch, Inc., 418 U.S. 323, 350 (1974) (“[A] private defamation

plaintiff who establishes liability under a less demanding standard than that stated by New

York Times may recover only such damages as are sufficient to compensate him for actual

injury.”).

64. Id. at 343.

65. Id. at 348–49.

66. Id. at 349; see also T. Michael Mather, Experience With Gertz “Actual Injury” in

Defamation Cases, 38 Baylor L. Rev. 917, 949 (1986) (noting “presumed damage cases

have substantially higher compensatory damage awards on average than actual injury

compensatory damage awards”).

67. 418 U.S. at 349.

68. Id. at 350.

69. Id.

70. Whether the Gertz rule achieved its objective of reducing damage awards is not

altogether clear. See, e.g., Gerald G. Ashdown, Gertz and Firestone: A Study in Constitu-

tional Policy-Making, 61 Minn. L. Rev. 645, 669–72 (1977) (arguing Gertz rule still allows

exorbitant damages awards).

71. See Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc., 472 U.S. 749, 761–62

(1985) (holding Gertz inapplicable when speech at issue does not involve “a matter of pub-

1008 COLUMBIA LAW REVIEW [Vol. 112:5

3. Public Employment. — Penalty sensitivity has also materialized

within case law involving the speech of public employees. The penalties

at issue here take the form of disciplinary measures that supervisors di-

rect at their subordinates. These measures vary greatly in their harshness:

At one end of the spectrum, disciplinary action might entail nothing

more than a verbal reprimand; at the other, it can amount to outright

dismissal. Given this wide range of severity levels, one might expect the

character of an employer’s disciplinary action to figure prominently in

judicial analysis of public employees’ free speech claims. It would not be

implausible, for example, to draw a constitutional distinction between

cases in which employers punish speech with short suspensions and cases

in which employers punish the same speech with outright termination.

For a number of years, some lower courts assumed as much, inter-

preting Supreme Court case law to mean that free speech violations were

more likely to occur when the government fired its employees for speech-

related reasons.

But in Rutan v. Republican Party of Illinois, the Court jet-

tisoned this approach.

Rutan involved a claim by Illinois civil service em-

ployees that their professional situations had suffered on account of their

political affiliation. Notably, none of these employees had been fired;

rather, they had suffered such lesser harms as promotion denials and

targeted exclusions from post-layoff recall appointments.

The Seventh Circuit analyzed the case by asking whether these em-

ployment actions amounted to the “substantial equivalent of a dis-

missal.”

This approach seemed reasonable under then-existing prece-

lic concern”).

72. Given the penalty-sensitive nature of the Gertz decision, it is perhaps surprising

that the Court has not drawn a similar constitutional distinction between civil and criminal

libel laws, subjecting the latter to something even more restrictive than the New York Times

standard. Cf. Thomas I. Emerson, Toward a General Theory of the First Amendment, 72

Yale L.J. 877, 924 (1963) [hereinafter Emerson, Toward a General Theory] (arguing

criminal libel laws should be deemed unconstitutional, in part because “in the context of a

public prosecution the dangers in punishing expression under the necessarily loose stan-

dards invoked are . . . great” (footnote omitted)). Indeed, the Court had an opportunity to

draw just such a penalty-sensitive distinction in a companion case to New York Times,

Garrison v. Louisiana, but it instead chose to apply the same standard to both types of re-

strictions, without suggesting that governments might enjoy less constitutional leeway

when regulating defamation through criminal means. 379 U.S. 64, 76 (1964). In New York

Times, the Court explained,

What a State may not constitutionally bring about by means of a criminal statute

is likewise beyond the reach of its civil law of libel. The fear of damage awards

under a rule such as that invoked by the Alabama courts here may be markedly

more inhibiting than the fear of prosecution under a criminal statute.

N.Y. Times Co. v. Sullivan, 376 U.S. 254, 277 (1964) (footnote omitted).

73. See, e.g., Delong v. United States, 621 F.2d 618, 623–24 (4th Cir. 1980) (confin-

ing inquiry to dismissal of employee and employment practices “that can be determined to

be the substantial equivalent of dismissal”).

74. 497 U.S. 62, 72 (1990).

75. Rutan v. Republican Party of Ill., 868 F.2d 943, 956 (7th Cir. 1989) (en banc),

2012] OF SPEECH AND SANCTIONS 1009

dent,

but the Supreme Court rejected it in no uncertain terms.

Whether or not substantially equivalent to dismissals, Justice Brennan

declared for the majority, the actions represented “significant penalties”

that “impermissibly encroach[ed] on First Amendment freedoms.”

continued,

We . . . determine that promotions, transfers, and recalls after

layoffs based on political affiliation or support are an impermis-

sible infringement on the First Amendment rights of public

employees. In doing so, we reject the Seventh Circuit’s view of

the appropriate constitutional standard by which to measure al-

leged patronage practices in government employment. The

Seventh Circuit proposed that only those employment decisions

that are the “substantial equivalent of a dismissal” violate a pub-

lic employee’s rights under the First Amendment. We find this

test unduly restrictive because it fails to recognize that there are

deprivations less harsh than dismissal that nevertheless press

state employees and applicants to conform their beliefs and as-

sociations to some state-selected orthodoxy.

Indeed, Justice Brennan went further, proclaiming that “the First

Amendment . . . protects state employees from ‘even an act of retaliation

as trivial as failing to hold a birthday party for a public employee . . .

when intended to punish her for exercising her free speech rights.’”

On his view, any form of retaliation, no matter how minimal, would seem

to violate the First Amendment. His stance was avowedly penalty-neutral.

Yet even the far-reaching rhetoric of Rutan has not erased penalty

sensitivity from the public employment picture. With the discharge/

nondischarge distinction now a thing of the past, some lower courts, re-

luctant to follow Rutan to its logical conclusion, have developed a differ-

ent set of punishment-based boundaries. Dismissing the “birthday party”

language as “colorful dicta,”

these courts have declared that not every

form of speech-based retaliation can violate an employee’s free speech

rights. Instead, they have held that a First Amendment violation cannot

materialize unless the employer has taken an “adverse employment ac-

tion” against the employee-speaker.

In so doing, they have drawn from

aff’d in part, rev’d in part, 497 U.S. 62.

76. See, e.g., Branti v. Finkel, 445 U.S. 507, 515 (1980) (holding employee may not

be discharged solely on basis of political beliefs); Elrod v. Burns, 427 U.S. 347, 360 (1976)

(holding patronage dismissals due to political affiliation violated First Amendment).

77. 497 U.S. at 74–75.

78. Id. at 75 (citing 868 F.2d at 954–57).

79. Id. at 75 n.8 (quoting 868 F.2d at 954 n.4).

80. Lybrook v. Members of Farmington Mun. Sch. Bd. of Ed., 232 F.3d 1334, 1340

(10th Cir. 2000) (noting “[a]lthough . . . employers’ acts short of dismissal may be action-

able as First Amendment violations, we have never ruled that all such acts, no matter how

trivial, are sufficient to support a retaliation claim”).

81. See, e.g., id.; Suppan v. Dadonna, 203 F.3d 228, 235 (3d Cir. 2000) (arguing “[i]t

would trivialize the First Amendment to hold that harassment for exercising the right of

1010 COLUMBIA LAW REVIEW [Vol. 112:5

the Supreme Court’s Title VII precedents, which make clear that an “ad-

verse employment action” requirement envisions “materially adverse” ac-

tion, thereby “separat[ing] significant from trivial harms.”

These courts,

in other words, have not abandoned penalty sensitivity in Rutan’s wake.

Rather, they have simply lowered the penalty-sensitive threshold—replac-

ing the old discharge/nondischarge distinction with a new distinction

between materially and nonmaterially adverse employment actions.

The new threshold is more demanding than might first meet the

eye. Consider the Tenth Circuit’s decision in Lybrook v. Members of

Farmington Municipal School Board of Education.

The court there rejected

the First Amendment claim of a school teacher whose supervisors had

placed her on a “Professional Development Plan” and required her to at-

tend weekly one-on-one meetings with the principal.

Lybrook lost her

case, but not because the court deemed her speech unworthy of protec-

tion. Rather, her problem lay in her penalty—a penalty that, in the

court’s view, did not rise to the level of an “adverse or detrimental em-

ployment decision.”

In particular, “while the Professional Development

Plan and the Monday morning meetings may have been unwelcome[] to

her,” Lybrook’s punishment was “of insufficient gravity to premise a First

Amendment violation.”

Lybrook’s penalty was hardly the stuff of Hammurabi’s Code; one

can imagine fates far worse than participating in professionalism training

and meeting weekly with the principal. But it was a penalty all the same.

The school made Lybrook do things she did not want to do, and it took

this action because Lybrook had engaged in a form of expressive con-

duct. What drove the Lybrook analysis, then, was not a conclusion regard-

ing the value of Lybrook’s speech, but the fact that the state’s chosen

means of punishment was an especially lenient one. For this reason, the

Lybrook decision and the adverse action rule it reflects qualify as penalty-

sensitive.

4. Vagueness/Overbreadth. — The doctrines discussed in the preceding

pages draw substantive boundaries between types of speech. The vague-

ness and overbreadth doctrines, by contrast, impose procedural limits on

lawmakers’ ability to propound restrictions on speech. Vagueness rules

free speech was always actionable no matter how unlikely to deter a person of ordinary

firmness from that exercise”); Pierce v. Tex. Dep’t of Criminal Justice, 37 F.3d 1146, 1150

n.1 (5th Cir. 1994) (explaining Fifth Circuit has “require[d] more than a trivial act to es-

tablish constitutional harm”).

82. Burlington N. & Santa Fe Ry. Co. v. White, 548 U.S. 53, 68 (2006) (emphasis

added).

83. 232 F.3d at 1334.

84. Id. at 1337. The plan required Lybrook “to ‘[s]trive to create an atmosphere that

will nurture collaboration with all colleagues’” and “‘to conduct affairs with a conscious

concern for the highest standards of professional commitment.’” Id.

85. Id. at 1339.

86. Id. at 1341.

2012] OF SPEECH AND SANCTIONS 1011

permit facial invalidation of laws that set forth imprecise standards of le-

gality, while overbreadth rules permit facial invalidations of laws that re-

strict both protected and unprotected forms of expression.

A brief look

at both doctrines reveals further examples of penalty-sensitive review.

Reflecting penalty sensitivity within First Amendment vagueness doc-

trine is the Court’s “greater tolerance of enactments with civil rather

than criminal penalties.”

This “greater tolerance” arises from the fact

that “the consequences of imprecision are qualitatively less severe” in the

civil context (where punishments generally involve nothing worse than

monetary fines),

than in the criminal context (where punishments can

entail lengthy prison sentences, and where the collateral effects of prose-

cution and conviction are of the highest order).

Consistent with this notion is the Court’s decision in NEA v. Finley.

This case presented the question whether Congress violated First

Amendment vagueness rules when it instructed officials at the National

Endowment for the Arts (NEA) to “tak[e] into consideration general

standards of decency” when evaluating grant applications.

The Ninth

87. Though rooted in the Fifth and Fourteenth Amendments’ due process guaran-

tees, void-for-vagueness review is traditionally understood to furnish a special sort of pro-

tection to First Amendment rights. This is because the Court has long adhered to the rule

that “[w]here a statute’s literal scope, unaided by a narrowing state court interpretation, is

capable of reaching expression sheltered by the First Amendment, the doctrine demands

a greater degree of specificity than in other contexts.” Smith v. Goguen, 415 U.S. 566, 573

(1974); see also Cramp v. Bd. of Pub. Instruction, 368 U.S. 278, 287 (1961) (“The vice of

unconstitutional vagueness is further aggravated where, as here, the statute in question

operates to inhibit the exercise of individual freedoms affirmatively protected by the

Constitution.”); Smith v. California, 361 U.S. 147, 151 (1959) (“[T]his Court has intimated

that stricter standards of permissible statutory vagueness may be applied to a statute having

a potentially inhibiting effect on speech; a man may the less be required to act at his peril

here, because the free dissemination of ideas may be the loser.”).

88. Village of Hoffman Estates v. Flipside, Hoffman Estates, Inc., 455 U.S. 489, 499

(1982).

89. Id. at 499.

90. See id.; Barenblatt v. United States, 360 U.S. 109, 137 (1959) (Black, J., dissent-

ing) (observing “it would be unthinkable to convict a man for violating a law he could not

understand”); Winters v. New York, 333 U.S. 507, 515 (1948) (“The standards of certainty

in statutes punishing for offenses is higher than in those depending primarily upon civil

sanction for enforcement.”); see also Maldonado v. Morales, 556 F.3d 1037, 1045 (9th Cir.

2009) (noting “courts must take extra care in determining whether criminal statutes are

vague”); Aaron Xavier Fellmeth, Civil and Criminal Sanctions in the Constitution and

Courts, 94 Geo. L.J. 1, 2–3 (2005) (observing criminal sanction “may impose a social

stigma that permanently impairs the defendant’s quality of life in a manner rarely equaled

by a civil judgment”); Frederick Schauer, Fear, Risk and the First Amendment: Unraveling

the Chilling Effect, 58 B.U. L. Rev. 685, 697 (1978) (“The possibility of imprisonment

coupled with the stigma and disabilities which accompany a criminal conviction will most

often lead an individual to view the criminal penalty as more harmful than a civil sanc-

tion.”).

91. 524 U.S. 569 (1998).

92. Id.

1012 COLUMBIA LAW REVIEW [Vol. 112:5

Circuit answered in the affirmative, characterizing the standard as “not

susceptible to objective definition” and thus highly vulnerable “to the

danger of arbitrary and discriminatory application.”

The Supreme

Court reversed. While never rejecting the lower court’s characterization

of the statutory language as deeply indeterminate, the Court explained

that the Ninth Circuit had applied an unduly stringent standard of preci-

sion. As Justice O’Connor explained, “[t]he terms of the provision are

undeniably opaque, and if they appeared in a criminal statute or regula-

tory scheme, they could raise substantial vagueness concerns.”

This

case, however, did not involve a “criminal statute or regulatory scheme,”

and so “the consequences of imprecision [we]re not constitutionally se-

vere.”

Several considerations supported the majority’s position, including,

for example, the voluntariness of the artists’ decision to submit them-

selves to the guidelines, the inherent subjectivity of artistic judgments,

and the inadvisability of second guessing the NEA’s (and Congress’s)

chosen standards. In addition to these points, however, Justice O’Connor

emphasized that, given the “penalties” attached to the law, “[i]t is

unlikely . . . that speakers will be compelled to steer too far clear of any

‘forbidden area’ in the context of grants of this nature.”

This argument

makes sense only in penalty-sensitive terms. Grant applicants are less

likely to “steer far clear of any forbidden area” because the consequence

of entering the area—i.e., the penalty imposed on the speaker—is merely

the disappointment of not winning the grant. By contrast, the subjects of

an equally vague criminal regulation, faced with the prospect of fines or

even terms in prison, would be sure to err on the side of caution.

Similar thinking drove Justice Kennedy’s overbreadth analysis in

Ashcroft v. Free Speech Coalition.

The Child Pornography Protection Act

(CPPA) criminalized the possession and distribution of images that ap-

peared to depict—but in fact did not depict—minors engaged in sexually

explicit conduct.

The government defended the Act as a constitution-

ally permissible restriction of child pornography under New York v.

Ferber

and, in the alternative, a constitutionally permissible restriction of

obscenity under Miller v. California.

100

The Court, however, rejected both

contentions, deeming Ferber inapplicable to “virtual” child pornogra-

93. Finley v. NEA, 100 F.3d 671, 680 (9th Cir. 1996), rev’d, 524 U.S. 569.

94. 524 U.S. at 588.

95. Id at 588–89.

96. Id. at 588.

97. 535 U.S. 234 (2002).

98. Id. at 239.

99. 458 U.S. 747 (1982) (holding child pornography not generally entitled to First

Amendment protection).

100. 413 U.S. 15 (1973) (establishing guidelines for defining obscenity).

2012] OF SPEECH AND SANCTIONS 1013

phy,

101

and finding only a small portion of the speech criminalized by the

CPPA to qualify as obscene under Miller.

102

This left open the question

whether the Court should invalidate the CPPA on its face, or only as ap-

plied to speech not punishable under Miller.

Opting for facial invalidation, Justice Kennedy stressed the severity

of the CPPA’s penalties:

The CPPA’s penalties are indeed severe. A first offender may be

imprisoned for 15 years. A repeat offender faces a prison sen-

tence of not less than 5 years and not more than 30 years in

prison. While even minor punishments can chill protected

speech, this case provides a textbook example of why we permit

facial challenges to statutes that burden expression. With these

severe penalties in force, few legitimate movie producers or

book publishers, or few other speakers in any capacity, would

risk distributing images in or near the uncertain reach of this

law.

103

Accordingly, the Court struck down the CPPA in its entirety, point-

ing not just to the overbreadth of its coverage, but also to what might be

fairly called the “overdepth” of the penalties imposed.

104

5. Indecency. — An additional area of penalty-sensitive doctrine in-

volves “indecent” expression and the communications media. The key

case is FCC v. Pacifica Foundation.

