ROUGH DRAFT

IS SOLITARY CONFINEMENT A PUNISHMENT?
John F. Stinneford

Nulla poena sine lege—no punishment without law—is one of the oldest and most
universally accepted principles of English and American law. 1 Today, thousands of American
prisoners are placed in long-term solitary confinement2 despite the fact that such placement is
authorized neither by penal statute nor by judicial sentence. 3 Is solitary confinement “punishment
without law,” or is it a mere exercise of administrative discretion?
Imagine a prisoner is present during a violent altercation in the prison yard. The warden
suspects that the prisoner was an instigator of the fight, and orders that the prisoner be publicly
flogged to deter both this prisoner and others from engaging in future acts of violence. Is this
flogging a punishment?
Now imagine a slightly different scenario. A federal judge sentences a gang member to the
statutory maximum sentence of five years in prison and a $250,000 dollar fine for illegally growing
twenty-five marijuana plants. When the offender arrives at prison, the warden informs him that he
will add one day to the prisoner’s sentence for every day the prisoner fails to identify the other
gang members he knows in the prison. As a result, the offender spends an extra five years in prison.
Is this additional prison time a punishment?
Now imagine a third scenario. Imagine that instead of a flogging or extra prison time, the
warden transfers the prisoner to a higher security facility to minimize the risk of violence or gang
activity. As a result, the prisoner suffers greater restrictions on his liberty and experiences more
discomfort than he would in a lower security facility. Is this transfer a punishment?
Many of us would probably consider the first two scenarios to be clear examples of
punishment, but we might not be so sure about the third. In all three cases, a government official
inflicts additional pain or imposes additional restrictions on a prisoner’s liberty. In all three cases,


Professor of Law, University of Florida Levin College of Law. I thank the organizers of and participants in the
Northwestern University Law Review symposium on “Rethinking Solitary Confinement,” from which this essay
sprang. I also thank both Judith Resnik and the participants in a faculty workshop at the University of Florida for
helpful comments and suggestions. Finally, I thank the University of Florida for a generous research grant that made
this Essay possible.
1
See, e.g., Jerome Hall, Nulla Poena Sine Lege, 47 YALE L.J. 165 (1937) (tracing the historical roots of the
principle).
2
See ASCA & LIMAN CTR. FOR PUB. INTEREST L., REFORMING RESTRICTIVE HOUSING: THE 2018
ASCA-LIMAN NATIONWIDE SURVEY OF TIME-IN-CELL 4, 14 (2018) [hereinafter REFORMING
RESTRICTIVE HOUSING] (estimating that 60,000 prisoners are currently kept in solitary confinement, of whom
approximately 11,500 have spent a year or more in solitary confinement and nearly 3,000 have spent 6 years or more
in solitary confinement).
3
See Alexander A. Reinert, Solitary Troubles, 93 NOTRE DAME L. REV. 927, 959–60 (2018).

1

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
the official’s actions are motivated by a desire to enhance prison security. The third scenario seems,
on its face, less extreme than the other two—so we might think of it as a mere administrative
measure rather than an additional punishment.4 But before deciding, we might want to know more
about conditions at the new facility. If they are sufficiently harsh, the reassignment might also start
to look like an additional punishment.
These examples illustrate two things: First, the most important factor in determining the
line between punishments and non-punishments, at least intuitively, is penal effect. Flogging and
extension of a prison sentence are new punishments because they inflict pain or restrict liberty well
beyond what was authorized by the original sentence. The fact that the warden’s purpose might be
characterized as “regulatory” rather than “penal” is not enough to transform these punishments
into mere administrative acts. Second, although prison officials need discretion to protect guards
and inmates from prison violence, this discretion does not include the power to impose new
punishments beyond what was authorized by the offender’s sentence. 5 Otherwise, prison officials
would have the authority to do what legislatures, judges, and juries may not: Impose punishment
without law.6
The government’s authority to impose punishments is limited by several constitutional
provisions. The Ex Post Facto clauses of Article I, sections 9 and 10 prohibit the government from
increasing an offender’s punishment after he commits a crime. 7 The Fifth Amendment’s8 Double
Jeopardy Clause prohibits punishing an offender twice for the same offense, and its Compelled
Self-Incrimination Clause prohibits the use of punishment as coercion to obtain incriminating
information from an individual.9 The Sixth Amendment prohibits punishment for conduct that is

4

See Meachum v. Fano, 427 U.S. 215, 225 (1976) (“Confinement in any of the State's institutions is within the
normal limits or range of custody which the conviction has authorized the State to impose”).
5
See infra Part I.
6
See infra Part II.
7
U.S. CONST. art. I, §§ 9, 10.
8
U.S. CONST. amend. V.
9
The Supreme Court has sometimes implied that government coercion can only be considered a “punishment” if it
is inflicted as the result of a criminal conviction. See, e.g., Ingraham v. Wright, 430 U.S. 651, 664 (1977) (holding
that paddling of schoolchildren did not come within scope of Eighth Amendment because “[a]n examination of the
history of the Amendment and the decisions of this Court construing the proscription against cruel and unusual
punishment confirms that it was designed to protect those convicted of crimes”). Early constitutional history
suggests, however, that the term “punishment” was also thought to encompass practices such as pretrial torture,
where pain was inflicted in order to obtain information. For example, during the Virginia ratification debate, George
Mason, the principle drafter of the Virginia Declaration of Rights, asserted the following reasons for concluding that
that document prohibited pretrial torture: “[F]or that one clause expressly provided that no man can give evidence
against himself; and that the worthy gentleman must know that, in those countries where torture is used, evidence
was extorted from the criminal himself. Another clause of the bill of rights provided that no cruel and unusual
punishments shall be inflicted; therefore, torture was included in the prohibition.” 3 J ONATHAN ELLIOT, DEBATES IN
THE SEVERAL STATE CONVENTIONS, ON THE ADOPTION OF THE FEDERAL CONSTITUTION, AS RECOMMENDED BY THE
GENERAL CONVENTION AT PHILADELPHIA, IN 1787, at 452 (2d ed. 1836) [hereinafter 3 ELLIOT’S DEBATES].
Similarly, Blackstone described pretrial torture as a “punishment.” 4 W ILLIAM BLACKSTONE, COMMENTARIES ON
THE LAWS OF ENGLAND 325–27 (16th ed. 1825); see also Celia Rumann, Tortured History: Finding Our Way Back
to the Lost Origins of the Eighth Amendment, 31 PEPP. L. REV. 661, 673 (2004) (discussing historical evidence that
the Eighth Amendment was originally understood to prohibit torture).

2

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
not proven to a jury beyond a reasonable doubt. 10 And of course, the Eighth Amendment prohibits
cruel and unusual punishments.11 If flogging, extending a prison sentence, or transferring a
prisoner to a higher security facility constitutes punishment, then prison officials may not inflict it
unless it was both authorized by the penal statute governing the offense of conviction at the time
the prisoner committed the offense and imposed by the sentence actually given by a judge or jury.
Moreover, the Fifth and Eighth Amendments may bar certain punishments altogether, either
because they involve an effort to force the prisoner to provide incriminating evidence or because
they are cruel and unusual.12
These limits apply to all three branches of government. Indeed, they arguably should apply
with greatest force to the executive branch because it has strong incentives to order the infliction
of punishments, but no independent constitutional authority to do so. Basic separation of powers
principles dictate that only the legislature may authorize punishments and only the judge or the
jury may impose them. Executive officials are supposed to implement the punishments authorized
by the other branches of government. They do not have the authority to enhance punishments on
their own.13 At the same time, executive officials interact with individuals more frequently and
directly than representatives of the other branches of government, and have strong incentives to
use punishment to exert control over them. Thus it is important to make sure that executive officials
comply with the constitutional provisions summarized above.
The Supreme Court’s modern prison conditions jurisprudence shows little awareness of the
separation of powers principles prohibiting executive officials from imposing punishments on their
own authority.14 Instead, the Court has focused on a different separation of powers problem: The
need to prevent the judiciary from involving itself in the running of prisons. To decide
constitutional cases without intruding upon the prerogatives of the other branches of government,
courts need a judicially administrable standard of adjudication. As discussed in Part I below, the
Supreme Court’s modern punishment-related jurisprudence notably lacks such a standard. The
tests the Court employs for distinguishing punishments from non-punishments are so vague and
ambiguous that they provide little real guidance. Thus the Court must either substitute its own
judgment for that of prison officials or defer to prison officials’ constitutionally questionable
10

U.S. CONST. amend. VI. See also, e.g., Apprendi v. New Jersey, 530 U.S. 466, 490 (2000) (“Other than the fact of
a prior conviction, any fact that increases the penalty for a crime beyond the prescribed statutory maximum must be
submitted to a jury, and proved beyond a reasonable doubt.”).
11
U.S. CONST. amend. VIII.
12
Because prison conditions are part of the punishment the offender receives with his sentence, the Eighth
Amendment applies to the conditions even if the decision to place a prisoner in a certain facility does not constitute
a separate punishment above and beyond the prisoner’s formal sentence. See, e.g., Estelle v. Gamble, 429 U.S. 97,
104–06 (1976) (applying the Cruel and Unusual Punishments Clause to prison conditions); see also John F.
Stinneford, The Original Meaning of “Cruel”, 105 GEO. L.J. 441 (2017) [hereinafter Stinneford, Original Meaning
of “Cruel”] (discussing the Supreme Court’s prison conditions jurisprudence under the Eighth Amendment). Judith
Resnik has recently identified a nascent “anti-ruination principle” tying together the various strands of the Supreme
Court’s Eighth Amendment jurisprudence over the past seventy years, including its prison conditions case law. See
Judith Resnik, (Un)Constitutional Punishments: Eighth Amendment Silos, Penological Purposes, and People's
"Ruin", 129 YALE L.J. FORUM 365, 408 (2020).
13
See supra nn. 7-12 and accompanying text.
14
See infra Part I.