105

A New York radio station had broad-

cast George Carlin’s “Seven Dirty Words” monologue during a weekday

afternoon, thereby prompting an angry listener to complain to the

FCC.

106

The FCC responded by issuing a declaratory order, which clari-

fied that, but for the good graces of the agency, the station’s broadcast

“could have been the subject of administrative sanctions,” and warned

that “in the event that subsequent complaints are received, the

Commission will then decide whether it should utilize any of the avail-

able sanctions it has been granted by Congress.”

107

The station owner

challenged the order, claiming that it violated the station’s free speech

rights. The Court found in favor of the FCC.

101. Free Speech Coal., 535 U.S. at 249–51.

102. Id. at 246–49.

103. Id. at 244 (citations omitted).

104. Interestingly, Justice O’Connor, who had earlier recognized a penalty-based dis-

tinction in Finley, refused to go along with Kennedy’s overbreadth analysis, and left unad-

dressed his claims about the CPPA’s punishments. Id. at 265–66 (O’Connor, J., concurring

in part and dissenting in part).

105. 438 U.S. 726 (1978).

106. The letter, incidentally, utilized four of the seven words but only for the purpose

of recounting the content of Carlin’s monologue. Angela J. Campbell, Pacifica Reconsid-

ered: Implications for the Current Controversy over Broadcast Indecency, 63 Fed. Comm.

L.J. 195, 201–02 (2010) (reproducing letter).

107. 438 U.S. at 730.

1014 COLUMBIA LAW REVIEW [Vol. 112:5

The Court’s analysis in Pacifica was fact-specific. In a section of the

opinion agreed to by a majority of the Court, Justice Stevens emphasized

the “narrowness” of the holding—pointing out, for example, that the

Court had not decided whether the FCC could apply similar indecency

standards to “a two-way radio conversation between a cab driver and a

dispatcher, or a telecast of an Elizabethan comedy.”

108

Of particular im-

portance, the Carlin monologue teemed with profanities; thus, Pacifica

did not address whether the FCC could sanction a broadcast in which

expletives made only a passing appearance.

109

In addition, the broadcast

had aired at an hour when children were likely to be tuned in; thus, the

Court did not consider whether the FCC could target equally profane

broadcasts that aired late at night.

110

All of these factors played a role in the Court’s analysis in Pacifica.

But another factor mattered, too: As the Court explained, the penalty at

issue was nothing more than a shot across the bow. The FCC had asserted

the authority to impose sanctions, but it had not exercised that authority

directly. Accordingly, the Court declined to decide whether “an occa-

sional expletive [in different settings] would justify any sanction, or, in-

deed, [whether] this broadcast would justify a criminal prosecution.”

111

The Court elaborated no further on this point, but its message was clear:

The government’s power to punish indecency on the airwaves was not

unlimited. Criminal punishments would bear a higher constitutional

burden, and so too might harsher civil sanctions.

112

Indeed, at least two

lower court cases later read the case in just this way. In Carlin

Communications, Inc. v. FCC, the Second Circuit relied on Pacifica’s

“criminal prosecution” language to hold that a criminal statute would be

108. Id. at 750.

109. Id.

110. Id.

111. Id. This qualification may have been included to help bring the Pacifica holding

in line with the Court’s earlier assertion—in Cohen v. California—that “the State may not,

consistently with the First and Fourteenth Amendments, make the simple public dis-

play . . . of [profanity] a criminal offense.” 403 U.S. 15, 26 (1971) (emphasis added). As

Justice Stevens explained, “It should be noted that the Commission imposed a far more

moderate penalty on Pacifica than the state court imposed on Cohen. Even the strongest

civil penalty at the Commission’s command does not include criminal prosecution.”

Pacifica, 438 U.S. at 747 n.25.

112. In Smith v. United States, 431 U.S. 291 (1977), decided one year prior to Pacifica,

Justice Stevens had pushed for a similar approach to the regulation of obscenity. Dissent-

ing from the majority’s opinion in that case, Justice Stevens argued that criminal and civil

obscenity laws should trigger application of different constitutional standards, noting in

particular that

[a]lthough the variable nature of a standard dependent on local community atti-

tudes is critically defective when used to define a federal crime, that very flexibil-

ity is a desirable feature of a civil rule designed to protect the individual’s right

to select the kind of environment in which he wants to live.

Id. at 317 (Stevens, J., dissenting).

2012] OF SPEECH AND SANCTIONS 1015

unconstitutional if its prohibitions were construed to encompass the

transmission of indecent material.

113

And in Community Television of Utah,

Inc. v. Wilkinson, a federal district court interpreted the same passage to

mean that “[i]f the penalty [in Pacifica] had been greater the result in

the case may have been different.”

114

The Court reaffirmed the penalty-sensitive aspects of its Pacifica rul-

ing in Reno v. ACLU.

115

There, the Court passed judgment on the

Communications Decency Act (CDA), which criminalized the online

transmission of obscene and indecent materials to minors. Arguing

against invalidation, the government relied on Pacifica, noting that there

the Court had withheld constitutional protection from indecent speech.

The Court, however, rejected this analysis, concluding that a close look at

Pacifica “raises—rather than relieves—doubts concerning the constitu-

tionality of the CDA.”

116

In particular, among the “significant differences

between the order upheld in Pacifica and the CDA” was one related to

penalties: Unlike the CDA, which called for punishments of up to two

years in prison, the “declaratory order [in Pacifica] was not punitive.”

117

This difference was significant, the Court explained, because in Pacifica it

had “expressly refused to decide whether the indecent broadcast ‘would

justify a criminal prosecution.’”

118

6. School Speech. — One further example of penalty-sensitive free

speech review comes from Justice Brennan’s concurrence in Bethel School

District v. Fraser.

119

Bethel asked whether the First Amendment permitted a

public high school to suspend a student for delivering an on-campus

speech rife with sexual double entendres. The Ninth Circuit had an-

swered this question in the negative, holding that the Court’s earlier de-

cision in Tinker v. Des Moines Independent Community School District

120

pro-

hibited the school from disciplining the speaker, Matthew Fraser, in any

way at all.

121

The Supreme Court reversed, with a majority of the Justices

asserting that the “School District acted entirely within its permissible

authority in imposing sanctions upon Fraser.”

122

Thus, while reaching op-

113. 837 F.2d 546, 560 (2d Cir. 1988) (“Pacifica did not decide that the indecent

broadcast ‘would justify a criminal prosecution.’ Here the statute specifically authorizes

prosecution. Were the term ‘indecent’ to be given meaning other than Miller obscenity, we

believe the statute would be unconstitutional.” (quoting 438 U.S. at 750)).

114. 611 F. Supp. 1099, 1111 (D. Utah 1985).

115. 521 U.S. 844 (1997).

116. Id. at 864.

117. Id. at 867.

118. Id. (quoting 438 U.S. at 750).

119. 478 U.S. 675 (1986).

120. 393 U.S. 503 (1969).

121. Fraser v. Bethel Sch. Dist. No. 403, 755 F.2d 1356, 1363 (9th Cir. 1985) (finding

“the First Amendment prohibited the District from punishing Fraser”), rev’d, 478 U.S.

675.

122. 478 U.S. at 685.

1016 COLUMBIA LAW REVIEW [Vol. 112:5

posite results, the Ninth Circuit and the Supreme Court majority at least

seemed to agree that the nature of the punishment inflicted on Fraser

was not relevant to his First Amendment claim.

In contrast, Justice Brennan suggested in an opinion concurring in

the judgment that a school’s chosen means of disciplinary action should

not lie outside the First Amendment inquiry. Like the majority, Brennan

believed that “school officials did not violate the First Amendment in de-

termining that [Fraser] should be disciplined.”

123

But unlike the major-

ity, he harbored doubts about the gravity of Fraser’s punishment:

Respondent served two days’ suspension and had his name re-

moved from the list of candidates for graduation speaker at the

school’s commencement exercises, although he was eventually

permitted to speak at the graduation. While I find this punish-

ment somewhat severe in light of the nature of respondent’s

transgression, I cannot conclude that school officials exceeded

the bounds of their disciplinary authority.

124

Thus, although he did not conclude that Fraser’s punishment was

unduly harsh in this case, Justice Brennan’s ruminations left open the

possibility that a harsher punishment might have met a different consti-

tutional fate.

125

7. Three Red Herrings. — This subsection turns briefly to three areas

of doctrine that, while showing some signs of the penalty-sensitive ap-

123. Id. at 690 (Brennan, J., concurring).

124. Id. at 690 n.3. For a survey of recent lower court cases exploring the possibility

of punishment-based review of school speech claims, see James F. Ianelli, Note, Punish-

ment and Student Speech: Straining the Reach of the First Amendment, 33 Harv. J.L. &

Pub. Pol’y 885, 890–92 (2010) (outlining and critiquing approach).

125. For some of the examples discussed in this Part, it is a fair question whether

penalty-sensitive considerations—though expressly entertained by the Court—actually af-

fected the outcome of the cases being decided. “Come on,” a skeptic might say, “the sever-

ity of the penalty didn’t really matter in cases like Pacifica, Finley, or Free Speech Coalition. The

Court just employed penalty-sensitive reasoning for rhetorical effect; the penalty-sensitive

aspects of these holdings amount to nothing more than window dressing applied to cases

that were penalty-neutral at their core.”

As a descriptive observation, this claim may have some merit. But even assuming its

rightness, the point does not much matter. Whatever the Court’s motivations for discuss-

ing the severity of a speaker’s penalty, two things are clear: First, the Court’s use of penalty-

sensitive reasoning for rhetorical effect would itself be significant; it would tell us that

penalty-sensitive reasoning carries rhetorical purchase, and hence some innate appeal, at

least to members of the Supreme Court. Second, and more important, if penalty-sensitive

arguments find support in the Court’s precedents, those arguments are fair game for liti-

gants and judges to cite in future cases, whatever the realist origins for those precedents

might be. Yesterday’s window dressing can inspire tomorrow’s redesign; throwaway lan-

guage in one decision can become the doctrinal linchpin of another. If penalty-sensitive

language appears within prior free speech precedents, then readers of such precedents

may take this language at face value. Penalty sensitivity, in other words, may not explain the

true basis for all of the decisions discussed above, but once invoked, it belongs to First

Amendment law. And that fact, in itself, must matter to anyone who desires to understand

this law in full.

2012] OF SPEECH AND SANCTIONS 1017

proach, do not quite embody the mode of free speech review that this

Article addresses. The primary aim here is to sharpen definitional con-

tours. Having identified illustrations of what falls within the boundaries

of the penalty-sensitive approach, the Article will briefly identify some

examples that fall outside these boundaries.

a. Prior Restraints. — Under the prior restraint rule, governments

enjoy greater constitutional leeway to regulate speech through

postpublication prosecutions than through prepublication restraint

orders.

126

Does this rule count as penalty-sensitive? Initial indicators are

positive: The rule draws distinctions based on the government’s means of

enforcing speech restrictions, and these means of enforcement employ

different types of punishments (e.g., contempt citations on the one hand

and traditional criminal/civil sanctions on the other). As a result, the

rule creates categories of speech that are punishable, but only in certain

ways.

Given the criteria identified in Part I, however, the prior restraint

rule counts as penalty-sensitive only if the in-kind distinction between

prior restraints and subsequent punishments tracks an in-degree distinc-

tion between “more exacting” and “less exacting” punishments. Here,

such a showing cannot be made, since, as Professor Marin Scordato has

noted, “there is no apparent logical or practically necessary correlation

between the severity of the sanction imposed in response to a violation of

the law and the law’s status as either a prior restraint or a subsequent

sanction.”

127

Indeed, postpublication punishments often are draconian;

just ask Jacob Abrams, who faced twenty years’ imprisonment based on a

postpublication prosecution. Prior restraints in the form of injunction

decrees, by contrast, are often quite limited.

128

Consequently, scholarly

126. See Neb. Press Ass’n v. Stuart, 427 U.S. 539, 559 (1976) (“The thread running

through [many First Amendment] cases is that prior restraints on speech and publication

are the most serious and the least tolerable infringement on First Amendment rights.”).

For examples of decisions disfavoring prior restraint, see generally N.Y. Times Co. v.

United States, 403 U.S. 713, 714 (1971) (per curiam) (finding government had not met

“heavy burden” necessary to justify prior restraint on publication of Pentagon Papers);

Org. for a Better Austin v. Keefe, 402 U.S. 415, 419 (1971) (holding government “carries a

heavy burden of showing justification for the imposition of such a restraint”); Bantam

Books, Inc. v. Sullivan, 372 U.S. 58, 70 (1963) (“Any system of prior restraints of expres-

sion comes to this Court bearing a heavy presumption against its constitutional validity.”);

Near v. Minnesota, 283 U.S. 697, 713 (1931) (noting “chief purpose [of First Amendment

is] to prevent previous restraints upon publication”).

127. Marin Scordato, Distinction Without a Difference: A Reappraisal of the Doctrine

of Prior Restraint, 68 N.C. L. Rev. 1, 12–13 (1989). But see John E. Nowak & Ronald D.

Rotunda, Constitutional Law § 16.16(e), at 1305–06 (8th ed. 2010) (speculating that

courts may tend to punish prior restraint violations more harshly because “[c]ourts are

more adamant in punishing contempt of their orders than . . . criminal statutes,” in which

they have no “personal stake”).

128. See Blasi, supra note 16, at 27 (“[S]anctions for the violation of injunctions and

permit ordinances tend to be light in comparison with those commonly administered un-

der the subsequent punishment regimes.”).

1018 COLUMBIA LAW REVIEW [Vol. 112:5

discussion of the prior restraint rule has centered on dangers unrelated

to the degree of punishment: for example, the relative lack of procedural

protections in the prior restraint context,

129

the increased probability of

speech suppression through prior restraints,

130

and the tendency of prior

restraint proceedings to present free speech claims in the abstract, rather

than in the light of a fully developed record.

131

These accounts may pro-

vide a satisfactory explanation of the elevated judicial scrutiny of prior

restraints, but they have nothing to do with differences in penalty sever-

ity.

b. Unconstitutional Conditions. — Running through First Amendment

law—and constitutional rights review more generally—is the problem of

unconstitutional conditions. This problem arises whenever the

government conditions receipt of a public benefit—be it a government

salary, welfare assistance, a procurement contract, and so on—on the

recipient’s waiver of certain constitutional protections. The law in this

area is complicated and multifaceted, but it is generally understood that,

within First Amendment doctrine, the government may more freely

regulate speech through the conditional allocation of benefits than

through the direct imposition of criminal and civil sanctions.

132

To the

extent that such a principle governs First Amendment law, it might be

tempting to regard it as embodying the penalty-sensitive approach.

133

Unconstitutional conditions doctrine sometimes permits the imposition

129. See Thomas I. Emerson, The Doctrine of Prior Restraint, 20 Law & Contemp.

Probs. 648, 657 (1955) [hereinafter Emerson, Prior Restraint] (noting “the procedural

protections built around criminal prosecution—many of which are constitutional guaran-

tees—are not applicable to a prior restraint”); Martin H. Redish, The Proper Role of the

Prior Restraint Doctrine in First Amendment Theory, 70 Va. L. Rev. 53, 55 (1984) (observ-

ing that prior restraint schemes present “the threat that expression will be abridged, if

only for a short time, prior to a full and fair hearing before an independent judicial fo-

rum”).

130. See Emerson, Prior Restraint, supra note 129, at 657 (noting “a decision to sup-

press in advance is usually more readily reached, on the same facts, than a decision to pun-

ish after the event”).

131. See Blasi, supra note 16, at 49 (noting prior restraints typically involve “adjudica-

tion in the abstract,” under which “the final authoritative judicial decision regarding the

legal status of a disputed communication takes place before the moment of initial dis-

semination”).

132. See Daniel A. Farber, Another View of the Quagmire: Unconstitutional Condi-

tions and Contract Theory, 33 Fla. St. U. L. Rev. 913, 920 (2006) (“Freedom of speech may

be the paradigmatic constitutional right. Despite their importance, however, speech rights

are alienable, at least in some contexts. The Supreme Court has staunchly defended First

Amendment rights in modern years, but it has allowed the government to restrict those

rights in exchange for providing benefits.”).

133. Indeed, this Article has already shown that within subsets of unconstitutional

conditions cases—those involving the speech of public employees and those involving

NEA grant awards—courts have unambiguously embraced penalty-sensitive logic. See su-

pra Part II.A.3, 4. The question here, however, is broader: It is whether we can character-

ize as penalty-sensitive the general rule that funding conditions are reviewed more leni-

ently than direct prohibitions.

2012] OF SPEECH AND SANCTIONS 1019

of one type of sanction (i.e., the withholding of benefits), but not

another (i.e., criminal and civil penalties), and given that the latter type

of sanction is often harsher than the former, all key penalty-sensitive

criteria seem to be met. When, for instance, the Court in Connick v. Myers

permitted the government to fire an employee for engaging in speech

that could not have given rise to a civil or criminal proceeding,

134

handed down what could plausibly be characterized as a penalty-sensitive

decision.

Conditional funding cases, however, present a special sort of diffi-

culty. Even if we stipulate that (a) governments enjoy greater leeway to

regulate speech through funding conditions than through direct prohi-

bitions and (b) the sanction of denying benefits is systematically less se-

vere than the sanction of imposing criminal/civil penalties, penalty-

sensitive analysis materializes only to the extent that (b) causes (a)—that

is, the relative leeway afforded to a funding condition is a consequence of

the relative leniency of its underlying sanctions. And this is not always the

case. It is true, for example, that the government is freer to restrict

speech as a condition of public employment than it is to restrict speech

by way of the criminal law.

135

But does this result obtain because the gov-

ernment as employer inflicts lighter penalties than the government as

law enforcer? Or because the government as employer cannot function

without enhanced abilities to monitor and restrict the speech of its em-

ployees? If the latter (as the Court has often indicated

136

), then the dis-

parity in constitutional treatment arises not from a disparity in penalty

severity, but rather from the idea that the special challenges of managing

a workforce necessitate an extra degree of constitutional deference to

the government’s managerial decisions. It is also true that the govern-

ment is freer to restrict libraries’ distribution of information through

funding conditions than it is to do the same through criminal prohibi-

tions.

137

But is that because the penalty of subsidies foregone is harsher

134. 461 U.S. 138, 147 (1983) (“[A]n employee’s false criticism of his employer on

grounds not of public concern may be cause for his discharge but would be entitled to the

same protection in a libel action accorded an identical statement made by a man on the

street.”).

135. See, e.g., Garcetti v. Ceballos, 547 U.S. 410, 418 (2006) (explaining that First

Amendment does not prohibit government employers from punishing employees’ speech

when speech is not “of public concern” or when employer has “adequate justification for

treating the employee differently from any other member of the general public”).

136. See Pickering v. Bd. of Educ., 391 U.S. 563, 568 (1968) (recognizing govern-

ment’s interest in “promoting the efficiency of the public services it performs through its

employees”); see also Borough of Duryea v. Guarnieri, 131 S. Ct. 2488, 2497 (2011) (“The

government’s interest in managing its internal affairs requires proper restraints on the in-

vocation of rights by employees when the workplace or the government employer’s re-

sponsibilities may be affected.”); Snepp v. United States, 444 U.S. 507, 510 n.3 (1980) (per

curiam) (upholding speech restriction on CIA employee).

137. See United States v. Am. Library Ass’n, 539 U.S. 194, 203–05 (2003) (plurality

opinion) (discussing government’s broad authority to further policy goals through fund-

1020 COLUMBIA LAW REVIEW [Vol. 112:5

than the penalty of fines incurred? Or is it because “‘when the

Government appropriates public funds to establish a program it is enti-

tled to define the limits of that program’”?

138

Decisions like these may

have the effect of permitting light penalties where heavy ones are forbid-

den, but the effect may be nothing more than a collateral consequence

of rules whose justifications lie elsewhere.

139

That said, it would be incorrect to characterize all of unconstitu-

tional conditions doctrine as unrelated to the penalty-sensitive approach.

Some conditional funding cases do incorporate penalty-sensitive consid-

erations into the constitutional analysis—drawing a clear causal link be-

tween distinctions in penalty severity and distinctions in constitutional

treatment. Indeed, the Court made just such a move in Finley, articulat-

ing an unambiguously penalty-sensitive justification for its decision to af-

ford the government increased flexibility in setting standards for its allo-

cation of NEA grants.

140

When courts employ reasoning of this sort, the

causation problem disappears; in these cases, differences in penalty se-

verity do generate differences in constitutional results.

The important point, then, is that the Court’s unconstitutional con-

dition cases are sui generis. Decisions in these cases sometimes will rest

on penalty-sensitive logic, but other times will not. This being so, one

cannot generally categorize this broad and complex area of constitu-

tional doctrine as consistently penalty-sensitive. Penalty sensitivity cer-

ing decisions).

138. Id. at 211 (quoting Rust v. Sullivan, 500 U.S. 173, 194 (1991)).

139. There are other non-severity-related reasons for giving the government greater

leeway to regulate through funding conditions than direct prohibitions. Courts and com-

mentators have argued, for example, (a) that funding cases are different because they im-

plicate contractual bargains between the government and its subsidy recipients, see

Richard Epstein, The Supreme Court, 1987 Term—Foreword: Unconstitutional Condi-

tions, State Power, and the Limits of Consent, 102 Harv. L. Rev. 4, 15, 40–41 (1988) (de-

veloping theory of “how the doctrine of unconstitutional conditions does, and should,

function in a variety of contexts as a check against the political perils of monopoly, collec-

tive action problems, and externalities”); Farber, supra note 132, at 915 (noting that “hav-

ing a right often means being free to decide on what terms to exercise it” and suggesting

that contract theory provides useful framework for analyzing unconstitutional conditions

issues), (b) that funding cases are different because the government’s freedom not to pro-

vide subsidies implies a further freedom to condition their receipt on the waiver of right,

cf. Cass R. Sunstein, Why the Unconstitutional Conditions Doctrine Is an Anachronism

(with Particular Reference to Religion, Speech, and Abortion), 70 B.U. L. Rev. 593, 597–

98 (1990) (discussing and criticizing argument), and (c) that funding cases are different

because the government acts as proprietor rather than sovereign, Bd. of Educ. v. Pico, 457

U.S. 853, 908–10, 920 (1982) (Rehnquist, J., dissenting) (“[T]he role of the government as

sovereign is subject to more stringent limitations than is the role of government as em-

ployer, property owner, or educator.”). These arguments do not seem to depend, at least

directly, on the relative leniency of the penalties embodied in funding restrictions.

140. See supra Part II.A.4 (discussing Finley); see also Harris v. McRae, 448 U.S. 297,

317 n.19 (1980) (noting in abortion context that “[a] refusal to fund protected activity,

without more, cannot be equated with the imposition of a ‘penalty’ on that activity”).

2012] OF SPEECH AND SANCTIONS 1021

tainly resides within this body of law, but it is identifiable only on a case-

by-case basis.

c. Sentencing Enhancements. — In Dawson v. Delaware, a sentencing

jury heard testimony about a convicted murderer’s membership in the

white-supremacist “Aryan Brotherhood.”

141

The testimony, in the Court’s

view, “proved nothing more than Dawson’s abstract beliefs,” which

“ha[d] no bearing on the issue being tried.”

142

Hence, the First

Amendment prohibited the trial court from admitting the evidence, and

the Court remanded for resentencing. But the Court did not call into

question the validity of Dawson’s original conviction; it only undid the

results of his sentencing proceeding. In this respect, the Dawson

holding—which passed judgment on a defendant’s sentence without

passing judgment on his conviction—may appear to be penalty-

sensitive.

143

On close inspection, however, Dawson’s penalty-neutral colors come

into plain view. Notice that the question asked and answered in Dawson

was whether, and not to what extent, the belief-based evidence could bear

on the severity of Dawson’s punishment. Even though the Court evalu-

ated a sentencing enhancement, it did so in a way that did not discrimi-

nate between large and small sentencing enhancements. The Court

would not have cared, that is, whether Dawson’s affiliation with the Aryan

Brotherhood had accounted for a slight uptick in a term of imprison-

ment, or had carried the more dramatic effect of converting a noncapital

sentence into a capital one. As long as the evidence created a risk of any

heightening of punishment, a new sentencing hearing had to be held.

Still, Dawson is worth keeping in mind, because it illustrates an im-

portant point about penalties and the First Amendment. Central to the

Court’s conclusion in Dawson was its concern that the sentencing jury,

once exposed to the belief-based evidence, may have punished Dawson

not just for his crime, but also for his beliefs. The Court thus feared that

some portion of Dawson’s sentence reflected purely belief-based con-

demnation, driven by nothing more than a disdain for the ideologies es-

poused by Dawson and his fellow Aryan Brothers. At its core, then,

Dawson was a case about improper motives. That Dawson received a

death sentence was not necessarily a problem; the problem was that this

death sentence may have stemmed from considerations that the First

Amendment disallows.

144

141. 503 U.S. 159 (1992).

142. Id. at 167–68.

143. In the end, the Court’s decision in Dawson had no effect on his ultimate fate.

Even with the belief-based evidence excluded, the resentencing proceeding assigned

Dawson the same sentence that he had originally received. See State v. Dawson, 681 A.2d

407, 411 (Del. Super. Ct. 1995) (noting Dawson was again sentenced to death following

Supreme Court’s reversal).

144. 503 U.S. at 167 (“Delaware might have avoided this problem if it had presented

1022 COLUMBIA LAW REVIEW [Vol. 112:5

In Dawson, the source of these considerations could be identified

and eliminated; the Court needed only to remand the case with instruc-

tions not to admit the Aryan Brotherhood evidence at resentencing. But

one can imagine situations in which proper and improper motives are

not so easily disaggregated. What if, for example, the jurors in Dawson

had discovered for themselves that Dawson was a racist and had escalated

his punishment accordingly? What if a judge sentences a breach-of-the-

peace defendant especially harshly for the covert reason that he loathes

the views espoused? To take a more familiar example, what if a legisla-

ture ratchets up penalties for draft-card burning, for the unspoken rea-

son that it detests the radicalism of draft-card burners? All of these cases

share with Dawson the problem of impermissible influences on the pun-

ishment imposed on a speaker. Unlike Dawson, however, they present the

problem in a way that reviewing courts cannot easily unravel. In these

cases, jury, judge, and legislature leave no paper trail of their improper

motives. Thus, a Dawson-like curative of a new and untainted government

decisionmaking process becomes difficult to achieve.

Perhaps, however, courts can perform the same “cleansing” act un-

dertaken in Dawson, by scrutinizing punishments rather than motives. On

this approach, rather than identifying improper motives and preventing

them from enhancing sentences, a penalty-sensitive court would identify

exceptionally onerous sentences and treat them as evidencing improper

motive. The finding of bad motives, in other words, would not lead the

court to the reduction of excessive punishments; the excessiveness of

punishments would lead the court to the detection of improper mo-

tives—which, in turn, would require judicial relief.

B. Penalty Sensitivity in Other Constitutional Contexts

The previous section identified some examples (and nonexamples)

of penalty-sensitive free speech review. The conclusion drawn there was

straightforward: Penalty sensitivity exists within First Amendment law,

and its presence there is more widespread than typically has been sug-

gested. But the descriptive claim can be broadened, as a look beyond

First Amendment borders confirms the solid doctrinal foundations of the

penalty-sensitive approach. Indeed, as this next section demonstrates, the

Court and its individual Justices have relied on penalty-sensitive review in

interpreting not just the free speech right, but many other constitutional

protections as well.

Take Justice Powell’s concurrence in Bowers v. Hardwick.

145

Bowers

presented a constitutional challenge to a Georgia antisodomy law,

evidence showing more than mere abstract beliefs on Dawson’s part, but on the present

record one is left with the feeling that the Aryan Brotherhood evidence was employed

simply because the jury would find these beliefs morally reprehensible.”).

145. 478 U.S. 186, 197 (1986) (Powell, J., concurring).

2012] OF SPEECH AND SANCTIONS 1023

founded on the claim that the law violated constitutional privacy protec-

tions. A five-Justice majority rejected this claim, concluding that criminal-

izing the acts in question did not infringe on the constitutional right to

privacy. Justice Powell joined the opinion, but he wrote separately to ex-

press serious hesitations about the penalties authorized by the law. In

particular, he worried that, under the law, a judge could impose prison

sentences of up to twenty years “for a single private, consensual act,” thus

creating “a serious Eighth Amendment issue.”

146

By fusing concerns

about the law’s privacy-infringing effects with concerns about its espe-

cially severe punishments, and by suggesting that the Eighth

Amendment—in joint operation with the privacy guarantee—might limit

the government’s power to punish private consensual behavior, Justice

Powell’s analysis of Bowers resonated with strong penalty-sensitive over-

tones.

147

146. Id.

147. It might be argued that, given Justice Powell’s express invocation of the Eighth

Amendment, his analysis of Bowers is better understood as an instance of straightforward

proportionality review, not as an instance of penalty-sensitive rights review. But that char-

acterization cannot be squared with the opinion’s specific allusion to the “private, consen-

sual” nature of the conduct being criminalized, id., an allusion that makes sense only in

reference to the Court’s longstanding recognition of a constitutional right to privacy.

What seems to have bothered Justice Powell was not merely the prospect of a harsh pen-

alty, nor was it merely the prospect of an intrusion on privacy. Rather, it was the possibility

of a harsh penalty coupled with an intrusion on privacy. Such an understanding necessarily in-

volves a penalty-sensitive view of the privacy right.

Justice Powell’s Eighth Amendment analysis is unusual in suggesting that especially

strict proportionality requirements might apply to punishments that target conduct lying

on the edge of other freestanding constitutional protections. In theory, however, courts

could frame penalty-sensitive free speech analysis in a manner reminiscent of this opinion,

reading the Eighth Amendment and Due Process Clause to impose especially strict limits

on the government’s power to punish expressive conduct. And such an approach would

hardly mark a significant departure from constitutional traditions. Equal protection analy-

sis, for instance, calls for heightened judicial skepticism of discriminatory action that im-

plicates fundamental rights. See, e.g., Kramer v. Union Free Sch. Dist. No. 15, 395 U.S.

621, 628–29 (1969) (noting need for “exacting judicial scrutiny” to review state law giving

voting right to some but not others); Shapiro v. Thompson, 394 U.S. 618, 641–42 (1969)

(holding statutory denial of social security benefits based on length of residency created

unconstitutionally discriminatory classification); Griffin v. Illinois, 351 U.S. 12, 17–20

(1956) (holding law limiting indigent defendants’ access to appellate review of criminal

convictions was discriminatory and unconstitutional). And free speech protections are said

to “take[] on an added dimension” when speech-infringing action also implicates right-to-

privacy interests. Stanley v. Georgia, 394 U.S. 557, 564 (1969). Along these lines, Eighth

Amendment/Due Process Clause proportionality protections might similarly strengthen

when the punishments at issue implicate conduct that falls within the First Amendment’s

domain. The goal of such analysis would be to locate punishments that, while insufficiently

excessive to give rise to a “pure form” Eighth Amendment or due process violation, and

also insufficiently speech-infringing to give rise to a “pure form” First Amendment viola-

tion, are sufficiently punitive and speech-infringing to give rise to a hybrid constitutional vio-

lation. Cf. Ariel Porat & Eric A. Posner, Aggregation and Law, 122 Yale L.J. (forthcoming

2012) (manuscript at 37), available at http://ssrn.com/abstract=1974565 (on file with the

1024 COLUMBIA LAW REVIEW [Vol. 112:5

Consider also the Court’s criminal procedure jurisprudence. Here,

the Court has held, for instance, that a criminal defendant acquires a

constitutional right to appointed counsel when he faces the possibility of

imprisonment but not when he faces the possibility of fines.

148

In like

fashion, the Court has refused to apply the constitutional jury trial right

in cases involving “petty offenses,” which, as a general matter, are crimes

“punishable by no more than six months in prison and a $500 fine.”

149

And with respect to the Fifth Amendment’s grand jury requirement, both

the Court and the Federal Rules Advisory Committee have interpreted

the provision’s reference to “infamous crimes” to encompass only crimes

subject to an “infamous punishment,” meaning a punishment greater

than a one-year prison sentence.

150

All of these rulings are penalty-

sensitive in the sense that they link the validity of a constitutional claim to

the severity of the punishment that a defendant confronts.

Penalty sensitivity is equally apparent on the civil side of the ledger,

where, in outlining the scope of the procedural due process guarantee,

the Court has asserted that “the degree of potential deprivation that may

be created by a particular decision is a factor to be considered in assess-

ing the validity of any administrative decisionmaking process.”

151

Goldberg v. Kelly, for example, the Court required an expansive pre-

deprivation hearing, emphasizing the “immediately desperate” nature of

the defendant’s circumstances, and the fact that deprivation of his wel-

fare benefits would leave him lacking “the very means by which to live.”

152

In Mathews v. Eldridge, by contrast, the Court viewed the “hardship im-

posed upon the erroneously terminated disability recipient” as less severe

Columbia Law Review) (showing how hybrid rights analysis can permit “cross-claim norma-

tive aggregation,” under which a statute that does not independently violate either of two

constitutional provisions might nonetheless “violate them jointly”). Regardless of whether

such analysis is properly described as “penalty-sensitive” First Amendment review or

“speech-sensitive” proportionality review, the end result is the same: penalty-based limita-

tions on the government’s power to restrict speech.

148. Scott v. Illinois, 440 U.S. 367, 373–74 (1979); Argersinger v. Hamlin, 407 U.S.

25, 37 (1972) (holding “absent a knowing and intelligent waiver, no person may be im-

prisoned for any offense, whether classified as petty, misdemeanor, or felony, unless he

was represented by counsel at his trial”); see also Alabama v. Shelton, 535 U.S. 654, 658

(2002) (holding state must appoint counsel where conviction results in suspended sen-

tence that “may ‘end up in the actual deprivation of a person’s liberty’” if probation is re-

voked (quoting Argersinger, 407 U.S. at 40)).