3

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
conduct. In other words, in the absence of a clear, workable standard of adjudication, the Court
must either violate separation of powers by taking over the functions of the executive or tolerate
the executive’s violation of separation of powers by deferring to its decisions to inflict punishments
on the basis of its own authority. Such toleration also permits prison officials to violate individuals
rights by imposing punishments without law.
In recent decades, the Supreme Court has chosen to tolerate punishment by executive fiat.
The Court has implied that once a prisoner is incarcerated, changes to prison conditions will not
be considered punishments unless they are cruel and unusual under the Eighth Amendment—a
situation that is only triggered, under current case law, by proof that the responsible prison official
had a culpable state of mind.15 Even if conditions of confinement impose a “significant and atypical
hardship” on the prisoner, they are not considered punishments and therefore trigger only minimal
due process protection.16 The Court’s deference to administrative discretion means that executive
officials can easily evade constitutional restrictions on the infliction of punishment.
An examination of English and American constitutional history demonstrates three facts
that are relevant to this situation: First, the need to limit the government’s discretion over
punishment has been a central theme of English and American constitutionalism from the Magna
Carta through the adoption of the American Bill of Rights. Second, executive officials’ exercise
of undue discretion over punishment has been recognized for centuries as a central characteristic
of arbitrary and tyrannical government. Third, the core standard for distinguishing between
punishments and non-punishments is penal effect viewed in light of tradition. If a government
action has traditionally been used as a punishment, or imposes pain or deprivation equivalent to a
method traditionally used as a punishment, it is a punishment for constitutional purposes regardless
of the label the government attaches to it.17
The Supreme Court’s older case law reflects these principles. In 1890, the Court held, in
the case of In re Medley,18 that transfer of a condemned offender from a county jail to solitary
15

See id.
See, e.g., Wilkinson v. Austin, 545 U.S. 209, 223 (2005). See also infra Section I.b. Moving the analysis from the
“punishment” question to the “due process” question has not solved the problem arising from the Court’s lack of an
administrable standard for differentiating punishments from non-punishments, for the Court also lacks an
administrable standard for determining which inflictions of pain or deprivations of liberty within prison trigger due
process protection. See, e.g., Wilkinson, 545 U.S. at 223 (noting the courts lack a “baseline from which to measure
what [deprivation] is atypical and significant in any particular prison system”).
17
This is not to say that the reasons for the imposition are unimportant. If the government imposes a deprivation that
has traditionally been used for non-penal as well as penal reasons, the purpose of the deprivation may tell us whether
or not the deprivation is punishment. For example, denial of the right to practice law has traditionally been imposed
for non-penal reasons – failure to pass the bar exam, failure to meet character and fitness requirements, etc. But if
the government imposes this deprivation in order to punish a person for prior conduct, it is a punishment for
constitutional purposes. See Ex parte Garland, 71 U.S. 333, 377 (1867). The government’s purpose is also important
in determining whether a punishment is justified and proportionate to the offense. See, e.g., Richard S. Frase,
Excessive Prison Sentences, Punishment Goals, and the Eighth Amendment: “Proportionality” Relative to What?,
89 MINN. L. REV. 571 (2005); Youngjae Lee, The Constitutional Right Against Excessive Punishment, 91 VA. L.
REV. 677 (2005); John F. Stinneford, Rethinking Proportionality Under the Cruel and Unusual Punishments Clause,
97 VA. L. REV. 899 (2011) [hereinafter Stinneford, Rethinking Proportionality].
18
134 U.S. 160 (1890).
16

4

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
confinement in a penitentiary prior to execution was a new punishment for constitutional purposes
for two related reasons: it was historically used as a heightened form of punishment and it inflicted
substantial suffering beyond what is normally imposed by a prison sentence. 19 The fact that the
government’s purpose in imposing solitary confinement on Medley was regulatory rather than
penal was irrelevant to the Court’s analysis. Medley is still good law, and answers the question
posed by this Essay. Solitary confinement is a punishment, not a mere exercise of administrative
discretion, and is thus subject to the constitutional constraints listed above.
Part I of the Essay that follows describes the Supreme Court’s existing case law governing
prison officials’ discretion to impose harsher conditions on inmates. Part II analyzes English and
American constitutional history relating to the need to limit discretion over punishment, the danger
of executive discretion in the infliction of punishment, and the distillation of a standard relevant
to conditions of confinement. Finally, Part III checks the accuracy of the Supreme Court’s
conclusion in Medley that the harshness of solitary confinement makes it a new punishment by
examining historical and modern empirical data relating to the effects of solitary confinement, and
concludes that the Medley court was correct.
I. THE DOMINANCE OF DISCRETION: CURRENT APPROACHES TO PUNISHMENT AND DUE PROCESS
Current Supreme Court doctrine governing conditions of confinement focuses almost
exclusively on the need to respect the discretionary decisions of prison officials and focuses little
on the need to constrain this discretion by law. This deferential approach stems from the Court’s
failure to identify a workable, coherent definition of “punishment,” which has led to an inability
to develop a standard to differentiate permissible from impermissible exercises of discretion over
conditions of confinement.
a. The Supreme Court’s Failure to Define “Punishment”
The Supreme Court has not adopted a clear, consistent standard for determining whether
conditions imposed by prison officials constitute an additional punishment beyond what has been
authorized by penal statute and judicial sentence. Instead it has adopted at least two different
standards, one of which is inconsistent with the plain meaning of the constitution’s text, and the
other of which is largely incoherent.
The Court’s first definition of punishment, which it employs in Eighth Amendment cases,
comes from its decision in Wilson v. Seiter.20 In that case, a prisoner brought a lawsuit claiming
that certain prison conditions—namely “overcrowding, excessive noise, insufficient locker storage
space, inadequate heating and cooling, improper ventilation, unclean and inadequate restrooms,
unsanitary dining facilities and food preparation, and housing with mentally and physically ill

19
20

Id. at 173.
501 U.S. 294 (1991).

5

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
inmates”21—constituted cruel and unusual punishment. The Supreme Court held that such
conditions amount to a punishment only if the prison official responsible for them displayed a
“wanton” or “culpable” state of mind.22 Such wantonness or culpability could be shown by
establishing a prison official’s “deliberate indifference” 23 to a prisoner’s well-being, or that the
official acted “maliciously and sadistically for the very purpose of causing harm.” 24
The requirement that responsible prison officials must be shown to have a culpable state of
mind might make sense as a standard for determining which punishments are “cruel,” 25 but not as
a means of distinguishing punishments from non-punishments. Indeed, this standard seems flatly
inconsistent with the plain meaning of the word “punishment.” Webster’s 1828 Dictionary, for
example, defines “punishment” as “[a]ny pain or suffering inflicted on a person for a crime or
offense, by the authority to which the offender is subject, either by the constitution of God or of
civil society.”26 Similarly, Merriam-Webster’s online dictionary defines punishment, in relevant
part, as “suffering, pain, or loss that serves as retribution . . . a penalty inflicted on an offender
through judicial procedure; severe, rough, or disastrous treatment.” 27 As these older and newer
dictionary definitions imply, punishment involves intent to inflict pain or suffering, but not
necessarily culpable intent. Moreover, once we consider punishment outside the context of the
Cruel and Unusual Punishments Clause, the culpability requirement loses even its surface appeal.
What does a government official’s culpability have to do with the question of whether a prisoner
is suffering an ex post facto punishment, or double jeopardy, or deprivation of the right to a jury
trial?
The other test the Court sometimes uses to distinguish punishments from non-punishments
was first articulated in Kennedy v. Mendoza-Martinez,28 and was applied to prison conditions for
pretrial detainees in Bell v. Wolfish29 and United States v. Salerno.30 The question in MendozaMartinez was whether a statute that stripped citizenship from certain draft dodgers imposed a
punishment within the meaning of the Constitution.31 To answer this question, the Court set forth
the following multi-factor test: (1) whether the sanction imposes “an affirmative disability or