149. Baldwin v. New York, 399 U.S. 66, 71 (1968). For cases fleshing out the meaning

of this rule, see, for example, Codispoti v. Pennsylvania, 418 U.S. 506, 512 (1974) (holding

jury trial right applies when consecutive sentences exceed six months in aggregate), and

Muniz v. Hoffman, 422 U.S. 454, 477–78 (1975) (casting doubt on “the proposition that a

contempt [violation] must be considered a serious crime under all circumstances where

the punishment is a fine of more than $500, unaccompanied by imprisonment”).

150. See Fed. R. Crim. P. 7(a)(2) advisory committee’s note (interpreting Supreme

Court’s Fifth Amendment decision in Duke v. United States, 301 U.S. 492 (1937)).

151. Mathews v. Eldridge, 424 U.S. 319, 341 (1976).

152. 397 U.S. 254, 264–65 (1970).

2012] OF SPEECH AND SANCTIONS 1025

than “that of a welfare recipient” on the ground that entitlement to dis-

ability payments was not need-based.

153

Consequently, there was “less rea-

son [in Mathews] than in Goldberg to depart from the ordinary principle

that something less than an evidentiary hearing is sufficient prior to ad-

verse administrative action.”

154

The Court’s equal protection jurisprudence also reveals traces of the

penalty-sensitive approach. In McLaughlin v. Florida, for example, the

Court struck down a Florida cohabitation law as unconstitutionally dis-

criminatory.

155

To support this holding, Justice White’s majority opinion

emphasized the severity of the law’s penalties: “We deal here with a racial

classification embodied in a criminal statute. In this context, where the

power of the State weighs most heavily upon the individual or the group,

we must be especially sensitive to the policies of the Equal Protection

Clause.”

156

Likewise, Justice Powell’s dispositive opinion in Wygant v.

Jackson Board of Education drew a constitutional distinction between race-

based layoffs and race-based hiring decisions.

157

The latter, he argued,

were less problematic than the former because, “[w]hile hiring goals im-

pose a diffuse burden, often foreclosing only one of several opportuni-

ties, layoffs impose the entire burden of achieving racial equality on par-

ticular individuals, often resulting in serious disruption of their lives.”

158

Both McLaughlin and Wygant illustrate penalty-sensitive conceptions of

the equal protection guarantee. Finding a constitutional violation de-

pends not only on the classification employed, but also on the signifi-

cance of the harm the government imposes.

Further examples come from Takings Clause and Contract Clause

doctrine. In Lucas v. South Carolina Coastal Council, the Court held that

153. 424 U.S. at 342.

154. Id. at 343; see also Goss v. Lopez, 419 U.S. 565, 581, 584 (1975). Goss holds that

[s]tudents facing temporary suspension have interests qualifying for protection

of the Due Process Clause, and due process requires, in connection with a sus-

pension of 10 days or less, that the student be given oral or written notice of the

charges against him and, if he denies them, an explanation of the evidence the

authorities have and an opportunity to present his side of the story.

Id. It does note, however, that “[l]onger suspensions or expulsions for the remainder of

the school term, or permanently, may require more formal procedures.” Id.

155. 379 U.S. 184 (1964).

156. Id. at 192. See also Justice Stewart’s concurrence, in which he said,

There might be limited room under the Equal Protection Clause for a civil law

requiring the keeping of racially segregated public records for statistical or other

valid public purposes. But we deal here with a criminal law which imposes crimi-

nal punishment. And I think it is simply not possible for a state law to be valid

under our Constitution which makes the criminality of an act depend upon the

race of the actor. Discrimination of that kind is invidious per se.

Id. at 198 (Stewart, J., concurring) (citation omitted).

157. 476 U.S. 267, 282–83 (1986) (Powell, J.) (plurality opinion) (contrasting race-

based hiring goals with more burdensome race-based layoffs).

158. Id. at 283.

1026 COLUMBIA LAW REVIEW [Vol. 112:5

“total” regulatory takings operate in a per se fashion to trigger just-

compensation obligations, whereas nontotal takings do not.

159

Likewise,

in Allied Structural Steel Co. v. Spannaus, the Court held that the govern-

ment violates the Contract Clause only when it causes “substantial im-

pairment of a contractual relationship.”

160

“Minimal alteration of contrac-

tual obligations,” the Court explained, “may end the inquiry at its first

stage.”

161

But “[s]evere impairment, on the other hand, will push the in-

quiry to a careful examination of the nature and purpose of the state leg-

islation.”

162

Notice that, for both of these restrictions, the Court has as-

sessed the severity of an individual’s harm not for the purpose of deter-

mining how much compensation the government owes that individual

but rather for the purpose of determining whether the government owes

compensation in the first place. Thus, the weightiness of the punitive

burden placed by the government on the individual is closely linked to

the question of whether the individual’s constitutional rights have or

have not been violated.

None of this is to say that cases of this kind litter the constitutional

landscape, or that these “non-First Amendment” examples correspond in

perfect fashion to their counterparts within the First Amendment do-

main. The existence of this case law does suggest, however, that the

Court’s forays into penalty-sensitive free speech review are by no means

radical or isolated events. Indeed, the examples illustrate that penalty-

sensitive analysis has surfaced within many areas of constitutional rights

review, thus raising the question whether expanding its presence within

free speech doctrine is a worthwhile goal to pursue.

III.

JUSTIFYING THE PENALTY-SENSITIVE APPROACH

Part II of this Article catalogued several existing examples of penalty-

sensitive free speech doctrine. The intent was not to suggest that modern

free speech doctrine is predominantly or even substantially penalty-

sensitive. To the contrary, the previously discussed examples may be

viewed as departures from the penalty-neutral norm—areas of doctrine

that are noteworthy precisely because they deviate from standard First

Amendment practice. Even so, as Part II makes clear, penalty sensitivity is

certainly present within First Amendment doctrine, and its presence

raises important questions: Does penalty sensitivity make sense as a

method of free speech adjudication? Should courts increase their use of

the penalty-sensitive approach in free speech cases? And, if so, how

should courts put penalty-sensitive First Amendment analysis to further

use?

159. 505 U.S. 1003, 1026–31 (1992).

160. 438 U.S. 234, 244 (1978) (emphasis added).

161. Id. at 245.

162. Id.

2012] OF SPEECH AND SANCTIONS 1027

This next Part addresses these questions. It argues that penalty sensi-

tivity does make sense, and that courts should expand its existing presence

within free speech law. To support these claims, this Part identifies five

ways in which penalty sensitivity can improve courts’ implementation of

the free speech right. Specifically, it argues that penalty sensitivity can (a)

increase the fairness of doctrinal results, (b) mitigate chilling effects on

protected speech, (c) facilitate the “efficient breach” of constitutionally

borderline speech restrictions, (d) improve courts’ understanding of leg-

islative motives, and (e) promote the transparency of judicial decision-

making. The lesson to draw from this discussion is not that all First

Amendment cases should be resolved in a penalty-sensitive fashion; it is

rather that courts have thus far underutilized penalty-sensitive analysis,

which, if properly undertaken, can assist in the resolution of difficult

First Amendment cases.

A. Fairness

Penalty-sensitive adjudication promotes fairness. I have in mind here

the principle that like cases should receive like treatment,

163

or, more ac-

curately, its corollary that “almost like” cases should receive “almost like”

treatment.

164

Governments offend these principles when they mete out

substantially different punishments for substantially similar forms of be-

havior.

165

And courts enable such behavior when they adjudicate free

speech claims in a penalty-neutral manner.

To see the point, imagine a one-dimensional axis measuring the

constitutional value of various expressive activities. Toward the left end of

the axis are activities of low constitutional value, and toward the right are

activities of high and recognized constitutional importance. For example,

a comparison of different forms of sexually explicit speech might result

in a diagram like this

166

163. This principle has long been recognized as fundamental to the rule of law. See,

e.g., Lon Fuller, The Morality of Law 211 (1969) (associating principle with requirement

that laws be general in nature); H.L.A. Hart, The Concept of Law 156 (1961) (“[T]he idea

of justice . . . consists of two parts: a uniform or constant feature, summarized in the pre-

cept ‘Treat like cases alike’ and a shifting or varying criterion used in determining when,

for any given purpose, cases are alike or different.”); John Rawls, A Theory of Justice 237

(1971) (“The rule of law also implies the precept that similar cases be treated similarly.”).

164. Congress has given this principle statutory recognition in 18 U.S.C. § 3553(a)(6)

(2006), which instructs sentencing courts to consider “the need to avoid unwarranted sen-

tence disparities among defendants with similar records who have been found guilty of

similar conduct.”

165. See, e.g., Richard G. Singer, Just Deserts: Sentencing Based on Equality and

Desert 11 (1979) (“Every just scheme of punishment must be fair. And part of fairness is

equality. Prima facie, equality would demand that two offenders who have committed the

same offense receive the same sentence.”).

166. In case it does not go without saying, the chart’s estimation of “constitutional

value” reflects only an impressionistic and inexact reading of First Amendment law. It

does, however, accord with the Court’s suggestions that activities toward the left end of the

1028 COLUMBIA LAW REVIEW [Vol. 112:5

PECTRUM OF SEXUALLY EXPLICIT SPEECH

Let us stipulate that speech falling far to the low end of the spec-

trum merits no constitutional concern, while speech falling far to the

high end of the spectrum merits absolute constitutional protection. With

these stipulations in place, the doctrinal challenge becomes apparent:

We must allocate constitutional protections between these two extremes.

Applying the penalty-neutral approach to First Amendment adjudi-

cation, we simply draw a constitutional borderline at the point on the spec-

trum where constitutional value becomes high enough to mandate con-

stitutional intervention. All activities falling to the left of the borderline

are fully punishable,

167

while all activities falling to the right of the bor-

derline are not punishable at all. (On a simplified model of existing First

Amendment doctrine,

168

this borderline would rest somewhere near the

obscenity standard of Miller v. California.

169

) Thus, if we were to add to

our original diagram a y-axis measuring maximum allowable penalty se-

verity, the penalty-neutral approach would yield something like the fol-

lowing:

spectrum are of lower constitutional value than the activities toward the right end of the

spectrum. See, e.g., New York v. Ferber, 458 U.S. 747, 762 (1982) (“The value of permit-

ting live performances and photographic reproductions of children engaged in lewd sex-

ual conduct is exceedingly modest, if not de minimis.”); FCC v. Pacifica Found., 438 U.S.

726, 745–46 (1978) (plurality opinion) (characterizing “patently offensive words dealing

with sex and excretion” as “ordinarily lack[ing] literary, political, or scientific value,” but

“not entirely outside the protection of the First Amendment”); Miller v. California, 413

U.S. 15, 34 (1973) (“The First Amendment protects works which, taken as a whole, have

serious literary, artistic, political, or scientific value, regardless of whether the government

or a majority of the people approve of the ideas these works represent.”); Chaplinsky v.

New Hampshire, 315 U.S. 568, 571–72 (1942) (“There are certain well-defined and nar-

rowly limited classes of speech, the prevention and punishment of which have never been

thought to raise any Constitutional problem. These include the lewd and obscene . . . . It

has been well observed that such utterances are no essential part of any exposition of

ideas . . . .”).

167. Note that “fully punishable” here means something like “punishable to the full

extent permitted by independent constitutional limitations, such as the Eighth

Amendment or the Due Process Clause.”

168. This model omits, for example, the penalty-sensitive distinction introduced in

Pacifica and Reno. See supra Part II.A.5 (discussing evolution of penalty-sensitive doctrine

governing “indecent” expression and communications media).

169. 413 U.S. 15 (1973).

Child

pornography

Miller

obscenity

Indecen

erotica

Rac

B-movie

rts

A-movie

Scientific

documentar

2012] OF SPEECH AND SANCTIONS 1029

ENALTY-NEUTRAL APPROACH

So why does penalty neutrality fail to treat like alike? The problem is

that, as this diagram illustrates, when conduct approaches the Miller

standard, small changes in constitutional value translate into huge differ-

ences in constitutional protection.

170

This precipitous drop means that

the constitutional regime must assign very different constitutional fates

to very similar activities. Suppose, for instance, that Hank and Harry have

posted lewd videos on their respective websites, and suppose further that

Hank’s video qualifies as “barely patently-offensive” while Harry’s video

qualifies as “almost patently-offensive.” On a penalty-neutral application

of the Miller rule, Hank is subject to grave punishment—potentially many

years in prison—while Harry is immune from sanction of any kind. This

outcome is not good.

171

A modified diagram reveals how penalty sensitivity ameliorates the

fairness problem. In most areas, the penalty sensitive approach parallels

the penalty-neutral approach, with activities at the left end entitled to no

170. This phenomenon is by no means isolated to penalty-neutral modes of constitu-

tional adjudication. For a general analysis of the problem—and a general recommenda-

tion in favor of “smoother” laws—see Adam Kolber, Smooth and Bumpy Laws 3–4 (Feb.

14, 2012) (unpublished manuscript), available at http://ssrn.com/ abstract=1992034 (on

file with the Columbia Law Review) (discussing benefits of “smooth” doctrine in various le-

gal contexts).

171. Notice that this problem has nothing to do with the standard’s lack of clarity.

The problem, in other words, does not arise from the difficulty of distinguishing between

“patently offensive” and “not patently offensive” speech; rather, it arises from the conse-

quences of making such a distinction in the first place—consequences that in no way de-

pend on the coherence of the distinction itself. Hence, courts cannot solve the fairness

problem merely by increasing doctrinal determinacy.

Eighth Amendment limit

Child

pornography

Miller

obscenity

Indecen

erotica

Rac

B-movie

rts

A-movie

Scientific

documentar

1030 COLUMBIA LAW REVIEW [Vol. 112:5

protection, and activities at the right end entitled to protection that is ab-

solute. But in the area surrounding Miller obscenity, the penalty-sensitive

approach employs gradated limits—rather than a single, stark border-

line—to achieve a rough correspondence between expressive value and

maximum punishability. Such an approach might, for example, permit

states to punish publishers of child pornography to the full extent per-

mitted by the Eighth Amendment, publishers of Miller obscenity with a

prison sentence not to exceed one year, publishers of pornography fal-

ling just short of the Miller standard with monetary fines, and publishers

of indecent erotica to no extent at all.

172

ENALTY-SENSITIVE APPROACH

This approach does not eliminate fairness problems, but it has a

mitigating effect. Under this new scheme, Hank and Harry would still re-

ceive substantially different constitutional treatments (with Hank ex-

posed to the possibility of a one-year prison term and Harry exposed to

the possibility of a fine), but the difference is less dramatic than the

penalty-neutral alternative would dictate. Courts might also improve the

picture by introducing new tiers, further lessening the disparities in the

treatment of constitutionally similar acts. Eventually, the benefits of fur-

ther gradations will give way to losses in administrability, but these losses

are not likely to be substantial when the tiers are few in number.

173

172. It bears emphasizing that the diagram reflects maximum punishments—not

mandatory punishments. That is, the Court would not be telling legislatures that they must

punish Miller obscenity with one-year prison terms. Rather, the Court would be saying that

the First Amendment prohibits any such punishment exceeding one year.

173. See infra Part IV.B (discussing and responding to implementation-related objec-

tions to penalty-sensitive analysis).

Child

pornography

Miller

obscenity

Indecen

erotica

Rac

B-movie

rts

A-movie

Scientific

documentar

Eighth Amendment limit

One-year prison term

$1,000 fine

2012] OF SPEECH AND SANCTIONS 1031

The foregoing discussion has focused on equality, but it also helps to

illustrate a related point, grounded in the concept of proportionality.

While penalty-sensitive review helps to ensure similar treatment for simi-

lar speakers, it also fosters adherence to the time-honored maxim that

“the punishment should fit the crime.”

174

Penalty neutrality, as the charts

above indicate, fails to distinguish between two different inquiries: (a)

whether speech has insufficient value to merit total constitutional protec-

tion; and (b) whether speech has insufficient value to merit any constitu-

tional protection at all. If a penalty-neutral analyst answers “yes” to the

first question, then he must also answer yes to the second; from the con-

clusion that the government may punish a speech act lightly, it automati-

cally follows that the government may do so harshly as well. Strict adher-

ence to such a scheme raises the risk of disproportionate punishments,

and it permits governments to punish speech carrying some First

Amendment value as if it carried no First Amendment value at all. Pen-

alty sensitivity avoids this problem by allowing courts to take note of a

speech’s constitutional value even after concluding that such speech does

not warrant all-out constitutional protection. And in so doing, penalty

sensitivity helps to prevent the imposition of punishments that are

marked by undue harshness when viewed in relation to the expressive

value of the act in question.

175

B. Chilling Effects

Modern First Amendment doctrine manifests a keen awareness of

chilling effects. Chilling effects, in the words of Professor Frederick

Schauer, occur when “individuals seeking to engage in activity protected

by the first amendment are deterred from so doing by governmental

174. BMW of N. Am., Inc. v. Gore, 517 U.S. 559, 575 n.24 (1996) (“The principle that

punishment should fit the crime ‘is deeply rooted and frequently repeated in common-law

jurisprudence.’” (quoting Solem v. Helm, 463 U.S. 277, 284 (1983))). For philosophical

development of this idea, see, for example, H.L.A. Hart, Prolegomenon to the Principles

of Punishment, in Punishment and Responsibility 1, 25 (2d ed. 2008) (discussing “the

principle that different kinds of offence of different gravity (however that is assessed)

should not be punished with equal severity”).

175. One could achieve similar results by modifying Eighth Amendment doctrine so

as to permit a “hybridized” proportionality analysis of speech-restricting punishments, tar-

geting government actions that, though not individually violative of Eighth Amendment

proportionality requirements or First Amendment prohibitions, would still violate a pro-

portionality rule informed by First Amendment considerations. See supra note 147 (out-

lining potential hybrid constitutional analysis arising from joint application of First and

Eighth Amendments). Federal sentencing courts, moreover, could engage in a similar

analysis in connection with their application of 18 U.S.C. § 3553(a), concluding that the

expressive value of an act, while insufficient to bar its punishment under the First

Amendment (or Eighth Amendment), at least necessitates a downward variance from the

Guidelines range. See 18 U.S.C. § 3553(a)(2)(A) (2006) (providing “[t]he court, in de-

termining the particular sentence to be imposed, shall consider . . . the need for the sen-

tence imposed . . . to reflect the seriousness of the offense, to promote respect for the law,

and to provide just punishment for the offense”).

1032 COLUMBIA LAW REVIEW [Vol. 112:5

regulation not specifically directed at that protected activity.”

176

That is,

governments “chill” protected speech by restricting some other form of

speech that, while unprotected, is similar to the speech getting chilled.

As Schauer explains, “[I]f a statute which is directed at hard core por-

nography has the actual effect of deterring an individual from publishing

the Decameron or Lady Chatterley’s Lover, that effect is properly deemed a

chilling effect.”

177

But the same statute does not have a chilling effect if it

merely deters the publication of unprotected hard core pornography.

At first glance, the notion of a chilling effect may seem puzzling. If a

statute targets only behavior A1, why would it also deter behavior A2?

Schauer answers this question by disaggregating two distinct states of

mind: fear and uncertainty.

178

Chilling effects can arise, that is, when

individuals fear that their participation in behavior A2 will result in an

erroneous conviction under the ban on behavior A1, or when individuals

do not know whether the ban on behavior A1 also encompasses behavior

A2. A bookseller, for instance, might decline to sell nonobscene books

for fear that the state’s legal apparatus will misclassify them as obscene (a

fear-based chilling effect). Or he might not sell the books because he

does not know whether they are in fact obscene (an uncertainty-based

chilling effect).

179

The upshot of these observations is that courts cannot safeguard the

exercise of free speech rights by protecting only speech acts that intrinsi-

cally deserve free speech protection. Rather, as Schauer explains, courts

must take additional steps to counteract the dangers of both fear-based

and uncertainty-based chilling effects. One strategy is to extend the

range of constitutional protection beyond the range of constitutional

value, thereby creating a “buffer zone” between conduct that deserves

protection and conduct that receives it.

180

The Court employed this strat-

egy in New York Times: What compelled its adoption of the “actual malice”

standard was not the intrinsic value of defamatory falsehoods, but rather

the need for “a measure of strategic protection” to promote vigorous and

uninhibited public debate.

181

Another strategy is to insist on precise legal

language, thereby reducing uncertainty regarding the scope of a speech

prohibition.

182

The vagueness and overbreadth rules serve this anti-

chilling purpose by reducing the risk that individuals will mistakenly de-

cline to engage in protected expression as a result of miscomprehending

the ambit of a speech-restricting rule.

176. Schauer, supra note 90, at 693.

177. Id.

178. Id. at 694.

179. Id. at 699.

180. Id. at 707.

181. See Gertz v. Robert Welch, Inc., 418 U.S. 323, 342 (1974) (explaining result in

New York Times).

182. Schauer, supra note 90, at 698–700.

2012] OF SPEECH AND SANCTIONS 1033

Schauer analyzes these strategies in depth, but he does not discuss a

third way in which courts can reduce both fear-based and uncertainty-

based chilling effects. This strategy is rooted in Schauer’s own suggestion

that the intensity of a law’s chill “may be likened to the product of the

probability of an erroneous verdict times the harm produced by such a

verdict,” the “most obvious measure” of which “is the harshness of the

penalty.”

183

Schauer’s strategies target the first component of this equa-

tion: Strategic buffer zones and heightened clarity requirements reduce

the probability of erroneous verdicts, thereby reducing the risk of silenc-

ing protected speech. But given that punishment severity also contributes

significantly to a law’s deterrent effect, penalty-sensitive adjudication also

can operate to combat chilling effects.

Consider again the seller of erotic literature, who works in a jurisdic-

tion where purveying obscenity is punishable by twenty years’ imprison-

ment. Even with a pretty clear definition of obscenity, and even with a

low likelihood of an erroneous verdict, the prospect of spending two

decades in prison may well deter the bookseller from carrying non-

obscene books. Here, the best way to reduce chilling effects is to lessen

the severity of the punishment—holding, for instance, that the state can-

not punish obscenity with anything more severe than a monetary fine or

the forfeiture of law-offending publications. While leaving the probability

of an erroneous verdict unchanged, this alteration would still reduce the

severity of the law’s chilling effect, giving the bookseller a worst case sce-

nario he might well be willing to live with.

This example is not fanciful. Indeed, some of the cases discussed in

Part II illustrate just this sort of approach to the problem of deterring

protected speech. What necessitated Gertz’s limit on presumed and puni-

tive damages was not the inherent constitutional value of negligently ut-

tered falsehoods; this much was clear from the Court’s assertion that

“there is no constitutional value in false statements of fact,” whether

caused by the “the intentional lie” or the “careless error.”

184

Rather, the

Court’s holding was designed to limit chilling effects on constitutionally

valuable speech. But unlike New York Times, which mitigated chilling ef-

fects by reducing the likelihood of erroneous verdicts, Gertz mitigated

chilling effects by reducing the harms that erroneous verdicts could gen-

erate. As Justice Powell explained, “[t]he largely uncontrolled discretion

of juries to award damages where there is no loss unnecessarily com-

pounds the potential of any system of liability for defamatory falsehood

to inhibit the vigorous exercise of First Amendment freedoms.”

185

Put

183. Id. at 695–96.

184. Gertz, 418 U.S. at 340; see also N.Y. Times Co. v. Sullivan, 376 U.S. 254, 300

(1964) (Goldberg, J., concurring) (“It may be urged that deliberately and maliciously false

statements have no conceivable value . . . .”).

185. Gertz, 418 U.S. at 349.

1034 COLUMBIA LAW REVIEW [Vol. 112:5

simply, the Court fashioned restrictions on punitive and presumed dam-

ages to guard against the “danger of media self-censorship.”

186

Concerns about chilling effects also account for the penalty-sensitive

sides of the vagueness and overbreadth doctrines. Recall Justice

O’Connor’s assurances in Finley that “[i]t is unlikely . . . that speakers will

be compelled to steer too far clear of any ‘forbidden area’ in the context

of grants of this nature.”

187

Translated into Schauer’s terms, the point is

that low penalty severity in this context operated to minimize

uncertainty-based chilling effects; with relatively little on the line, grant

applicants were not likely to temper their artwork to avoid offending

even the murky guidelines that the NEA applied.

188

Consider also Justice

Kennedy’s observation in Free Speech Coalition that “[w]ith [the CPPA’s]

severe penalties in force, few legitimate movie producers or book pub-

lishers, or few other speakers in any capacity, would risk distributing im-

ages in or near the uncertain reach of this law.”

189

Here, high penalty se-

verity threatened to exacerbate the chilling effects of the CPPA: Given

the terrible consequences of a CPPA violation, individuals would not go

anywhere near the law’s boundaries.

Gertz, Finley, and Free Speech Coalition all took account of the close

connection between the severity of a law’s penalty and the intensity of its

deterrent effect on valuable speech. But courts cannot attend to this

connection when stuck in penalty-neutral mode. Take the majority opin-

ion in Bethel: Having evaluated the content of Matthew Fraser’s speech,

and having then concluded that such speech should not receive protec-

tion, the majority moved on to other issues.

190

Had it looked at Fraser’s

punishment—a multiday suspension—the Court might well have noticed

a chilling effects problem. School officials had imposed, with hardly any

warning at all,

191

a serious sanction that most students would take pains

to avoid. What is more, Fraser’s speech achieved its effect through allu-

sion, relying entirely on innuendo that was inoffensive on its face. These

186. Id. at 350; see also Carol Rice Andrews, The First Amendment Problem with the

Motive Restrictions in the Rules of Professional Conduct, 24 J. Legal Prof. 13, 57–58

(2000) (explaining Gertz rule in similar terms).

187. NEA v. Finley, 524 U.S. 569, 588 (1998).

188. The Court acknowledged the connection between chilling effects and uncertain

legal language in Spieser v. Randall, where it noted,

The man who knows that he must bring forth proof and persuade another of the

lawfulness of his conduct necessarily must steer far wider of the unlawful zone

than if the State must bear these burdens. This is especially to be feared when

the complexity of the proofs and the generality of the standards applied provide

but shifting sands on which the litigant must maintain his position.

357 U.S. 513, 526 (1958) (citations omitted).

189. Ashcroft v. Free Speech Coal., 535 U.S. 234, 244 (2002).

190. Bethel Sch. Dist. No. 403 v. Fraser, 478 U.S. 675, 686 (1986).

191. See id. at 691 (Stevens, J., dissenting) (arguing Fraser was not given fair notice

of school’s prohibitions).

2012] OF SPEECH AND SANCTIONS 1035

facts in tandem raised a chilling effects problem. Future student speak-

ers, aware of Fraser’s punishment, might well hesitate to give public ad-

dresses, or they might at least go out of their way to excise potentially

controversial material from their prepared remarks. Had Fraser’s sanc-

tion not been so severe, the increase in self-censorship would not have

been as great. By raising the disciplinary stakes to a significant level, the

school board’s actions threatened to suppress much more than just the

“patently offensive” speech that Bethel purported to deem out of bounds.

C. “Preserving” Unprotected Speech

A related, though distinct, role of penalty sensitivity involves avoid-

ing the overdeterrence of unprotected speech. This strategy begins with

the thought that, under certain conditions, courts may wish to deny abso-

lute constitutional protection to an expressive activity, while simultane-

ously taking measures to protect against that activity’s total extinction.

Whether prompted by uncertainty about the correctness of the constitu-

tional decision, or by the desire to fashion a compromise solution to a

“close” constitutional case, courts may employ penalty-sensitive rules to

reduce the deterrent impact of a refusal to protect speech in an all-out

way. The predictable effect of such rules would be to suppress some, but

not all, instances of the conduct denied protection—permitting the gov-

ernment to soften, but not to silence, speech lying just beyond the zone

of absolute constitutional protection.

The benefits of this “speech-preserving” strategy go beyond its capac-

ity to reduce the sum quantity of speech that is deterred: There is a qual-

ity dimension, as well. The key idea is this: So long as the Court maintains

the “punitive price” of speaking at a nonprohibitive level, individuals can

reasonably choose to flout a government’s speech restriction, calculating

that the benefits of speaking exceed the constitutionally capped costs of

punishment. When this happens, we can further surmise that the speech

left undeterred is the speech that matters most to would-be speakers. If

an individual only marginally desires to engage in the unprotected con-

duct, then the presence of even limited punishments will deter her from

running afoul of the government’s restriction; if she yearns to speak,

then these same punishments will not deter her. The penalty-sensitive

rule thus functions as a screening device, filtering out instances of a

speech act of little concern to the speaker, while preserving instances of

the act that she values the most.

This point is constitutionally significant for two reasons. First, insofar

as the objective constitutional value of a speech act correlates with the

subjective value that an individual attaches to it, the penalty-sensitive rule

will have the salutary effect of effectively facilitating valuable forms of

unprotected speech. Suppose, for example, that the Court were to adopt

hard-and-fast limits on punishments accompanying contempt citations—

i.e., punishments that judges impose on individuals engaging in aggres-

1036 COLUMBIA LAW REVIEW [Vol. 112:5

sive, profane, or otherwise disruptive forms of in-court expression.

192

Even if there were no substantive free speech restrictions on the con-

tempt power (i.e., rules identifying forms of contempt of court that were

absolutely unpunishable),

193

placing limits on the degree of punishments

for contempt may go a long way toward preserving those forms of con-

temptuous conduct that are most worth preserving. That is because, in

this context, an individual’s subjective desire to breach courtroom deco-

rum should at least roughly correlate with the objective constitutional

value of his conduct. Compare, for example, two speakers: the witness

who peppers his testimony with gratuitously profane language, and the

defendant who vigorously protests treatment he believes to be unfair.

Generally speaking, the penalty-sensitive rule should end up deterring

the former sort of conduct more intensely than the latter, as individuals

will tend to value more greatly the opportunity to protest treatment they

perceive to be unfair than the opportunity to toss about profanities.

From a constitutional perspective, this may be as it should be: We might

well prefer to preserve vigorous protestations of perceived government

oppression over needless embellishments of in-court speech, on the

ground that the former type of communication is more likely than the

latter to further First Amendment values.

194

In this respect, a penalty-

192. For general discussions of First Amendment limitations on the contempt power,

see Louis S. Raveson, Advocacy and Contempt: Constitutional Limitations on the Judicial

Contempt Power, 65 Wash. L. Rev. 477, 497–506 (1990); Ronald J. Rychlak, Direct Crimi-

nal Contempt and the Trial Attorney: Constitutional Limitations on the Contempt Power,

14 Am. J. Trial Advoc. 243, 272–86 (1990); see also Ira P. Robbins, Digitus Impudicus: The

Middle Finger and the Law, 41 U.C. Davis L. Rev. 1403, 1476–84 (2008) (examining con-

tempt power as it applies to use of middle finger gesture in court).

193. I assume here that a desirable constitutional regime permits the suppression of

some, but not all, disruptive or disrespectful speech acts in the courtroom. Cf. United

States ex rel. Lynch v. Werksman, 319 F. Supp. 353, 354 (N.D. Ill. 1970) (“If every dis-

respectful comment of losing counsel and litigants were to constitute criminal contempt,

the prison population in the United States would be substantially increased and the First

Amendment would have a substantial new exception to its protection.”).

194. The defendant’s “vigorous protestations” seem to me more valuable than the

witness’s “gratuitous slurs” under any number of free speech theories. One might defend

the distinction, for instance, by reference to Professor Meiklejohn’s “self-government”

framework, arguing that the defendant’s conduct represents a vital act of public dissent,

whereas the witness’s conduct does not in any meaningful way promote democratic self-

government. See generally Alexander Meiklejohn, Free Speech and Its Relation to Self-

Government (1948) (outlining First Amendment theory based on role of free speech in

ensuring proper functioning of democratic system). The distinction also makes sense

within a Millian “marketplace-of-ideas” framework: The defendant’s vigorous protestations

draw attention to a competing assessment of a judge’s decision, while the witness’s gratui-

tous embellishments add little of substance to a view already being espoused. Cf. John

Stuart Mill, On Liberty 86–120 (David Bromwich & George Kateb eds., Yale Univ. Press

2003) (1859) (defending freedom of speech as allowing for airing of competing opinions

and, through this process, facilitating eventual emergence of truth).

2012] OF SPEECH AND SANCTIONS 1037

sensitive denial of speech protection creates room for a sort of “efficient

breach” of government imposed speech restriction.

195

Second, the penalty-sensitive rule fits well with the First Amendment

values of enhancing autonomy and promoting individual self-realization.

When compared to an absolute denial of speech protection, a penalty-

sensitive denial furnishes a greater degree of choice to individuals con-

templating participation in expressive activity. This choice, to be sure, is

not unrestrained. By withholding absolute protection, the Court denies

individuals the freedom to speak without facing punishment as a result.

But with strict enough limits on punishment in place, the choice remains

open in the sense that speakers can reasonably decide to absorb the costs

of punishment—purchasing, as it were, their right to speak at a manage-

able and predictable price.

196

Whereas penalty-neutral denials of speech

195. Cf. Matthew Stephenson, The Price of Public Action, 118 Yale L.J. 2, 4–6 (2008).

Professor Stephenson’s article, which is also about constitutional rights review, draws a

similar lesson from the efficient breach phenomenon. In particular, Stephenson argues

that “courts often can, do, and should craft doctrines that raise the costs to government

decisionmakers of enacting constitutionally problematic policies, rather than attempting

to designate certain government actions, or categories of government actions, as permissi-

ble or impermissible.” Id at 4. His argument begins from the thought that many types of

constitutional cases call for an implicit weighing of legitimate governmental purposes

against constitutionally relevant harms. Id. at 5. In an ideal world, all-knowing courts could

strike the appropriate balance between these interests, but in the real world, courts face

serious informational difficulties in determining what this balance should be. As a result, a

strictly category-oriented approach to judging will produce constitutional rules that either

overprotect or underprotect individual rights. To reduce informational difficulties,

Stephenson favors an “enactment cost” strategy, which permits the passage of constitu-

tionally problematic laws, but only in a manner that imposes significant costs on their en-

actors. Id. at 11. Such a strategy—implemented, for example, by way of clear statement

rules, narrow tailoring requirements, legislative history canons, etc.—allows courts to dele-

gate informational burdens to the legislature without fully abdicating their rights-

protecting responsibilities. Id. at 6. As Stephenson explains, “constitutional doctrines that

raise the costs associated with problematic government enactments may help deter policies

that are ‘inefficient’—in the broad sense of failing a hypothetical ideal constitutional bal-

ancing test—while allowing what might be thought of as ‘efficient breaches’ of constitu-

tional rights.” Id.