21

Id. at 296.
Id. at 299, 302, 305 (italics omitted).
23
Id. at 302 (quoting Estelle v. Gamble, 429 U.S. 97, 106 (1976)).
24
Id. (quoting Whitley v. Albers, 475 U.S. 312, 320–1 (1986)).
25
As I have demonstrated elsewhere, however, this would be an incorrect standard as a matter of original meaning.
See Stinneford, Original Meaning of “Cruel”, supra note 11. See also Samuel L. Bray, "Necessary and Proper" and
"Cruel and Unusual": Hendiadys in the Constitution, 102 VA. L. REV. 687, 717 (2016) (arguing that “cruel and
unusual” is an hendiadys—that is, a multi-word phrase with a single, complex meaning—but agreeing that it does
not refer to the mental state of the punisher).
26
Noah Webster, AN AMERICAN DICTIONARY OF THE ENGLISH LANGUAGE (New York, S. Converse 1828).
27
Punishment, MERRIAM-WEBSTER, https://www.merriamwebster.com/dictionary/punishment?utm_campaign=sd&utm_medium=serp&utm_source=jsonld
[https://perma.cc/DKQ9-579G].
28
372 U.S. 144 (1963).
29
441 U.S. 520 (1979).
30
481 U.S. 739 (1987).
31
372 U.S. at 165–166.
22

6

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
restraint;”32 (2) whether the sanction “has historically been regarded as a punishment;” 33 (3)
whether the sanction “comes into play only on a finding of scienter;” 34 (4) whether the sanction
“promote[s] the traditional aims of punishment—retribution and deterrence;” 35 (5) whether the
sanction applies to behavior that is “already a crime;”36 (6) whether the sanction is “rationally
connected” to a possible “alternative purpose;” 37 and (7) whether the sanction “appears excessive
in relation to the alternative purpose.” 38
This test is confusing and amorphous. The Court does not specify how the factors relate to
each other, nor how they are supposed to reveal whether a government action is a punishment.
Nonetheless, there is an implicit logical relationship between the test’s seven factors. The first two
focus on the government action itself: Does it have a penal effect, and is it the sort of action that
has historically been used as punishment? The remaining factors focus on the government’s
purpose in taking the action: Is it directed only at culpable actors? (Factors 3 and 5.) Does the
government label the action as penal or regulatory? (Factors 4 and 6.) If the government labels the
action as regulatory, is this labeling plausible in light of the action’s effect? (Factors 6 and 7.)
When one examines the precedents underlying the Mendoza-Martinez factors, it becomes
clear that a government-imposed sanction is a punishment if it has a clear penal effect, and that
questions about the government’s purpose in imposing a sanction only arise when the sanction is
also often used for non-penal purposes. Every case cited in support of the factors relating to
government purpose (factors 3 through 7) involved a type of deprivation that has historically been
imposed for regulatory purposes. Seven of the nine cases involved monetary deprivations labeled
as taxes, customs duties, denial of social security benefits, or contractual liquidated damages
provisions.39 The two remaining cases involved “status” deprivations—disqualification from
public office40 and denationalization41—that are often imposed for regulatory purposes. States
routinely impose regulatory requirements that must be met before a person can hold public office,
and the exclusion of those who fail to meet these requirements is not generally regarded as a
punishment. Similarly, as a plurality of the Supreme Court noted in Trop, the federal government
generally denationalizes nationalized citizens who falsified their citizenship applications “not . . .

32

Id. at 168.
Id.
34
Id.
35
Id.
36
Id.
37
Id. at 168–169.
38
Id. at 169.
39
See United States v. Constantine, 296 U.S. 287 (1935) (excise tax on liquor); United States v. La Franca, 282 U.S.
568 (1931) (tax on retail liquor business); Bailey v. Drexel Furniture Co. (Child Labor Tax Case), 259 U.S. 20
(1922) (child labor tax); Lipke v. Lederer, 259 U.S. 557 (1922) (“tax” on unlawful liquor sales); Helwig v. United
States, 188 U.S. 605 (1903) (customs charge); Flemming v. Nestor, 363 U.S. 603 (1960) (termination of social
security benefits); Rex Trailer Co. v. United States, 350 U.S. 148 (1956) (contractual liquidated damages provision).
40
Cummings v. Missouri, 71 U.S. 277 (1867).
41
Trop v. Dulles, 356 U.S. 86 (1958).
33

7

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
to penalize the alien for having falsified his application for citizenship” but “in the exercise of the
power to make rules for the naturalization of aliens.” 42
It is not a coincidence that the precedents supporting factors 3 through 7 involve
deprivations that have historically been imposed for regulatory purposes. In order to determine
whether such a deprivation is actually a punishment, the Court must consider evidence that the
deprivation’s true purpose is not what it appears to be. Factors 3 through 7 identify the kinds of
evidence relevant to this inquiry. But if the deprivation is of the sort that has historically been
imposed as punishment, there is no need to look for evidence of purpose. The penal effect itself is
sufficient to qualify the deprivation as punishment.43 Thus a court examining whether a
government-imposed deprivation is actually a punishment should engage in a two-step inquiry.
First it should ask whether the action has traditionally been used as a punishment, or imposes pain
or deprivation equivalent to a method traditionally used as a punishment. If the answer to this
question is no, the court should use factors 3 through 7 to determine whether the action has been
imposed for punitive purposes. If the answer is yes, there is no need to inquire into the
government’s purpose.
The Supreme Court in recent decades has generally skipped the first step of this inquiry
and focused solely on evidence relating to a given sanction’s purpose. The Court has also
increasingly deferred to the label the government attaches to the sanction at the expense of other
types of evidence. For example, in Bell v. Wolfish,44 the Court considered whether prison
conditions such as double bunking and restrictions on the receipt of books, food, and other
packages counted as unconstitutional punishment of pretrial detainees. In answering this question,
the Court focused almost exclusively on factor 6—whether there was a rational relationship
between the challenged restrictions on liberty and the asserted non-punitive purpose of the
restrictions.45 The Court minimized the liberty interest at stake and emphasized the need to defer
to the expertise of prison administrators. Because detainees were already in prison, and were thus
already subject to intrusions on their privacy and restrictions on their liberty, these increases in the
level of intrusion were not sufficient to raise constitutional concerns. 46 Moreover, respect for
separation of powers dictated that the Court give deference to prison officials’ expertise concerning
the measures necessary to preserve security. 47 The Supreme Court observed that courts should not
be in the business of running prisons. 48 Thus, it concluded, the restrictions placed on pretrial

42

Id. at 98.
See, e.g., In re Medley, 134 U.S. 160 (1890) (imposition of solitary confinement); Wong Wing v. United States,
163 U.S. 228 (1896) (60 days imprisonment at hard labor). These cases are discussed in greater depth in Section
II.C, infra.
44
Bell v. Wolfish, 441 U.S. 520 (1979).
45
Id. at 538–539.
46
Id. at 546 (“A detainee simply does not possess the full range of freedoms of an unincarcerated individual.”).
47
Id. at 547 (“Prison administrators . . . should be accorded wide-ranging deference in the adoption and execution of
policies and practices that in their judgment are needed to preserve internal order and discipline and to maintain
institutional security.”).
48
See id. at 548.
43

8

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
detainees did not constitute punishment because they were rationally related to the goal of
protecting prison security.49
In United States v. Salerno,50 the Supreme Court made its deference to government labeling
more explicit. It asserted that unless the government labels its action as a punishment, the action
will be considered a non-penal regulation so long as it has a rational relationship to a non-penal
purpose and is not excessive in relation to that purpose. 51 Thus, it held that pretrial detention on
grounds of dangerousness was not a punishment because Congress labeled this detention as nonpenal and because such detention was rationally related to the goal of protecting the community. 52
The Supreme Court later used similar reasoning to hold that indefinite detention of persons
considered sexually dangerous was not a punishment and thus not subject to the various limits
stated in the Constitution.53
b. The Minimal Due Process Approach
As the cases discussed above demonstrate, outside the context of an Eighth Amendment
claim, the Supreme Court strongly defers to the labeling the government attaches to a given
deprivation or infliction. If the government labels a sanction as non-penal, the Court generally
accepts the label. The cases discussed below demonstrate the Supreme Court will sometimes
analyze these “non-punishments” under the due process clause—but only in extreme cases—and
even then, it provides far less procedural and substantive protection than is required by the
constitutional provisions that govern the infliction of punishment.
The Supreme Court’s reluctance to interfere with prison officials’ decisions concerning
conditions of confinement is evident in three cases involving imposition of solitary confinement:
Hutto v. Finney,54 Sandin v. Conner,55 and Wilkinson v. Austin.56
In Hutto v. Finney,57 the Supreme Court upheld a district court’s determination that the
Arkansas prison system’s use of punitive isolation was a cruel and unusual punishment. 58 The
conditions of isolation were particularly egregious—numerous prisoners were crowded into a
single cell and subjected to malnourishment and exposure to infectious disease. 59 But even as the
Court condemned these conditions, it implied that prison officials have implicit authority to impose
long-term solitary confinement so long as they provide for the physical needs of the offender: “If
new conditions of confinement are not materially different from those affecting other prisoners, a
49

See id. at 555.
United States v. Salerno, 481 U.S. 739 (1987).
51
See id. at 747.
52
Id.
53
See Kansas v. Hendricks, 521 U.S. 346, 366 (1997).
54
437 U.S. 678 (1978).
55
515 U.S. 472 (1995).
56
545 U.S. 209 (2005).
57
437 U.S. 678 (1978).
58
Id. at 680–81.
59
Id. at 682–83 (citations omitted).
50

9

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
transfer [to punitive isolation] for the duration of a prisoner’s sentence might be completely
unobjectionable and well within the authority of the prison administrator.” 60
The Supreme Court confirmed prison officials’ discretion to send prisoners to solitary
confinement in Sandin v. Conner.61 In that case, an inmate named Conner cursed at a prison guard
during a highly intrusive strip search.62 As a result, the prison conducted a disciplinary hearing
pursuant to prison regulations, but refused to allow Conner to present witnesses. 63 Conner was
sentenced to thirty days of “disciplinary segregation” in solitary confinement. 64 During this period,
Conner “had to spend his entire time alone in his cell (with the exception of 50 minutes each day
on average for brief exercise and shower periods, during which he nonetheless remained isolated
from other inmates and was constrained by leg irons and waist chains).” 65 By contrast, if he had
not been placed in disciplinary segregation, Conner “would have left his cell and worked, taken
classes, or mingled with others for eight hours each day.”66 Thus Conner argued that the denial of
his request to present witnesses at his disciplinary hearing violated his right to due process. 67
The Supreme Court acknowledged that the government’s purpose in sending Conner to
solitary confinement was “punitive”68 but held that the due process clause was not implicated
because such confinement did not implicate any “liberty interest.” 69 A liberty interest might be
implicated if prison officials impose a constraint that “exceed[s] the sentence in such an
unexpected manner as to give rise to protection by the Due Process Clause of its own force”—for
example, transfer to a mental hospital or coercive administration of a psychotropic drug. 70 Such
interests might also be implicated if the state “create[s]” liberty interests by giving prisoners certain
rights—such as the right to “good time” credits—and then takes them away from a particular
prisoner without adequate procedures. 71 But the latter situation only gives rise to due process
concerns if the deprivation “imposes atypical and significant hardship on the inmate in relation to
the ordinary incidents of prison life.”72 The Supreme Court held that the due process clause did
not apply of its own force in this case because “[d]iscipline by prison officials in response to a
wide range of misconduct falls within the expected parameters of the sentence imposed by a court
of law.”73 Moreover, the Court held, Conner’s sentence to solitary confinement “did not present
the type of atypical, significant deprivation in which a State might conceivably create a liberty
60

Id. at 686.
515 U.S. 472 (1995).
62
Id. at 475.
63
Id.
64
Id. at 475–476.
65
Id. at 494 (Breyer, J., dissenting).
66
Id.
67
Id. at 476.
68
Id. at 485.
69
Id. at 486.
70
See id. at 484 (citing Vitek v. Jones, 445 U.S. 480, 493 (1980) (transfer to mental hospital); Washington v.
Harper, 494 U.S. 210, 221–22 (1990) (coercive administration of a psychotropic drug)).
71
Id. at 477–478 (citing Wolff v. McDonnell, 418 U.S. 539, 557 (1974)).
72
Id. at 484.
73
Id. at 485.
61

10

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
interest.”74 This was so because the conditions of disciplinary segregation “mirrored” those of
“administrative segregation and protective custody” and because inmates in the general population
were confined to their cells for twelve to sixteen hours per day. 75
The Supreme Court returned to this question ten years later in Wilkinson v. Austin.76 That
case concerned the procedures necessary to assign a prisoner to indefinite solitary confinement in
a “supermax” facility known as the Ohio State Penitentiary (OSP). 77 As in Conner, prisoners were
given notice and an opportunity to rebut the case against them before being assigned to solitary
confinement but were not allowed to present their own witnesses. 78 Thus, they claimed that the
procedure violated their right to due process. 79 The Court decided that the differences between the
indefinite solitary confinement at issue in this case and the disciplinary segregation at issue in
Conner were sufficient to create “an atypical and significant hardship,” thus implicating a liberty
interest under the due process clause:
For an inmate placed in OSP, almost all human contact is prohibited, even to the
point that conversation is not permitted from cell to cell; the light, though it may be
dimmed, is on for 24 hours; exercise is for 1 hour per day, but only in a small indoor
room. Save perhaps for the especially severe limitations on all human contact, these
conditions likely would apply to most solitary confinement facilities, but here there
are two added components. First is the duration. Unlike the 30-day placement in
[Conner], placement at OSP is indefinite and, after an initial 30-day review, is
reviewed just annually. Second is that placement disqualifies an otherwise eligible
inmate for parole consideration.80
Nonetheless, as in Conner, the Austin court upheld the decision to send the prisoner to indefinite
solitary confinement.81 The Court held that prisoners had a reduced liberty interest because they
were already incarcerated82 and that the state’s interest in security was “dominant” because of the
threat posed by prison gangs. 83
The Supreme Court’s decisions in Hutto, Conner, and Austin show that the Court lacks a
meaningful standard for determining whether the imposition of solitary confinement constitutes
an additional punishment or even a mere deprivation of liberty. The Hutto and Conner decisions
imply that every sentence of incarceration includes authorization to subject prisoners to solitary
confinement. The Austin Court expresses some discomfort with this idea but ultimately affirms
74

Id. at 486.
Id. at 486 n.8.
76
545 U.S. 209 (2005).
77
Id. at 213.
78
Id. at 216.
79
Id. at 218.
80
Id. at 223–224.
81
Id. at 213.
82
Id. at 225.
83
Id. at 227 (“Prison security, imperiled by the brutal reality of prison gangs, provides the backdrop of the State's
interest. Clandestine, organized, fueled by race-based hostility, and committed to fear and violence as a means of
disciplining their own members and their rivals, gangs seek nothing less than to control prison life and to extend
their power outside prison walls.”).
75

11

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
prison officials’ broad authority to impose even indefinite solitary confinement on prisoners. These
decisions show relatively little awareness of the extreme degree of additional suffering that solitary
confinement inflicts on prisoners. This suffering will be discussed in Section III below.
II. HISTORICAL LIMITS ON PUNISHMENT DISCRETION
This Part provides a brief overview of three aspects of English and American constitutional
history: First, the development of procedural and substantive doctrines to limit the government’s
discretion over the imposition of punishment and thus minimize the risk of punishment without
law. Second, recognition of the dangers to liberty that arise from executive discretion over
punishment. Third, the distillation in the nineteenth century of a standard for differentiating
punishments from non-punishments, focusing on penal effect in light of tradition.
a. Constitutional Doctrines Limiting Punishment Discretion
Of all governmental powers, the power to punish may be the most susceptible to abuse. If
one wishes to enhance one’s power, eliminate one’s enemies, or simply demonstrate that one is in
control, the easiest and most straightforward way to accomplish these goals is often the whip, the
prison, or the gallows. For this reason, much of English constitutional history may be seen as a
struggle between a power holder (whether it be king, judge, or parliament) seeking to exercise
unconstrained power to punish and others seeking to enforce common law limits on this power.
These discretion-constraining common law doctrines were later adopted by the drafters of the
United States Constitution and written into its original text or the Bill of Rights. Because I have
limited space, I will provide only a brief overview of these principles and doctrines.
English common law doctrines enforcing the nulla poena principle were both substantive
and procedural in nature. These included:
No punishment for conduct not prohibited by law at the time it occurred
In England, the nulla poena principle dates back at least to the Magna Carta. The thirteenth
century conflicts between King John and his barons led to a settlement in which he agreed to abide
by customary limits on royal power, including the famous promise: “No Free-man shall be taken,
or imprisoned, or dispossessed, of his free tenement, or liberties, or free customs, or be outlawed,
or exiled, or in any way destroyed; nor will we condemn him, nor will we commit him to prison,
excepting by the legal judgment of his peers, or by the laws of the land.” 84 This promise reflected