While Stephenson’s analysis identifies a means of managing informational difficulties

on the “government interest” side of the balancing equation, my analysis here identifies a

means of doing so on the “liberty interest” side. When, that is, the bulk of a court’s infor-

mational difficulties relate to estimating the “value” of the liberty interests implicated by a

constitutionally problematic enactment, penalty-sensitive doctrine provides an indirect

means of preserving the most valuable instances of a speech act while filtering out its less

valuable variants. Just as Stephenson’s enactment cost strategy creates room for “efficient

constitutional breaches” by public officials, a penalty-sensitive strategy creates room for

“efficient constitutional invocations” by private actors.

196. Penalty-sensitive doctrine of this sort might also be conceived of as implement-

ing a sort of constitutional liability rule, under which speakers are permitted to purchase

the entitlement to circumvent a speech-infringing law at a preset price. See generally

Guido Calabresi & A. Douglas Melamed, Property Rules, Liability Rules, and Inalienability:

One View of the Cathedral, 85 Harv. L. Rev. 1089, 1092 (1972); cf. David A. Strauss, First

Amendment Entitlements and Government Motives: A Reply to Professor Merrill, 93 Nw.

1038 COLUMBIA LAW REVIEW [Vol. 112:5

protection allow the government to set the punitive prices at levels that

individuals cannot seriously contemplate paying (such as a lengthy term

in prison), a penalty-sensitive denial of speech protection may set costs

that, given the right conditions, individuals can and will choose to en-

dure (for example, a small fine). Thus, for courts and scholars who see

the free speech right as grounded in the value of autonomy,

197

this

penalty-sensitive approach dovetails with one of the First Amendment’s

animating purposes.

In this respect, civil libertarians disaffected with the Court’s inde-

cency jurisprudence should find some consolation in the penalty-

sensitive aspects of the Pacifica holding. Pacifica may not have ushered in

a golden age of no-holds-barred programming, but the decision at least

deserves credit for staving off a dark age of enforced blandness. And that

is because, by stressing the importance of limited punishments, Pacifica

withheld from the government the tool of total, nonnegotiable censor-

ship. Under the reasoning of Pacifica, when broadcasters strongly desire

to air the seven dirty words, the government can do only so much to dis-

suade them: It can exact a price for the broadcaster’s act of defiance, but

it cannot erect insurmountable barriers to indecent broadcasts.

198

Thus,

whatever screen Pacifica and its progeny have placed in the way of inde-

cent programming, the penalty-sensitive aspects of this case law have

U. L. Rev. 1205, 1212 (1999) (“Even if the government may not forbid, buy out, or con-

demn tobacco advertising, perhaps it should be allowed to require tobacco advertisers to

pay for the harm that their advertising inflicts.”). For elaboration on this idea, see Gerard

N. Magliocca, Constitutional Liability Rules 23–26 (2011) (unpublished manuscript),

available at http://www.law.northwestern.edu/colloquium/constitutionallaw/documents/

Fall2011Magliocca-Constitutional-Liability-Rules.pdf (on file with the Columbia Law Review)

(proposing, among other things, tax-oriented campaign finance doctrine that “hold[s]

that corporations have the First Amendment entitlement that Citizens United identified, but

that this right can be taxed in the interest of protecting the political system from corrup-

tion or the appearance of corruption”). It bears noting that this sort of approach might

present untoward distributive consequences, insofar as it affords deep-pocketed parties

greater freedom to flout speech restrictions than their shallow-pocketed counterparts.

Whether this problem merits attention from courts charged with implementing constitu-

tional doctrine—as opposed to policymakers with direct control over the distribution of

wealth—is an important and interesting question, but it is ultimately one that lies beyond

the scope of this project.

197. See, e.g., Martin H. Redish, Freedom of Expression: A Critical Analysis 11

(1984) (arguing “the constitutional guarantee of free speech ultimately serves only one

true value, . . . labeled ‘individual self-realization’”); Emerson, Toward a General Theory,

supra note 72, at 879–81 (identifying “individual self-fulfillment” as First Amendment

value, in light of idea that “[t]he right to freedom of expression is justified first of all as the

right of an individual purely in his capacity as an individual”).

198. See supra notes 113–115 and accompanying text (discussing penalty-sensitive

aspects of Pacifica and lower courts’ reading of case as prohibiting excessive punishment of

indecency).

2012] OF SPEECH AND SANCTIONS 1039

loosened its mesh, ensuring that some forms of “unprotected” speech

can still get through.

199

D. Government Motives

Courts often face difficulties in ascertaining the purposes underlying

a law with speech-infringing effects. These difficulties prompted the

Court in United States v. O’Brien to declare that it “will not strike down an

otherwise constitutional statute on the basis of an alleged illicit legislative

motive.”

200

As then-professor and now-Justice Elena Kagan has shown,

however, this motive-neutral approach to First Amendment law has not

won the day. In fact, the Court often probes the reasons underlying a

law’s enactment, and it has crafted doctrines with an eye to the problem

of improper motive.

201

Kagan has thus explained in motive-based terms

the Court’s distinction between “content-based” and “content-neutral”

laws, as well as its distinction between direct and incidental restrictions

on speech. While these doctrinal categories do not explicitly reference

government motives, they provide tools that courts routinely use to “flush

out bad motives without directly asking about them.”

202

Penalty-sensitive review works the same way. Just as, for example,

testing for content neutrality can serve as a “prox[y] for a direct inquiry

into motive,” so too can scrutinizing the severity of a law’s penalties.

203

Recall, for example, Justice Holmes’s observation that anything more

than a nominal punishment would have caused the Abrams defendants

“to suffer not for what the indictment alleges but for the creed that they

199. That broadcasters continue to show indecent material notwithstanding the

threat of punishment is illustrated by the agency’s omnibus order of March 15, 2006. See

Complaints Regarding Various Television Broadcasts Between February 2, 2002 and March

8, 2005, 21 FCC Rcd. 2664 (2006) (cataloging and imposing fines for multiple television

broadcasts deemed indecent); see also FCC v. Fox Television Stations, Inc., 129 S. Ct.

1800, 1808–10, 1812 (2009) (describing order and upholding it against challenge under

Administrative Procedure Act), remanded to 613 F.3d 317, 319 (2d Cir. 2010) (invalidat-

ing order on First Amendment vagueness grounds), cert. granted, 131 S. Ct. 3065 (2011).

What matters for our purposes is not whether the FCC correctly identified the broadcasts

deserving of punishments, or whether the policy underlying the order was unconstitution-

ally vague. Rather, what matters is that broadcasters, fully aware of the potential for pun-

ishment, have continued to show indecent material.

200. 391 U.S. 367, 383 (1968).

201. Elena Kagan, Private Speech, Public Purpose: The Role of Governmental Motive

in First Amendment Doctrine, 63 U. Chi. L. Rev. 413, 414 (1996) (contending “First

Amendment law, as developed by the Supreme Court . . . has as its primary, though un-

stated, object the discovery of improper governmental motives”); see also Paul Brest, The

Conscientious Legislator’s Guide to Constitutional Interpretation, 27 Stan. L. Rev. 585,

590 (1975) (using O’Brien to support proposition that Supreme Court “decisions inform

the conscientious legislator that some motives are unconstitutional”); Jed Rubenfeld, The

First Amendment’s Purpose, 53 Stan. L. Rev. 767, 776 (2001) (arguing “the real function

of the O’Brien test is nothing other than ascertaining the law’s purpose”).

202. Kagan, supra note 201, at 443.

203. Id. at 414.

1040 COLUMBIA LAW REVIEW [Vol. 112:5

avow.”

204

If pressed to elaborate on this idea, Holmes might have said

something like the following:

Look, I recognize that the government has a legitimate interest

in safeguarding wartime production, and I recognize that

Abrams and his colleagues may have undermined this interest.

But when I turn to their punishments, I become extremely

skeptical of the government’s stated motivations. These people

were puny anonymities, whose leaflets circulated within a super-

limited sphere! Any adverse effects on wartime output would

have been negligible at best. The twenty-year prison sentences

thus speak volumes about the government’s true motives here,

motives relating not to a benign interest in preventing indus-

trial disturbances, but to a more sinister interest in persecuting

dissenters.

Applied in this manner, penalty-sensitive review serves an evidentiary

function, smoking out illegitimate purposes in much the same way as the

motive-oriented proxies that Justice Kagan has identified.

The Abrams example reveals how penalty sensitivity facilitates ex post

review of government motives. But motive-based monitoring of penalties

can assume an ex ante perspective, as well. This Article earlier character-

ized Gertz’s bar on presumed and punitive damages as a prophylactic

measure designed to alleviate chilling effects.

205

But the rule also carries a

motive-based justification, reflected in Justice Powell’s observation that

permitting juries to go beyond compensation for actual injury would

open the door to improperly motivated defamation awards. Allowing

presumed damages, he explained, “invites juries to punish unpopular

opinion,” and allowing punitive damages leaves juries “free to use their

discretion selectively to punish expressions of unpopular views.”

206

Put

briefly, the less regulated the jury’s power to punish, the greater the dan-

ger of belief-based condemnation. Thus arises the need for a limitation

on damage awards, which guards against the risk of punishing unpopular

views.

The idea can also be viewed through the lens of means-end analysis.

In its traditional formulation, this commonplace form of inquiry asks

whether the regulatory sweep of a speech restriction is no more expan-

sive than is appropriate to achieve a legitimate (or important) govern-

ment interest. Frequently surfacing within First Amendment law, for ex-

ample, is the rule of O’Brien, which asks whether “the incidental restric-

tion on alleged First Amendment freedoms is no greater than is essential

to the furtherance of [an important or substantial government] inter-

est.”

207

Penalty-sensitive analysis can involve much the same sort of exami-

204. Abrams v. United States, 250 U.S. 616, 629 (1919) (Holmes, J., dissenting).

205. See supra notes 184–186 and accompanying text (discussing Gertz’s role in miti-

gating chilling effects).

206. Gertz v. Robert Welch, Inc., 418 U.S. 323, 349–50 (1974).

207. United States v. O’Brien, 391 U.S. 367, 377 (1968).

2012] OF SPEECH AND SANCTIONS 1041

nation, focusing attention on whether a law’s sanction is harsher than a

legitimate government interest would require. Both Holmes’s Abrams dis-

sent and the Gertz opinion can be understood in this way. As Holmes saw

it, the Espionage Act did not just cover more speech than was necessary

to achieve the legitimate interest of protecting wartime supplies; it also

punished the speech more severely than necessary to achieve the inter-

est. Similarly, as the Gertz majority saw it, defamation actions are constitu-

tionally problematic not just because they threaten to restrict more

speech than necessary to remedy the injuries of private citizens but also

because they threaten to punish defendants more harshly than necessary

to achieve this interest. Both opinions identified a less restrictive means

in a less punitive means, diagnosing tailoring defects not just in the scope

of a speech restriction, but also in the severity of the penalties connected

to it.

The Court’s disposition of O’Brien itself might well have benefited

from an analysis informed by these considerations. The tailoring ques-

tion in that case was whether the 1965 Amendment to the Selective

Service Act (SSA)—which criminalized the willful destruction or mutila-

tion of draft cards—went beyond advancing the government’s interest in

“further[ing] the smooth and proper functioning of the system that

Congress has established to raise armies.”

208

The Court concluded that

the Act did not, reasoning that the prohibition covered no more speech

than necessary to serve the substantial government interest in administer-

ing the Selective Service System.

209

But by asking only the penalty-neutral

question of whether the SSA’s substantive restrictions were narrowly tai-

lored, the Court ignored troubling facts about the severity of O’Brien’s

punishment. As Professor Lucas Powe has noted, the district court sen-

tenced O’Brien to “six years under the Youth Offenders Act, which

meant that he might serve four years at the Federal Youth Correctional

Center in Chillicothe, Ohio.”

210

This fact alone should have raised eye-

brows, given that “[e]ven those who refused to report for induction in

the armed forces received lighter sentences, typically between one and

three years.”

211

Indeed, anyone convicted under the recently amended Act

“was subject to punishment that would last longer than a voluntary

enlistment in the armed forces.”

212

And on top of everything else, some

evidence suggested that the trial judge had imposed the maximum sen-

tence as a direct consequence of O’Brien’s refusal to renounce his politi-

cal beliefs.

213

In short, these and other considerations indicated that “the

208. Id. at 378–81.

209. Id. at 382.

210. Lucas A. Powe, Jr., The Warren Court and American Politics 326–27 (2000).

211. Id. at 327.

212. Id. at 326.

213. Id.; see also Brief for David Paul O’Brien at 9–10, 75–76, O’Brien, 391 U.S. 367

(Nos. 232, 233), 1968 WL 129291, at *9–10, *75–76 (“[T]he sentencing judge imposed a

1042 COLUMBIA LAW REVIEW [Vol. 112:5

punishment was keyed not to the conduct of burning a small piece of

paper but to the hated message the burning conveyed.”

214

The critical point is this: Even if the law’s substantive prohibitions

reflected an adequate “fit” with the government’s interest in administer-

ing the SSA, the law’s penalties seemed far more exacting than achieve-

ment of that interest would require. In O’Brien’s case, the six-year sen-

tence was hard to square with the permissible objective of implementing

the draft and easier to square with the impermissible objective of sup-

pressing dissent. Thus, much like the sentencing enhancement struck

down in Dawson,

215

O’Brien’s punishment could plausibly have been

characterized as carrying an impermissible belief-based “enhance-

ment”—penalizing him not just for the noncommunicative aspects of his

actions but also for the message that his actions conveyed. A determina-

tion to this effect would not have supported a wholesale invalidation of

O’Brien’s conviction; insofar as his conduct was punishable to some ex-

tent, the proper remedy would have been merely a remand for resen-

tencing. The Court, however, failed to consider this possibility. In doing

so, it overlooked an especially problematic element of the O’Brien prose-

cution.

E. Transparency

Recall that Justice Holmes’s decision to dissent in Abrams stemmed

in part from a visceral unease with the lengthy sentences involved in that

case. Along the same lines, the Court’s willingness to uphold the NEA

guidelines in Finley derived from its comfort with the mildness of the

“sanction” at issue there, just as its willingness to strike down the CPPA’s

child pornography prohibitions in Free Speech Coalition derived from its

discomfort with the high severity of that law’s punishments. These exam-

ples are noteworthy because they involve judges openly acknowledging

the First Amendment significance of a penalty’s severity. In another

sense, however, they may represent the tip of the iceberg. There almost

surely exist other, perhaps many other, cases in which penalty-related

considerations have exerted real, but invisible, effects on the outcomes of

cases.

six-year maximum indeterminate sentence with the express intent that respondent would

serve less than the maximum if he changed his beliefs and associations.”).

214. Powe, supra note 210, at 326.

215. Dawson v. Delaware, 503 U.S. 159, 160, 163 (1992). Dawson is not itself a penalty-

sensitive case, but it nonetheless illuminates the purposivist potential of penalty sensitivity

analysis. See supra text accompanying notes 141–144. Understood, that is, as a case about

motives, Dawson underscores the relationship between penalties and purposes, suggesting

that, all else equal, the more severe an individual’s punishment, the more likely it is to

bear the taint of improper motives. See id. And when that is true, courts can mitigate the

influence of improper motives by limiting the penalty that the government may impose.

2012] OF SPEECH AND SANCTIONS 1043

Empirical study is needed to test this hypothesis, but it requires no

great leap of imagination to see how penalty-sensitive considerations

might, sub silentio, affect judicial resolution of First Amendment claims.

The literature on federal sentencing is telling in this respect. Numerous

commentators, including some members of the federal judiciary, have

observed that judges manipulate legal and factual findings so as to avoid

imposing sentences they perceive to be unfair.

216

Similarly, it has been

observed that jurors become less likely to return convictions after learn-

ing that these convictions will result in especially serious sentences.

217

216. See, e.g., Kate Stith & Jose A. Cabranes, Fear of Judging: Sentencing Guidelines

in the Federal Courts 90 (1998) (suggesting “[m]any judges are not at ease operating

within [the Guidelines system], and may be sorely tempted to manipulate their Guidelines

calculations to avoid the results called for by the Guidelines”); Andrew D. Liepold, Why

Are Federal Judges So Acquittal Prone?, 83 Wash. U. L.Q. 151, 200–01 (2005) (speculating

“judges may acquit more often because they [find] it to be the only way to avoid imposing

an unjust sentence that they know would follow a conviction” and providing limited em-

pirical evidence supporting claim); Joanna Shepherd, Blakely’s Silver Lining: Sentencing

Guidelines, Judicial Discretion, and Crime, 58 Hastings L.J. 533, 560–61 (2007) (“[B]oth

judges and juries are more likely to acquit as the punishment following a conviction in-

creases.”); Jack B. Weinstein, A Trial Judge’s Second Impression of the Federal Sentencing

Guidelines, 66 S. Cal. L. Rev. 357, 365 (1992) (quoting anonymous colleague’s concession

that he and his colleagues “spend . . . time plotting and scheming, bending and twisting,

distorting and ignoring the law in an effort to achieve a just result” (internal quotation

marks omitted)).

217. See, e.g., Shannon v. United States, 512 U.S. 573, 579 (1994) (justifying long-

standing rule against providing jurors with sentencing information on grounds that doing

so would “invite[] them [inter alia] to ponder matters that are not within their province”);

Rachel E. Barkow, Recharging the Jury: The Criminal Jury’s Constitutional Role in an Era

of Mandatory Sentencing, 152 U. Pa. L. Rev. 33, 79–80 (2003) (arguing “although it is ad-

mittedly more difficult for juries to check laws that would, in their estimation, produce

overly harsh results when they are unaware of what the actual sentence will be,” modern

jury “does attempt to make such predictions, and if it predicts that the punishment will be

disproportionate to what the defendant did, it is less likely to convict”); Neal Kumar

Katyal, Deterrence’s Difficulty, 95 Mich. L. Rev. 2385, 2450–51 (1997) (“As penalties in-

crease, people may not be as willing to enforce them because of the disproportionate im-

pact on those caught.”); Erik Lillquist, The Puzzling Return of Jury Sentencing: Misgivings

About Apprendi, 82 N.C. L. Rev. 621, 656 (2004) (arguing that, in rational choice model,

“the jury’s guilt decision may be based upon assumptions about the probable sentence”);

Leading Cases, Sixth Amendment—Allocation of Factfinding in Sentencing, 121 Harv. L.

Rev. 225, 235 (2007) (“A jury verdict will meaningfully reflect the community’s conscience

and rein in an overly punitive legislature only when it is the product of knowledge, not ig-

norance, about sentencing.”). Although the law review literature provides only anecdotal

evidence for this claim, social scientists have accumulated some experimental support. See

Martin F. Kaplan, Setting the Record Straight (Again) on Severity of Penalty: A Comment

on Freedman et al., 18 Law & Hum. Behav. 697, 698 (1994) (reviewing experimental at-

tempts to verify existence of this phenomenon and concluding “there [is] evidence that

penalty does affect decisions under some circumstances”). But see Jonathan L. Freedman,

Kristen Krismer, Jennifer E. MacDonald & John A. Cunningham, Severity of Penalty,

Seriousness of the Charge, and Mock Jurors’ Verdicts, 18 Law & Hum. Behav. 189, 190

(1994) (“[T]here is no convincing support for [this] effect.”). And there is historical sup-

port as well. During the seventeenth and eighteenth centuries, for example, it was not un-

common for English juries—sometimes with the encouragement of their overseeing

1044 COLUMBIA LAW REVIEW [Vol. 112:5

Based on these examples, it is not hard to infer that penalty-sensitive free

speech review sometimes occurs behind closed doors, with judicial actors

contorting an outwardly penalty-neutral analysis to achieve a result that

satisfies inwardly penalty-sensitive aims.

This possibility points to a further benefit of the penalty-sensitive

approach: If penalty-sensitive concerns are already functioning as a hid-

den variable within the First Amendment calculus, then further recogniz-

ing the legitimacy of these concerns should increase the transparency of

judicial decisionmaking, which in turn should help everyone involved in

First Amendment analysis—from litigants to scholars to legislatures to

speakers—better understand the actual contours of the law.

It is true, of course, that not every hidden influence on constitu-

tional decisionmaking should be thrust into the light of full disclosure.

Even if we could prove the realist maxim that breakfast food influences

judicial decisionmaking,

218

we would still discourage judges from writing

opinions about bacon, toast, and eggs. A similar claim applies to penalty

severity: Even if judges are inwardly motivated by penalty-sensitive con-

cerns, this fact alone does not indicate that they should publicize, rather

than suppress, such inclinations. But, as we have seen from the discussion

above, far from being a problematic influence on free speech adjudica-

tion, penalty-sensitive analysis is in fact a meritorious method of First

Amendment review, capable of generating fairer and more finely tailored

rules of speech protection. Thus, to the extent that penalty sensitivity al-

ready operates as a hidden adjudicatory influence, judges should not

hesitate to make its presence known.

IV.

OBJECTIONS TO THE PENALTY-SENSITIVE APPROACH

With the justifications for penalty-sensitive review now laid on the

table, it is worth considering potential criticisms to the practice. This Part

takes on that task. Part IV.A speaks to a legitimacy-based critique, asking

whether penalty-sensitive analysis usurps the legislative role. Part IV.B

turns to objections based on administrability, evaluating the extent to

which courts are capable of conducting penalty-sensitive analyses without

generating arbitrary or abstruse doctrinal rules. And Part IV.C addresses

the claim that penalty-sensitive analysis will underprotect the free speech

right.

judges—to commit what Blackstone called “pious perjury,” deliberately convicting defen-

dants of lesser charges so as to spare them from especially harsh punishments. See John H.

Langbein, The Origins of Adversary Criminal Trial 57–60 (2003) (describing this phe-

nomenon).

218. Cf. Shai Danziger, Jonathan Levav & Liora Avnaim-Pesso, Extraneous Factors in

Judicial Decisions, 108 Proc. Nat’l Acad. Sci. U.S. 6889, 6890 (2011) (observing outcomes

of parole board hearings bear statistically significant correlation to temporal proximity to

food breaks).

2012] OF SPEECH AND SANCTIONS 1045

In parrying these thrusts, I do not mean to imply that penalty sensi-

tivity is immune to criticism. In fact, as the ensuing discussion will dem-

onstrate, the practice has its vulnerabilities, and these vulnerabilities, if

not carefully attended to, can create problems. That penalty-sensitive

analysis might cause problems, however, does not mean that courts

should forsake the practice altogether. Rather, the presence of potential

downsides means only that penalty-sensitive analysis should proceed in a

manner that takes them into account. In this respect, the objective of this

last Part is twofold: In addition to demonstrating that the problems with

penalty sensitivity are manageable, the ensuing discussion proposes some

ways in which courts can avert its dangers.

A. Legitimacy

A legitimacy-based attack on penalty-sensitive review might advance

the claim—often articulated in Eighth Amendment and due process

cases—that unelected judges have no business reviewing a legislature’s

chosen means of punishment. In the criminal context, the Court has de-

clared that “[r]eviewing courts . . . should grant substantial deference to

the broad authority that legislatures necessarily possess in determining

the types and limits of punishments for crimes, as well as to the discre-

tion that trial courts possess in sentencing convicted criminals.”

219

In the

civil context, this same principle—though perhaps less rigidly adhered

220

—finds expression in admonitions that courts should afford “‘sub-

stantial deference’ to legislative judgments concerning appropriate sanc-

tions for the conduct at issue.”

221

Along these lines, one might criticize

penalty-sensitive review as insufficiently deferential to coordinate

branches, arguing that it gives unelected judges too much authority to

meddle with the difficult punitive judgments made by democratically

elected officials. These concerns weighed on the Court, for instance, in

Fort Wayne Books, Inc. v. Indiana. There, in explaining the majority’s re-

fusal to consider the harsh penalties attached to an anti-obscenity law,

Justice White reasoned that “‘it is not for this Court . . . to limit the State

in resorting to various weapons in the armory of the law.’”

222

Arguments of this sort, however, overlook a critical distinction be-

tween pure-form proportionality review and penalty-sensitive free speech

219. Solem v. Helm, 463 U.S. 277, 290 (1983).

220. See Karlan, Pricking the Lines, supra note 30, at 920 (“Having embraced and

then largely abandoned a judicially enforceable constitutional requirement of proportion-

ality under the Eighth Amendment in criminal cases, the Court has articulated an increas-

ingly robust requirement of proportionality under the Due Process Clause in punitive

damages cases.”).

221. Browning-Ferris Indus. of Vt., Inc. v. Kelco Disposal, Inc., 492 U.S. 257, 300–01

(1989) (O’Connor, J., concurring in part and dissenting in part).

222. 489 U.S. 46, 60 (1989) (alterations in original) (quoting Kingsley Books, Inc. v.

Brown, 354 U.S. 436, 441 (1957)).

1046 COLUMBIA LAW REVIEW [Vol. 112:5

review. The former asks whether the severity of a penalty is justifiable in

light of traditional punitive objectives such as retribution, deterrence, in-

capacitation, and the like. But the latter adds a key variable to the pro-

portionality equation: the constitutional value of the behavior being sanc-

tioned. What is involved here is not simply the determination that a par-

ticular punishment is, as a policy matter, too harsh or too unfair; it is

rather the determination that First Amendment priorities demand appli-

cation of light penalties, trumping whatever policy considerations might

otherwise counsel in favor of more exacting sanctions. Because penalty-

sensitive review requires this second sort of judgment—a judgment well

within courts’ purview—the legitimacy objection does not carry persua-

sive force.

223

Indeed, the legitimacy of penalty-sensitive review may be said to fol-

low a fortiori from the unquestioned legitimacy of penalty-neutral review.

Penalty-neutral analysis empowers courts to strike down sanctions in their

entirety, while penalty sensitivity merely adds the option of lessening

sanctions’ severity. The argument is not difficult to see. So long as courts

wield the sledgehammer of statutory invalidation, why not hand them the

chisel of punitive moderation? Along with the greater power to bludgeon

penalties out of existence should come the lesser power to chip away at

their severity.

224

This last point helps to illustrate why penalty-sensitive review, far

from facilitating illegitimate judicial power grabs, actually comports with

a minimalist conception of the judge’s role.

225

By offering courts an inter-

mediate option between the twin extremes of “full constitutional protec-

tion” and “no constitutional protection,” penalty sensitivity can promote

evolutionary, rather than abrupt, doctrinal change. If the Court wishes to

protect a new category of speech, penalty sensitivity allows it to make this

223. Cf., e.g., United States v. Carolene Prods. Co., 304 U.S. 144, 152 n.4 (1938)

(“There may be narrower scope for operation of the presumption of constitutionality

when legislation appears on its face to be within a specific prohibition of the Constitution,

such as those of the first ten amendments, which are deemed equally specific when held to

be embraced within the Fourteenth.”).

224. Along similar lines, Professor William Stuntz once claimed that, to the extent

that constitutional law prohibits legislative constraints on juries’ power to acquit criminal

defendants, it should a fortiori prohibit legislative constraints on juries’ power to reduce

criminal sentences:

It is hard to understand why constitutional law should make it impossible for leg-

islatures to command that a given course of conduct be punished (the power to

acquit for any reason does away with that legislative power, at least in theory),

and yet leave legislatures free to require that, if behavior is to be punished, it

should be punished at least so much. Logically, the greater mercy ought to in-

clude within it the lesser.

William J. Stuntz, The Pathological Politics of Criminal Law, 100 Mich. L. Rev. 505, 596

(2001).

225. See generally Cass R. Sunstein, One Case at a Time: Judicial Minimalism on the

Supreme Court (1999) (developing theory of judicial minimalism).

2012] OF SPEECH AND SANCTIONS 1047

change gradually, transitioning into a rule of partial constitutional pro-

tection, letting coordinate constitutional actors respond accordingly,

and, if conditions seem appropriate, then proceeding to a rule of total

protection. The Court may likewise use penalty-sensitive adjudication to

“phase out” existing rules of speech protection—replacing them at first

with rules that furnish partial speech protection, followed eventually by

rules that furnish no protection at all. Penalty sensitivity, in short, makes

possible constitutional change through small steps rather than giant

leaps. It is therefore capable of “reduc[ing] the burdens of judicial deci-

sion,” “mak[ing] judicial errors less frequent,” “promot[ing] more de-

mocracy and more deliberation,” and providing other benefits attendant

to the minimalist approach.

226

B. Administrability

A more powerful objection to penalty sensitivity centers on imple-

mentation concerns. This objection posits that penalty-sensitive review—

while perhaps attractive in theory—is simply too difficult to operational-

ize in practice. Here, rather than the dishonoring of doctrinal traditions

or the arrogation of too much judicial power, the nightmare scenario en-

visions tangled webs of penalty-sensitive rules that cannot be easily ap-

plied or understood. On this view, penalty-sensitive adjudication is bad

because it threatens to undermine the coherence and predictability of

free speech doctrine.

This objection has some merit. Suppose, for example, that Holmes

had gotten his way in Abrams, such that seditious libel was subject only to

“nominal punishments.” This rule would leave important matters unre-

solved: What counts as a “nominal punishment”? Where do the bounda-

ries lie between speech that is fully punishable, nominally punishable,

and not punishable at all? Does the nominal-punishment limitation apply

only to criminal sanctions, or does it extend to civil sanctions as well? Are

all prison sentences per se not nominal? Can fines not be nominal? If so,

how hefty would the fines have to be? Does the rule tolerate higher pun-

ishments when the prosecution involves repeat offenders? And so on.

Such questions do not answer themselves, and they would create un-

certainty for courts, legislatures, and potential litigants. Moreover, ques-

tions like these are likely to arise in connection with almost every penalty-

sensitive holding the Court hands down. To address the implementation

objection, then, one must ask whether the costs of this confusion are

likely to outweigh the benefits penalty sensitivity can provide.

226. Id. at 3–4; see also Amanda Frost, The Limits of Advocacy, 59 Duke L.J. 447, 481

(2009) (“Minimalists favor incremental steps over sweeping changes in legal norms be-

cause they fear that broadly worded decisions will undermine democratic processes, lead

to unintended consequences, and put in place rigid rules that leave no flexibility for the

future.”).

1048 COLUMBIA LAW REVIEW [Vol. 112:5

At least as a general matter, this does not appear to be a likely out-

come. To begin, one must remember that some amount of doctrinal in-

determinacy is, and will always be, a fact of legal life. The Court routinely

adopts doctrinal formulations that require further clarification; indeed,

“it is an essential part of adjudication to draw distinctions, including fine

ones, in the process of interpreting the Constitution.”

227

First

Amendment doctrine is no stranger to this process. The Court, for in-

stance, has recognized constitutional distinctions between “public” and

“private” speech,

228

between “content” and “viewpoint” discrimination,

229

between “primary” and “secondary” effects,

230

between “substantial” and

“insubstantial” disruptions,

231

and between many other conceptual cate-

gories of uncertain definition.

232

It is difficult to see why similar distinc-

tions between and among types of sanctions would be any more difficult

to flesh out.

Recall, too, that the Court has already drawn, and continues to draw,

penalty-based distinctions within many other areas of rights-related doc-

trine.

233

This point is significant for two reasons. First, it suggests that the

line drawing required by penalty-sensitive judging is not an impossible

task. Courts have exercised sober judgment in performing this task, and

227. Walz v. Tax Comm’n, 397 U.S. 664, 679 (1970).

228. See, e.g., Snyder v. Phelps, 131 S. Ct. 1207, 1215 (2011) (noting that “[w]hether

the First Amendment prohibits holding [the defendant] liable for its speech in this case

turns largely on whether that speech is of public or private concern”).

229. See, e.g., Rosenberger v. Rector of the Univ. of Va., 515 U.S. 819, 830 (1995)

(noting operator of limited public forum may engage in “content discrimination, which

may be permissible if it preserves the purposes of that limited forum,” but not “viewpoint

discrimination, which is presumed impermissible when directed against speech otherwise

within the forum’s limitations”).

230. See, e.g., United States v. Playboy Entm’t Grp., Inc., 529 U.S. 803, 815 (2000)

(“We have made clear that the lesser scrutiny afforded regulations targeting the secondary

effects of crime or declining property values has no application to content-based regula-

tions targeting the primary effects of protected speech.”).

231. See, e.g., Morse v. Frederick, 127 S. Ct. 2618, 2637 (2007) (Alito, J., concurring)

(noting First Amendment “permits the regulation of student speech that threatens a con-

crete and ‘substantial disruption’” (citing Tinker v. Des Moines Indep. Cmty. Sch. Dist.,

393 U.S. 503, 514 (1969))).

232. See, e.g., Christian Legal Soc’y Chapter of the Univ. of Cal., Hastings Coll. of the

Law v. Martinez, 130 S. Ct. 2971, 2984 n.11 (2010) (summarizing different doctrinal rules

governing speech regulations in “traditional public forums,” “designated public forums,”

and “limited public forums”); United States v. Williams, 128 S. Ct. 1830, 1835–36 (2008)

(“We have long held that obscene speech—sexually explicit material that violates funda-

mental notions of decency—is not protected by the First Amendment. But . . . we have lim-

ited the scope of the obscenity exception, and have overturned convictions for the distri-

bution of sexually graphic but nonobscene material.” (citation omitted)); Wash. State

Grange v. Wash. State Republican Party, 128 S. Ct. 1184, 1190 n.6 (2008) (“[A] law may be

overturned as impermissibly overbroad because a ‘substantial number’ of its applications

are unconstitutional . . . .” (quoting New York v. Ferber, 485 U.S. 747, 769–71 (1982))).