84

The Third Great Charter of King Henry the Third c. 29 (1225), in THE ROOTS OF LIBERTY: MAGNA CARTA,
ANCIENT CONSTITUTION, AND THE ANGLO-AMERICAN TRADITION OF RULE OF LAW, APPENDIX: TEXT AND
TRANSLATION OF MAGNA CARTA 347 (Ellis Sandoz ed., Liberty Fund, Inc. 2008). The full text of the provision in
the original Latin is: “nullus liber homo decetero capiatur vel imprisonetur aut disseisiatur de aliquo libero
tenemento suo vel libertatibus vel liberis consuetudinibus suis, aut utlagetur, aut exuletur aut aliquo alio modo
destruatur, nec super eum ibimus, nec super eum mittemus, nisi per legale judicium parium suorum, vel per legem
terre.” See id. at 340; Claire Breay & Julian Harrison, Magna Carta in Context, BRITISH LIBRARY,

12

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
the idea that the king’s power to punish was limited by law. By the seventeenth century, common
law thinkers like Edward Coke identified this passage from Magna Carta as the source of the
requirement that life, liberty, or property could not be taken without due process. 85
The idea that the government may not inflict punishment for conduct that does not violate
preexisting law is reflected in the traditional common law prohibition of ex post facto
punishments.86 English rulers did not always honor this principle, of course, but when they violated
it they were ultimately condemned as acting unconstitutionally. For example, English monarchs
created prerogative courts, such as the Court of Star Chamber, in part to evade procedural and
substantive limits to government power generally respected by common law courts. 87The Court of
Star Chamber’s penchant for punishing those who had not violated preexisting law led parliament
not only to abolish it, but to condemn it on the ground that it had “undertaken to punish where no
law doth warrant, and to make decrees for things having no such authority, and to inflict heavier
punishments than by any law is warranted . . . . [Such judgment had proven] to be an intolerable
burden to the subjects, and the means to introduce an arbitrary power and government.” 88
Notice that in this statute, parliament criticizes the Court of Star Chamber not only for
violating substantive rights, but also for ignoring structural limits to its own power. The Court of
Star Chamber abused its power by inflicting punishments either unauthorized by law or heavier
than authorized by law, and also by issuing decrees it had no authority to issue. This statute was
later interpreted as condemning the Court’s refusal to follow established common law procedures
designed to protect the rights of defendants, and as requiring that any new courts of justice
established by the king “must proceed according to the old established forms of the common
law.”89 By insisting that the government can only inflict punishments in a manner that protects
substantive rights, follows established procedures, and respects structural limitations of

https://www.bl.uk/magna-carta/articles/magna-carta-in-context [https://perma.cc/7VHA-C6XJ] (describing
historical and legal issues that gave rise to the Magna Carta).
85
See EDWARD COKE, SECOND PART OF THE INSTITUTES OF THE LAWS OF ENGLAND 50 (6th ed. 1671).
86
The prohibition of ex post facto punishments is reflected in a story recounted by Blackstone concerning the
Russian ambassador to England during the reign of Queen Anne. The ambassador was apparently a profligate
spender who ran up debts he could not pay back. Ultimately, one of his creditors had him arrested and imprisoned
for debt. When the Czar learned of this, he angrily demanded that the creditor be executed for his effrontery. “But
the queen (to the amazement of that despotic court) directed her secretary to inform him, ‘that she could inflict no
punishment upon any, the meanest, of her subjects, unless warranted by the law of the land: and, therefore, was
persuaded that he would not insist upon impossibilities.” See 1 WILLIAM BLACKSTONE & ST. GEORGE TUCKER,
BLACKSTONE'S COMMENTARIES: WITH NOTES OF REFERENCE, TO THE CONSTITUTION AND LAWS, OF THE FEDERAL
GOVERNMENT OF THE UNITED STATES; AND OF THE COMMONWEALTH OF VIRGINIA 254–55 (1803) [hereinafter
1 TUCKER’S BLACKSTONE].
87
See, e.g., Frederic William Maitland, Selected Historical Essays of F.W. Maitland 127-130 (Helen M. Cam ed.,
1957).
88
The Act for the Abolition of the Court of Star Chamber, (July 5, 1641) Statutes of the Realm, v. 110. 17 Car. I.
cap. 10.
89
1 TUCKER’S BLACKSTONE, supra note 84, at 142.

13

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
government power, parliament sought to prevent future efforts “to introduce an arbitrary power
and government . . . .”90
No use of punishments that are harsher than the common law permits
As noted above, one of the grounds for condemning the Court of Star Chamber was that it
inflicted punishments that were heavier than the law authorized by law. Even if the law permits
punishment for a given offense, it also limits the punishments that may be inflicted for that offense.
To the extent punishment exceeds legal limits, it is a punishment without law. This limitation on
government power showed itself in two primary contexts under the English Constitution: The
absolute prohibition of torture and the requirement that punishment be proportionate to the offense.
One of the ways prerogative courts differed from common law courts was that at least some
of them used continental civil law practices such as torture to extract confessions from criminal
defendants.91 From the time English monarchs first introduced this practice, common lawyers
argued that it was illegal.92 Ultimately, the Court of King’s Bench declared in Felton’s Case (1628)
that torture was prohibited because “no such punishment is known or allowed by our law.” 93
The requirement that punishment be proportionate to the offense may be seen in
parliament’s condemnation of the Court of Star Chamber for inflicting heavier punishments than
the law permits for a given offense. It may be seen even more clearly in the prohibition of excessive
fines and cruel and unusual punishments in the English Bill of Rights. As I have shown elsewhere,
the prohibition of cruel and unusual punishments was directed not only at inherently cruel methods
of punishment such as torture, but also at punishments that are cruelly disproportionate to the
offense in light of longstanding prior practice. 94
The proportionality requirement may also be seen in rules governing conditions of
confinement while prisoners await trial. Blackstone wrote that because pretrial prisoners were held

90

The Act for the Abolition of the Court of Star Chamber, (July 5, 1641) Statutes of the Realm, v. 110. 17 Car. I.
cap. 10.
91
See, e.g., See, e.g., Anthony F. Granucci, “Nor Cruel and Unusual Punishments Inflicted”: The Original Meaning,
57 Cal. L. Rev. 839, 848-49 (1969) (describing the Court of High Commission’s use of torture); 3 Edward Coke,
Institutes of the Lawes of England ch. 2 (1608), as reprinted in 2 The Selected Writings and Speeches of Sir Edward
Coke 1025 (Steve Sheppard ed., 2003) (claiming that the first step toward introducing civil law judicial process into
England was the placement of a torture device – the rack – in the Tower of London).
92
See, e.g., SIR JOHN FORTESCUE, DE LAUDIBUS LEGUM ANGLIAE, A TREATISE IN COMMENDATION OF THE LAWS OF
ENGLAND 73 (Francis Gregor trans., Cincinnati, Robert Clarke & Co. 1874) (1470) (condemning torture: “[a]
practice so inhuman deserves not indeed to be called a law, but the high road to hell”); E DWARD COKE, THE THIRD
PART OF THE INSTITUTES OF THE LAWS OF ENGLAND: CONCERNING HIGH TREASON, AND OTHER PLEAS OF THE
CROWN AND CRIMINAL CAUSES 35 (1797) (“[T]here is no law to warrant tortures in this land . . . And there is no one
opinion in our books, or judicial record (that we have seen and remember) for the maintenance of tortures or
torments . . . .”).
93
3 T. B. HOWELL, Proceedings against John Felton for the Murder of the Duke of Buckingham, in HOWELL'S
STATE TRIALS 369 (1628).
94
Stinneford, Rethinking Proportionality, supra note 16, at 932–38.

14

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
“only for safe custody, and not for punishment,”95 they “ought to be used with the utmost
humanity; and neither be loaded with needless fetters, or subjected to other hardships than such as
are absolutely requisite for the purpose of confinement only.” 96 He lamented, however, that
conditions of confinement “must too often be left to the discretion of the gaolers; who are
frequently a merciless race of men, and, by being conversant in scenes of misery, steeled against
any tender sensation.”97 To protect against abuse of this discretion, parliament enacted laws
making gaolers liable to punishment for engaging in extortionate or abusive conduct, 98 and for
neglecting the health and sanitation of the prisoners. 99 In short, if a restriction on liberty is not
itself a punishment, the jailer must be careful not to inflict greater restrictions or pain than are
necessary, for any additional pain or restriction would be a punishment without law.
Protection of Common Law Procedural Rights
Common law criminal procedures were also designed to prevent the imposition of
wrongful punishment. These procedures included the right to indictment by grand jury, to jury trial
in the vicinage of the offense, to confront witnesses, and to seek a writ of habeas corpus. Such
procedures limited the power of the government to punish, either by requiring a finding by citizens
that punishment was warranted, or by permitting the defendant to challenge the basis of the
government’s case, or by providing that a judge may review the lawfulness of a person’s
incarceration.
The nulla poena principle was also reflected in jurisdictional rules designed to limit judicial
discretion at sentencing. For example, as Blackstone explains, courts of equity were prohibited
from exercising jurisdiction in criminal cases because of the risk they would use their equitable
powers to impose more punishment than was permissible by law: “For the freedom of our
constitution will not permit, that in criminal cases a power should be lodged in any judge, to
construe the law otherwise than according to the letter. . . . A man cannot suffer more punishment
than the law assigns, but he may suffer less. The laws cannot be strained by partiality to inflict a
penalty beyond what the letter will warrant; but, in cases where the letter induces any apparent
hardship, the crown has the power to pardon.”100
Strict Separation of Powers
The common law also required a strict separation of powers in the imposition and execution
of punishments. According to Blackstone, for example, only a judge could order the execution of
a man found guilty of murder, and he could only do so when acting upon the basis of a lawful

95

4 WILLIAM BLACKSTONE & ST. GEORGE TUCKER, BLACKSTONE’S COMMENTARIES: WITH NOTES OF REFERENCE,
TO THE CONSTITUTION AND LAWS, OF THE FEDERAL GOVERNMENT OF THE UNITED STATES; AND OF THE
COMMONWEALTH OF VIRGINIA 299 (1803) [hereinafter 4 T UCKER’S BLACKSTONE].
96

Id. at 300.
Id.
98
32 Geo. 2 c. 28, § XI (1758).
99
14 Geo. 3 c. 59 (1774).
100
1 TUCKER’S BLACKSTONE, supra note 84, at 92.
97

15

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
commission.101 The execution order could only be carried out by the proper officer or his deputy. 102
Finally—and most importantly for our purposes—the officer had no discretion to change the
ordered method of execution (for example, from beheading to hanging, or vice versa): “[F]or he is
merely ministerial, and therefore only justified when he acts under the authority and compulsion
of the law . . . .”103 If anyone violated any of these rules, he would himself be guilty of murder. 104
Even the king lacked the authority to substitute one method of execution for another, because this
would be imposition of a new punishment—although he could remit part of the sentence. 105
American Adoption of English Common Law Limits on Punishment Discretion
Americans of the Founding Era were at least as concerned about constraining governmental
punishment discretion as were English common law thinkers. 106 They were acutely aware of the
historical struggles to constrain this discretion, and were determined not to permit the same abuses
that had occurred in England.107 For example, when England tried to give an Admiralty Court
criminal jurisdiction over American colonists, Americans protested that the Admiralty Court used
the same civil law procedures as the Court of Star Chamber. As John Adams wrote: “Can you
recollect the complaints and clamors, which were sounded with such industry, and supported by
such a profusion of learning in law and history, and such invincible reasoning . . . against the StarChamber and High Commission, and yet remain an advocate for the newly-formed courts of
admiralty in America?”108
Similarly, Anti-Federalists opposed ratification of the United States Constitution on the
ground that it did not require Congress to provide traditional common law protections to criminal
defendants.109 Patrick Henry noted, for example, that although criminal courts of equity were
forbidden in England, “[Congress] will tell you that there is such a necessity of strengthening the

101

See id. at 178–79.
See id.
103
Id. at 179.
104
See id. at 178–179.
105
See id. at 179.
106
For a detailed discussion of American efforts to constrain governmental power within common law limits during
the revolutionary and founding periods, see John F. Stinneford, The Original Meaning of "Unusual": The Eighth
Amendment As A Bar to Cruel Innovation, 102 NW. U.L. REV. 1739, 1792–1810 (2008) [hereinafter Stinneford, The
Original Meaning of “Unusual”].
107
See id. at 1798.
108
John Adams, The Earl of Clarendon to William Pym, BOSTON GAZETTE (Supplement), Jan. 13, 1766, reprinted
in 3 CHARLES FRANCIS ADAMS, THE WORKS OF JOHN ADAMS, SECOND PRESIDENT OF THE UNITED STATES: WITH A
LIFE OF THE AUTHOR, NOTES AND ILLUSTRATIONS 469–70 (Charles C. Little, & James Brown 1851).
109
See, e.g., George Mason, Objections to this Constitution of Government, Sept. 15, 1787, reprinted in 2 THE
RECORDS OF THE FEDERAL CONVENTION OF 1787, at 637 (Max Farrand ed., 1911) (“There is no Declaration of
Rights, and the laws of the general government being paramount to the laws and constitution of the several States,
the Declaration of Rights in the separate States are no security. Nor are the people secured even in the enjoyment of
the benefit of the common law (which stands here upon no other foundation than its having been adopted by the
respective acts forming the constitutions of the several States).”)
102

16

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
arm of government, that they must have a criminal equity, and extort confession by torture, in
order to punish with still more relentless severity. We are then lost and undone.” 110
Indeed, many of the rights listed in the original constitutional text and the Bill of Rights
were specifically designed to limit the government’s discretionary power to punish. 111 These
include the right to trial by jury in the state and vicinage of the offense; 112 habeas corpus;113
prohibitions of ex post facto laws and bills of attainder; 114 the right to indictment by grand jury;115
the prohibition of double jeopardy; 116 the prohibition of compelled self-incrimination;117 the right
to due process of law;118 to be informed of the nature and cause of the accusation; 119 to be
confronted with the witnesses against him; 120 to have compulsory process for obtaining witnesses
in his favor;121 to have the assistance of counsel; 122 and the prohibition of excessive bail, excessive
fines, and cruel and unusual punishments.123
Like its English counterpart, American law required that even where a decision-maker has
discretion as to the type or quantity of punishment, such discretion must have legal limits. 124 For
example, the defendant in Commonwealth v. Wyatt125 was convicted of operating an illegal gaming
table126 and appealed his conviction on the ground that the criminal statute authorized the judge to
inflict a cruel and unusual punishment.127 The statute provided that a defendant could be
imprisoned for up to six months and “shall moreover be punished with stripes, at the discretion of
the Court, to be inflicted at one time, or at different times during such confinement, as such Court
may direct, provided the same do not exceed thirty-nine at any one time.” 128 Wyatt argued that
“the Court, by virtue of this Law, might exercise its discretion to subserve vindictive passions, and
so as to direct the party convicted to be subjected to thirty-nine stripes every day of the six months,
which would inevitably terminate in death; a death produced by the most cruel torture.”129 The
110

Patrick Henry, Speech to the Virginia Ratifying Convention for the United States Constitution (June 9, 1788), in
3 ELLIOT’S DEBATES, supra note 8 at 448.
111
The Bill of Rights was added to the Constitution largely to answer the Antifederalist critique that the original
constitutional text did not provide sufficient common law constraints on the power of the federal government. See,
e.g., Robert Allen Rutland, The Birth of the Bill of Rights 171-175 (1991).Stinneford, The Original Meaning of
“Unusual”, supra note 102, at 1800-1808.
112
U.S. CONST. art. III, § 2; U.S. CONST. amend. VI.
113
U.S. CONST. art. I, § 9.
114
U.S. CONST. art. I, §§ 9–10.
115
U.S. CONST. amend. V.
116
Id.
117
Id.
118
Id.
119
U.S. CONST. amend. VI.
120
Id.
121
Id.
122
Id.
123
U.S. CONST. amend. VIII.
124
See Stinneford, The Original Meaning of “Unusual”, supra note 102, at 1810–11.
125
27 Va. 694, 701 (Va. Gen. Ct. 1828).
126
Id. at 695.
127
Id. at 700.
128
Id. at 698.
129
Id. at 700.

17

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
General Court of Virginia rejected this argument, noting that the discretion authorized under this
statute was “of the same character with the discretion always exercised by Common Law Courts
to inflict fine and imprisonment, and subject to be restrained by the same considerations . . . .” 130
If the judge abused this discretion by inflicting excessive floggings, he could be “impeached” for
abuse of office.131
b. The Particular Danger of Executive Discretion over Punishment
Executive officials have strong incentives to impose punishments without law. A king or a
president may wish to use punishment to put a stop to civil disorder or to dispose of political
enemies. Similarly, a prison warden may wish to inflict punishment to establish order in the prison
or to put down prisoners who challenge or annoy him. Historically, the constitutional movement
to limit punishment discretion has been driven by executive officials’ attempts to impose
punishment without law. It was King John’s depredations that led to Magna Carta’s requirement
that punishment be according to the “law of the land.” 132 It was the unauthorized punishments
imposed by the Court of Star Chamber—a prerogative court composed largely of the king’s
ministers—that led to the emphasis on common law rights in England and America.
Both English and American thinkers recognized that executive discretion over punishment
was dangerous because it could lead so easily to tyranny. For example, the tendency of English
kings to impose arbitrary imprisonment led parliament to pass the Habeas Corpus Act of 1679. 133
This statute strengthened the ancient common law writ of habeas corpus, which provided for
judicial review of the lawfulness of incarceration.134 Blackstone described the Habeas Corpus Act
of 1679 as an important structural limit to royal and executive power: “[I]f once it were left in the
power of any, the highest, magistrate to imprison arbitrarily whomever he or his officers thought
proper, (as in France it is daily practised by the crown) there would soon be an end of all other
rights and immunities. . . . To bereave a man of life, or by violence to confiscate his estate, without
accusation or trial, would be so gross and notorious an act of despotism, as must at once convey
the alarm of tyranny throughout the whole kingdom: but confinement of the person, by secretly