233. See supra Part II.B (discussing, among other things, due process, equal protec-

tion, and Takings Clause jurisprudence).

2012] OF SPEECH AND SANCTIONS 1049

the consequences have not been disastrous. To take one example, the

Contract Clause rule that only “substantial” impairments of contracts

create a right to compensation is not a rule that lends itself to mechani-

cal application. But that fact has not stopped judicial actors from recog-

nizing a core constitutional distinction between substantial and nonsub-

stantial impairments, while leaving disputes at the margins to be resolved

as they arise.

234

Second, the existence of an already-developed body of penalty-

sensitive case law means that penalty-sensitive First Amendment analysts

need not paint on a blank canvas. In developing penalty-sensitive free

speech doctrine, courts can incorporate distinctions they have already

made within other areas of law—areas in which they have already spent

time and energy elaborating on these distinctions’ meaning. We have

seen such a move, for example, in the public employment context, where

some lower courts employ a penalty-sensitive rule of speech protection

derived from Title VII’s “adverse employment action” requirement. So

too, we might imagine First Amendment adjudicators plucking other

formulations from other doctrinal branches—for example, the “petty

offense” rule from right-to-counsel case law, the “infamous crimes” rule

from grand jury case law, or even the “substantial impairment” rule from

Contract Clause case law. In this field, as in others, courts can build on

their prior work, at once easing administrative burdens for themselves

and reducing doctrinal uncertainty for others.

The implementation objection must also deal with the demonstrated

adaptability of penalty-sensitive analysis. In some cases, such as Pacifica,

the severity of an individual’s penalty can function as a soft doctrinal

variable, exerting some force on the validity of the free speech claim, but

not carrying outcome-determinative weight. In other cases, such as Gertz,

234. A potential problem with case-by-case identification of constitutionally problem-

atic penalties is highlighted by Professor Eugene Volokh’s discussion of judge-drawn dis-

tinctions based on the severity of crimes. See Eugene Volokh, Crime Severity and

Constitutional Line-Drawing, 90 Va. L. Rev. 1957, 1978 (2004). As Professor Volokh notes,

these distinctions may become (and in some cases have become) unstable. Having identi-

fied a limited category of crimes as sufficiently severe to warrant application of a special

constitutional rule, courts may feel “tempted to move the lines down, treating more and

more crimes as severe.” Id. at 1983. So too, we might suppose that having declared that

some speech is not punishable by one means, courts will find it difficult in subsequent

cases to declare the speech punishable by a more severe means; or, conversely, having de-

clared that some speech is punishable by one means, courts may be unable to declare it

not punishable by a less severe means. The concern, in other words, is that what sets out to

be penalty-sensitive doctrine ultimately will collapse into penalty-neutral doctrine, owing

to courts’ reluctance to follow through on promises to distinguish penalties based on their

severity. That is a valid concern, to be sure. At the same time, the doctrines discussed in

Part II indicate that this outcome, while possible, is far from inevitable. What is more, even

if an attempt at implementing a penalty-sensitive rule ends up reverting back to the

penalty-neutral norm, it is perhaps better to have tried and failed than never to have tried

at all.

1050 COLUMBIA LAW REVIEW [Vol. 112:5

the severity of the penalty can function as a hard variable, taking on dis-

positive constitutional significance by way of a categorical doctrinal

command.

235

Courts might also formulate rules that occupy a middle

place on the soft-hard spectrum, employing weak presumptions of consti-

tutional invalidity, conditioning application of a categorical rule on soft-

at-the-edges definitional criteria, or finding some other way to reconcile

the need for flexibility with the competing need for doctrinal clarity.

This point is important because it means that courts can tailor the

penalty-sensitive aspects of their holdings to further whatever

implementation-related priorities they have chosen to pursue. Pacifica,

for example, was a self-consciously narrow decision—a point that Justice

Stevens took pains to emphasize at the conclusion of his opinion. Among

the many non-penalty-related factors on which the Court relied were (a)

the timing of the broadcast, (b) the number of expletives used in the

broadcast, (c) the nature of the broadcast medium, and (d) the type of

program on which the expletives were featured. Pacifica, in short, “re-

quire[d] consideration of a host of variables,”

236

no one of which exerted

outcome-determinative force. Thus, in deciding Pacifica, the Court com-

mitted itself to a highly contextualized, balancing-oriented type of consti-

tutional analysis, accepting as a necessary evil the holding’s failure to

provide clear guidance to future litigators and adjudicators of indecency-

related claims. Having made a commitment to multifactor analysis, the

Court did not much increase implementation difficulties by adding

penalty-sensitive considerations to the mix. Pacifica’s import was going to

be cloudy whether or not its holding took into account the significance

of the broadcaster’s penalty. Consequently, the Court could employ

penalty-sensitive analysis without much adding to administrative difficul-

ties.

237

Gertz, by contrast, built on the foundation of the New York Times’s

“actual malice” requirement—a requirement crafted in the interest of

clarity and predictability. Here, a Pacifica-like “soft” rule would have dis-

served these values, introducing too much uncertainty into an area of

doctrine in which the Court sought to outline a clear First Amendment

boundary. But the Court could (and did) still embrace penalty sensitivity

by way of a hard rule, articulating the penalty-sensitive aspects of the Gertz

235. For an especially helpful discussion of the distinction between categoricalist and

balancing modes of First Amendment adjudication, see Blocher, supra note 32, at 381–98.

236. FCC v. Pacifica Found., 438 U.S. 726, 750 (1978).

237. This is not necessarily to say that indecency cases are best dealt with by way of ad

hoc balancing. All told, Pacifica’s reliance on this approach may have created more prob-

lems than it solved—providing too little guidance as to the scope of the First

Amendment’s indecency protections and leaving too much up to the subjective prefer-

ences of regulators and lower court judges. But whether or not Pacifica’s doctrinal strategy

was the correct one, the point here is that this strategy, once adopted, did not suffer on

account of the Court’s willingness to treat penalty severity as one of the “host of variables”

bearing on its resolution of the case. Id.

2012] OF SPEECH AND SANCTIONS 1051

holding in precise and outcome determinative terms. Thus, just as the

Pacifica Court structured its penalty-sensitive analysis to cohere with ad

hoc balancing, the Gertz Court structured its analysis to cohere with rule-

based categoricalism.

There is a final implementation-related point worth considering

here—primarily applicable to penalty-sensitive rules of the hard variety.

In contrast to soft rules, which purposively strive for abstractness, hard

rules raise the problem of arbitrariness, because, under them, constitu-

tional results may hinge on criteria that appear to be unduly specific and

untethered to the First Amendment text. Many would not respond well,

for example, to a hypothetical rule forbidding sixty-month prison sen-

tences for an obscenity violation but permitting fifty-nine-month prison

sentences for identical conduct or to a rule permitting school administra-

tors to punish student speakers with five-day suspensions but not six-day

suspensions. Constitutional borderlines of this sort would strike most

people as ad hoc, if not absurd.

As Gertz reveals, however, judicial use of hard-edged penalty-sensitive

rules can sidestep this problem by focusing on distinctions in kind rather

than distinctions in degree. Thus, rather than locate a constitutional

borderline somewhere in between the sixty- and fifty-nine-month prison

sentence, a reviewing court might locate the borderline between proba-

tion and imprisonment, fines and probation, or other distinctions that

capture qualitatively different forms of government-imposed punish-

ment. This way of articulating limits has the added virtue of linking the

penalty-sensitive rule to other considerations of constitutional relevance.

In Gertz, for example, the wall erected between actual damages on one

side and presumed/punitive damages on the other cohered with a sepa-

rate argument about state interests—tracking, that is, the qualitative dis-

tinction between a state’s distinctively strong interest in compensating

proven injuries and its lesser interest in simply discouraging defamatory

speech.

238

The key point is that penalty-sensitive analysis can come in different

shapes and sizes. When circumstances demand flexibility, soft penalty-

sensitive holdings can accommodate this need. When circumstances de-

238. Additionally, to the extent that numerical distinctions are a necessary compo-

nent of penalty-sensitive review, their introduction would not constitute some sort of dra-

matic transformation of the constitutional terrain. On more than one occasion, the need

for clear guidance has compelled the court to incorporate numerical boundaries into con-

stitutional doctrine. Thus, for example, the Court has speculated that the Due Process

Clause generally forbids juries from awarding punitive damages that are at least ten times

greater than compensatory damages, State Farm Mut. Auto. Ins. Co. v. Campbell, 538 U.S.

408, 424–45 (2003), it has specified that certain self-incrimination protections of criminal

suspects lapse fourteen days after their release from custody, Maryland v. Shatzer, 130 S.

Ct. 1213, 1223 (2010), and it has read the Seventh Amendment to permit six-person juries

in civil cases, while acknowledging that “at some point the number becomes too small to

accomplish [the Amendment’s] goals,” Colgrove v. Battin, 413 U.S. 149, 160 n.16 (1973).

1052 COLUMBIA LAW REVIEW [Vol. 112:5

mand predictability, hard penalty-sensitive rules can be developed. And

when circumstances demand a middle course, “in-between” penalty-

sensitive approaches are there for courts to use. None of this is to say that

penalty-sensitive review will never raise problems in implementation. The

wide range of options for implementing penalty-sensitive doctrine, how-

ever, suggests that challenges of implementation are manageable.

C. Liberty

A third objection to the penalty-sensitive approach focuses on judges

and, in particular, penalty sensitivity’s potentially adverse effects on judi-

cial decisionmaking. This objection stems from the concern that, over

time, penalty sensitivity may vitiate the strength of the free speech right

by offering judges an all-too-tempting escape hatch from difficult, but

necessary, decisions to protect unpopular speech.

Suppose, for instance, that Judge Sympatico has before him a de-

fendant sentenced to ten years in prison for violating a fighting words law

of questionable constitutional validity. The defendant raises a First

Amendment claim, arguing that his conviction should be reversed be-

cause the law is unconstitutional. Judge Sympatico is torn: On the one

hand, he thinks the First Amendment argument is pretty strong; on the

other hand, he is reluctant to let the defendant get off scot-free. He rec-

ognizes, however, that his failure to protect the speech in question will

result in an especially harsh deprivation of the defendant’s liberty, and

he therefore reluctantly, but unambiguously, declares the speech pro-

tected.

If we approve of Judge Sympatico’s decision, then we might worry

about the consequences of introducing him to the penalty-sensitive ap-

proach. Under a penalty-sensitive framework, Judge Sympatico need not

have chosen between letting the defendant completely off the hook and

subjecting him to ten years in prison. Rather, Judge Sympatico could

have said,

It’s tough to know whether the speech was constitutionally pro-

tected; certain elements of it had expressive value, but other

elements seemed to me plainly punishable. But at the very least,

a ten-year prison sentence looks plainly impermissible to me.

Therefore, I rule that you may be punished, but that your pun-

ishment may not exceed one year.

Penalty sensitivity would thus prompt Judge Sympatico to uphold a one-

year sentence under circumstances that otherwise would have led him to

allow no sentence at all.

This is a troubling scenario. But does it prove that penalty-sensitive

analysis is biased against the protection of speech? In answering this

question, consider the same hypothetical case from the perspective of

Judge Draco, who, unlike his colleague Judge Sympatico, tends to err on

the side of denying speech protection. In a penalty-neutral world, Judge

2012] OF SPEECH AND SANCTIONS 1053

Draco would have upheld the ten-year sentence; but in a penalty-sensitive

world, he might prefer the one-year constitutional remedy. For Judge

Draco, then, penalty sensitivity exerts a liberty-promoting effect, causing

him to uphold a one-year sentence when he otherwise would have up-

held a ten-year sentence.

As illustrated by the crosscutting examples of Judge Sympatico and

Judge Draco, it is hard to know whether, on the whole, penalty sensitivity

is likely to expand or curtail the free speech right. It will give some judges

an easy way out of the distasteful decision to protect speech absolutely,

letting them furnish partial protection to speech whose punishment they

would otherwise fully disallow. But it will give other judges an easy way

out of the distasteful decision to uphold a speech restriction, allowing

them to furnish partial First Amendment protection instead of no pro-

tection at all. Which of these two effects is likely to predominate? Does

the judiciary consist of more Judge Dracos or Judge Sympaticos? The an-

swer to these questions is not clear.

It is worth noting, however, that within other constitutional settings,

scholars have suggested that when the only constitutional remedy at a

judge’s disposal is to strike down a punishment in its entirety, the judge is

more likely to uphold the punishment than to strike it down—an in-

stance of what Professor Daryl Levinson has characterized as “remedial

deterrence.”

239

For example, Judge Guido Calabresi has identified the

drastic remedial consequences of the exclusionary rule as a central

source of judges’ reluctance to identify Fourth Amendment violations,

240

and Professor Sonja Starr has made a similar argument with respect to

prosecutorial misconduct claims.

241

Along similar lines, Professor Pamela

Karlan has speculated that the “windfall” remedy of per se reversal for

Batson violations has resulted in appellate courts “surreptitiously rede-

fin[ing] the right.”

242

If a similar hypothesis obtains within the free

speech setting, then the penalty-sensitive approach might ultimately ex-

pand the contours of the free speech right, by affording judges the op-

239. See Daryl J. Levinson, Rights Essentialism and Remedial Equilibration, 99

Colum. L. Rev. 857, 889–99 (1999) [hereinafter Levinson, Rights Essentialism].

240. Guido Calabresi, The Exclusionary Rule, 26 Harv. J.L. & Pub. Pol’y 111, 112

(2003); see also Harry M. Caldwell & Carol A. Chase, The Unruly Exclusionary Rule:

Heeding Justice Blackmun’s Call To Examine the Rule in Light of Changing Judicial

Understanding About Its Effects Outside the Courtroom, 78 Marq. L. Rev. 45, 53 (1994)

(observing that “courts tend to strain the Fourth Amendment to avoid the suppression of

reliable evidence”).

241. Sonja B. Starr, Sentence Reduction as a Remedy for Prosecutorial Misconduct,

97 Geo. L.J. 1509, 1511 (2009) (noting that, because sentencing reduction remedy does

not offer “windfall” to defendant, “courts will likely be more willing to invoke it than cur-

rent, nominally stronger remedies—thus serving all remedial purposes better than the

current all-or-nothing choices that usually drive courts to pick ‘nothing’”).

242. Pamela S. Karlan, Race, Rights, and Remedies in Criminal Adjudication, 96

Mich. L. Rev. 2001, 2020–21 (1998); see also Levinson, Rights Essentialism, supra note

239, at 890–92 (discussing Karlan’s analysis in similar terms).

1054 COLUMBIA LAW REVIEW [Vol. 112:5

portunity to protect speech in a manner that does not commit them to

the significant constitutional remedy of absolutely precluding its pun-

ishment.

ONCLUSION

Decades ago, Hugo Black and Felix Frankfurter waged their famous

debates over the “balancing” and “absolutist” approaches to free speech

adjudication. Black vigorously defended the absolutist position, arguing

that “the First Amendment’s unequivocal command that there shall be

no abridgment of the rights of free speech and assembly shows that the

men who drafted our Bill of Rights did all the ‘balancing’ that was to be

done in this field.”

243

Frankfurter, who carried the balancing banner,

strongly disagreed; he believed that “[a]bsolute rules would inevitably

lead to absolute exceptions,” and that “[t]he demands of free speech in a

democratic society . . . are better served by candid and informed weigh-

ing of the competing interests, within the confines of the judicial process,

than by announcing dogmas too inflexible for the non-Euclidian prob-

lems to be solved.”

244

In assessing the competing virtues of the penalty-neutral and

penalty-sensitive approaches, one sees obvious parallels to the Black-

Frankfurter debate. Penalty neutrality brooks no compromise within the

punitive realm and in so doing affords comfort to those who view the

qualified protection of civil liberties as a danger to be avoided at all costs.

Penalty sensitivity, on the other hand, comports with the idea that the

drawing of First Amendment boundaries is a difficult and delicate task,

and it enables judges to carry out this task with a degree of nuance and

refinement that penalty neutrality woodenly eschews. On the whole,

then, penalty sensitivity should be more attractive to Frankfurter-like bal-

ancers and penalty neutrality to Black-like absolutists.

This Article has taken the position that penalty sensitivity’s upsides

outweigh its downsides and that judges grappling with difficult First

Amendment problems would do well to consider the penalty-sensitive

approach. Underlying this claim is a Frankfurterian belief that penalty

neutrality’s “all-or-nothing” framework will often prove too blunt a tool

for courts confronting the free speech complexities of a modern democ-

ratic society. That is not to say that penalty neutrality has no place within

First Amendment doctrine; the penalty-neutral approach, like the abso-

lutist approach, may merit unflinching adherence in some, perhaps

many, circumstances. But it is to say that penalty-sensitive analysis has

been, and remains, a useful tool of free speech adjudication and that

courts should look to expand its presence in First Amendment law.

243. Konigsberg v. State Bar, 366 U.S. 36, 61 (1961) (Black, J., dissenting).

244. Dennis v. United States, 341 U.S. 494, 524–25 (1951) (Frankfurter, J., concur-

ring).