130

Id. at 701.
Id. See also Ex parte Hickey, 12 Miss. (4 S. & M.) 751, 765–66 (Miss. Err. App. 1844) (holding that the trial
court’s claim of inherent power to punish newspaper editor for contempt for running an article critical of the judge
was invalid because it gave the judge unlimited power to punish). “It is a maxim of law that where a discretion is
allowed courts in the punishment of defined offenses, that discretion must be regulated by law. But in this instance,
the law, as claimed, sets to itself no bounds, and, under the influence of strong passions, punishment may be
inflicted to a cruel, an unusual and excessive degree.” Id. at 778.
132
See, e.g., Breay & Harrison, supra n.__.
133
See Habeas Corpus Act of 1679, (May 27, 1679) Statutes of the Realm, v. 935–38. 31 Car. II. c. 2, reprinted in
MABEL HILL, LIBERTY DOCUMENTS 431 (Albert Bushnell Hart ed., 2d ed. 1907); see e.g., Amanda L. Tyler, A
“Second Magna Carta”: The English Habeas Corpus Act and the Statutory Origins of the Habeas Privilege,
91 NOTRE DAME L. REV. 1949 (2016) (describing the origins and significance of the Habeas Corpus Act of 1679).
134
See id.
131

18

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
hurrying him to gaol, where his sufferings are unknown or forgotten, is a less public, a less striking,
and therefore a more dangerous engine of arbitrary government.” 135
Similarly, Joseph Story wrote that although the President’s duty to “take care that the laws
be faithfully executed” gives the President an important role in protecting the country’s well-being,
it must not be construed to give him power to impose punishments: “[W]e are not to understand,
that this clause confers on the President any new and substantial power to cause the laws to be
faithfully executed, by any means, which he shall see fit to adopt, although not prescribed by the
Constitution, or by the acts of Congress. That would be to clothe him with an absolute despotic
power over the lives, the property, and the rights of the whole people. A tyrannical President might,
under a pretence of this sort, punish for a crime, without any trial by jury, or usurp the functions
of other departments of the government.”136
c. Penal Effect and the Line between Punishment and Regulation
The question of how to draw the line between punishments and non-penal regulations has
arisen repeatedly from the very beginning of the republic. In Calder v. Bull,137 for example, the
Supreme Court held that a Connecticut statute enacted to overturn a judge’s decision disapproving
a will was not an ex post facto law, even though it retroactively changed legal rules in a way that
deprived the plaintiff of property, because the law did not impose a punishment. 138 In support of
this decision, Justice Chase wrote that an ex post facto law is:
1st. Every law that makes an action, done before the passing of the law, and which
was innocent when done, criminal; and punishes such action. 2nd. Every law that
aggravates a crime, or makes it greater than it was, when committed. 3rd. Every
law that changes the punishment, and inflicts a greater punishment, than the law
annexed to the crime, when committed. 4th. Every law that alters the legal rules of
evidence, and receives less, or different, testimony, than the law required at the time
of the commission of the offence, in order to convict the offender. 139
The Calder court held that the Connecticut law did not violate the ex post facto clause because it
did not impose a punishment—although in making this decision, the Court did not define
“punishment” so as to make the distinction between penal and non-penal laws and practices clear.

135

1 TUCKER’S BLACKSTONE, supra note 84, at 135–36.
See Joseph Story, A FAMILIAR EXPOSITION OF THE CONSTITUTION OF THE UNITED STATES: CONTAINING A BRIEF
COMMENTARY ON EVERY CLAUSE, EXPLAINING THE TRUE NATURE, REASONS, AND OBJECTS THEREOF, DESIGNED
FOR THE USE OF SCHOOL LIBRARIES AND GENERAL READERS, WITH AN APPENDIX CONTAINING IMPORTANT PUBLIC
DOCUMENTS, ILLUSTRATIVE OF THE CONSTITUTION 177-78 (1840) (italics omitted).
137
3 U.S. 386 (1798).
138
See id. at 387, 390. There has been sustained controversy from the time Calder was decided to today as to the
correctness of its decision that the ex post facto clause only applies to criminal and not civil statutes. See, e.g., John
Mikhail, James Wilson, Early American Land Companies, and the Original Meaning of “Ex Post Facto Law”,
17 GEO. J.L. & PUB. POL'Y 79 (2019). This controversy is beyond the scope of the present essay.
139
Calder, 3 U.S. at 390.
136

19

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
The distinction between punishment and regulation repeatedly arose in the nineteenth
century in relation to the power to deport non-citizens. The issue was first debated after passage
of the Alien and Sedition Acts in 1798. 140 These statutes were enacted during a period of
heightened tensions between the United States and France, and gave the President the authority to
arrest and deport such non-citizens “as he shall judge dangerous to the peace and safety of the
United States, or shall have reasonable grounds to suspect are concerned in any treasonable or
secret machinations against the government thereof,”141 or who are male, fourteen years old or
older, and “natives, citizens, denizens, or subjects of [a] hostile nation or government” during a
time of actual conflict.142
The Alien and Sedition Acts generated numerous constitutional objections. Most notable
for our purposes was the Virginia legislature’s complaint that the deportation provisions gave the
President the power to impose punishments without trial, and therefore “subvert[ed] the general
principles of free government” by “uniting legislative and judicial powers to those of executive.” 143
In response, Congress rejected the argument that the Alien Acts gave the President the authority
to punish without trial, arguing that deportation was not a punishment “but . . . merely the removal
from motives of general safety, of an indulgence which there is danger of their abusing, and which
we are in no manner bound to grant or continue.”144 James Madison replied on behalf of the
Virginia legislature that deportation is the same thing as banishment, and that it is therefore a
punishment regardless of whether Congress’s motives were “preventive” or “penal.” Because
deportation has the effect of removing a person from a country where he may have made a
permanent home, acquired property, and established friends and family ties and exposing him to
the dangers of travel at sea during a time of possible conflict, Madison argued that it must be
classified as a punishment: “[I]f a banishment of this sort be not a punishment, and among the
severest of punishments, it will be difficult to imagine a doom to which the name can be
applied.”146
This debate did not definitively resolve the constitutionality of the Alien Acts’ deportation
provisions. Nonetheless, it is illuminating because both sides of the debate focused primarily on
the question of whether deportation has a penal effect. The congressional committee argued that
the Acts were not a punishment but “merely the removal, from motives of general safety, of an
140

See The Alien Friends Act, 1798 Laws, ch. 58, 1 Stat. 570 (1798); The Alien Enemies Act, 1798 Laws ch. 66, 1
Stat. 577 (1798); The Sedition Act, 1798 Laws, ch. 74, 1 Stat. 596 (1798).
141
Alien Friends Act, supra note 140.
142
Alien Enemies Act, supra note 140.
143
JAMES MADISON, VIRGINIA RESOLUTIONS OF 1798, PRONOUNCING THE ALIEN AND SEDITION LAWS TO BE
UNCONSTITUTIONAL, AND DEFINING THE RIGHTS OF THE STATES (1798), reprinted in 4 JONATHAN ELLIOT, DEBATES
IN THE SEVERAL STATE CONVENTIONS, ON THE ADOPTION OF THE FEDERAL CONSTITUTION, AS RECOMMENDED BY
THE GENERAL CONVENTION AT PHILADELPHIA, IN 1787, at 528 (2d ed. 1836) [hereinafter 4 Elliot’s Debates].
144
Report of Select Committee on the Petitions Praying for a Repeal of the Alien and Sedition Laws (February 25,
1799), in ANNALS OF THE CONGRESS OF THE UNITED STATES, FIFTH CONGRESS 2987 (1851) [hereinafter Select
Committee Report]
146
JAMES MADISON, REPORT OF THE COMMITTEE TO WHOM WERE REFERRED THE COMMUNICATIONS OF VARIOUS
STATES, RELATIVE TO THE RESOLUTIONS OF THE LAST GENERAL ASSEMBLY OF THIS STATE, CONCERNING THE ALIEN
AND SEDITION LAWS, H.D. 1799–1800 Sess. (Va. 1800) [hereinafter M ADISON, REPORT], reprinted in 4 Elliott’s
Debates, supra note 138, at 554–55.

20

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
indulgence . . . .”147 Madison, on the other hand, asserted that deportation of those who may have
lived here for years, formed relationships, acquired property, and sought citizenship is
“banishment,” a traditional method of punishment. 148 He argued that Congress may not use a
punishment to achieve a regulatory purpose unless it complies with the Constitution’s procedural
restrictions on the government’s power to punish. He further argued that the executive should not
possess independent authority to inflict punishment.149
The issues raised by the Alien Acts arose once again nearly a century later in Wong Wing
v. United States.150 The defendants in this case were Chinese citizens convicted under a law
providing that any Chinese person found to be illegally present in the United States should be
sentenced to up to a year of imprisonment at hard labor followed by deportation. 151 The defendants
were convicted in a summary hearing before a commissioner152 and sentenced to 60 days at hard
labor in a house of correction before deportation. 153 They argued that imprisonment at hard labor
is an infamous punishment within the meaning of the Constitution, and that therefore they had
been unconstitutionally denied their rights to indictment by grand jury and to a jury trial. 154 The
Supreme Court agreed, noting that “for more than a century[,] imprisonment at hard labor in the
state prison or penitentiary or other similar institution has been considered an infamous punishment
in England and America . . . .”155 The fact that Congress’s purposes were regulatory did not
prevent the statute from being subject to the constitutional limits on punishment: “[W]hen
Congress sees fit to further promote such a policy by subjecting the persons of such aliens to
infamous punishment at hard labor, or by confiscating their property, we think such legislation, to
be valid, must provide for a judicial trial to establish the guilt of the accused.” 156
The question of the line between punishment and non-penal regulation also arose in the
context of the late-nineteenth century trend (which persists to this day) 157 toward housing
condemned offenders awaiting execution in solitary confinement in the state penitentiary. 158 This
147

Select Committee Report, supra note 144.
See MADISON, REPORT, supra note 146; see also, e.g., 4 TUCKER’S BLACKSTONE, supra note 91, at 112, 121,
138, 187, 245, 376–77 (providing various examples, ancient and modern, in which banishment is prescribed as a
punishment).
149
See Madison, REPORT, supra note 146. Nearly a century later, the Supreme Court concluded that deportation is
not the equivalent of banishment because traditionally banishment was only a punishment for citizens, not aliens.
See Fong Yue Ting v. United States, 149 U.S. 698, 730 (1893) (“The order of deportation is not a punishment for
crime. It is not a banishment, in the sense in which that word is often applied to the expulsion of a citizen from his
country by way of punishment.”).
150
163 U.S. 228 (1896).
151
Id. at 233–34.
152
Id. at 229.
153
Id. at 234.
154
Id.
155
Id. at 237.
156
Id.
157
See Marah Stith McLeod, Does the Death Penalty Require Death Row? The Harm of Legislative Silence,
77 OHIO ST. L.J. 525, 527 (2016).
158
See, e.g., DAVID GARLAND, PECULIAR INSTITUTION: AMERICA'S DEATH PENALTY IN AN AGE OF ABOLITION 52,
116 (2010) (describing the historical trend toward moving executions out of the public square and into the private
confines of the prison).
148

21

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
trend seems not to have been intended to further any penal purpose; rather, it appears to have been
motivated by the desire to hide executions from public view.159 For most of the nineteenth century
– indeed, for nearly all of history – executions were conducted in the public square. 160 Toward the
end of the nineteenth century, a number of states moved executions out of the public square and
behind prison walls.161 This move was supported by some people who believed that public
executions coarsened public sensibility by turning death into a spectacle. 162 But there is evidence
that states were actually motivated by a desire to undercut the death penalty abolition movement
by removing executions from public view and thus minimizing the emotional force of the
abolitionist argument. 163
The Colorado statute considered by the Supreme Court in the case In re Medley is a good
example of these new death penalty statutes.164 In 1889, Colorado adopted a new law governing
executions in the state. The old law had provided that prisoners awaiting execution should be kept
in a county jail for fifteen to twenty-five days before execution by hanging. 165 Both the jail term
and the execution were supervised by the county sheriff. 166 The new law, by contrast, required that
a prisoner sentenced to death be kept in solitary confinement in the state penitentiary for two to
four weeks prior to execution.167 The execution should be “enclosed from public view within the
walls of the penitentiary . . . .”168 The warden had the power to set the time and date of the
execution, but was required to keep this information “secret” from everyone except those people
invited to witness the execution.169 These witnesses were also required to keep the time and date
of the execution secret.170 All witnesses were also prohibited from divulging the details of what
happened at the execution “beyond the statement of the fact that such convict was on the day in
question duly executed according to law at the state penitentiary . . . .” 171 Thus under the new
159

See, e.g., STUART BANNER, THE DEATH PENALTY: AN AMERICAN HISTORY 148 (2002) (noting that in nineteenth
century America, many criticized public executions on the ground that they generated public sympathy for offenders);
GARLAND, supra note 158, at 135 (stating that moving executions out of public view “is best viewed as the ongoing
effort of government officials to exert ever-tighter control over a fraught undertaking and to manage the meanings that
it put into circulation”); LOUIS MASUR, RITES OF EXECUTION: CAPITAL PUNISHMENT AND THE TRANSFORMATION OF
AMERICAN CULTURE, 1776–1865, at 96 (1989) (describing the perception among some nineteenth century elites that
public executions undermined public order); Michael Madow, Forbidden Spectacle: Executions, the Public and the
Press in Nineteenth Century New York, 43 BUFF. L. REV. 461, 493–497 (1995) (describing debate among historians
concerning the reasons executions were moved out of the public square).
160
See, e.g., Garland, supra note 154, at 52, 116; JOHN D. BESSLER, DEATH IN THE DARK: MIDNIGHT EXECUTIONS IN
AMERICA 25 (1997) (noting that executions in early America were designed to encourage public visibility).
161
See, e.g., Garland, supra note 154, at 116.
162
See BANNER, supra note 154, at 146, 153; LAWRENCE FRIEDMAN, CRIME AND PUNISHMENT IN AMERICAN
HISTORY 76 (1993); MASUR, supra note 155, at 97.
163
See, e.g., BANNER, supra note 155, at 148; GARLAND, supra note 154, at 135; MASUR, supra note 155, at 97;
Madow, supra note 155, at 491, 557.
164
134 U.S. 160 (1890).
165
Id. at 167.
166
Id.
167
Id. at 163–64.
168
Id. at 163.
169
Id. at 164.
170
Id.
171
Id.

22

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
Colorado statute, the offender was kept in solitary confinement in a location likely to be remote
from the place of crime and conviction; neither the offender nor the public was to be informed of
the execution’s time and place; the execution took place behind prison walls; and the details of the
execution—including, for example, the suffering undergone by the offender—were to be kept
secret as well.172 The whole focus of the statute was on isolation and secrecy.
An offender named James Medley challenged the constitutionality of this statute. Medley
committed a murder in May of 1889, two months before Colorado adopted its new execution
law.173 Medley was tried and convicted after adoption of the new execution law, and was sentenced
under its provisions.174 He challenged his conviction on the ground that the new law imposed an
ex post facto punishment in violation of Article I section 10 of the United States Constitution. 175
He argued that the imposition of solitary confinement in a state penitentiary—albeit only for a
period of two to four weeks—substantially added to the pain and suffering of his execution. 176
Because the new statute imposed additional punishment, and because it was adopted after Medley
committed his crime, it violated the constitutional prohibition of ex post facto punishments. 177
The Supreme Court agreed.178 It held that placing the prisoner in solitary confinement prior
to execution added so much suffering to the execution that it had to be considered an additional
punishment for constitutional purposes. 179 Thus the new statute could not be applied to a prisoner
whose crime occurred before the statute took effect without violating the prohibition of ex post
facto punishments.180 In reaching this decision, the Court took note of both England’s and
America’s prior experience with solitary confinement. During the reign of George II, the English
parliament authorized solitary confinement prior to execution in order to add “further terror and
[a] peculiar mark of infamy” to the sentence.181 But parliament repealed this law during the reign
of William IV because “[i]n Great Britain, as in other countries, public sentiment revolted against
this severity. . . .”182 Similarly, the Court noted, America had experimented with solitary
confinement earlier in the nineteenth century and found it unduly harsh:
[E]xperience demonstrated that there were serious objections to it. A considerable
number of the prisoners fell, after even a short confinement, into a semi-fatuous
condition, from which it was next to impossible to arouse them, and others became
violently insane; others, still, committed suicide; while those who stood the ordeal
better were not generally reformed, and in most cases did not recover sufficient
mental activity to be of any subsequent service to the community. . . . [S]ome thirty
172

See id. at 161-64.
Id. at 161–62.
174
Id.
175
Id. at 162.
176
Id. at 170, 172.
177
Id. at 172-73.
178
Id. at 173.
179
Id. at 171.
180
Id. at 173.
181
Id. at 170.
182
Id.
173

23

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
or forty years ago the whole subject attracted the general public attention, and its
main feature of solitary confinement was found to be too severe. 183
Given the fact that solitary confinement was historically used as a heightened form of punishment,
and given its extremely harsh effects, the Supreme Court held that it could not be considered a
mere administrative measure. Rather, it was “an additional punishment of the most important and
painful character . . . .”184 Thus Medley’s punishment violated the ex post facto clause, and
Medley’s conviction had to be overturned—even though this meant that Medley would escape
punishment for murder altogether. 185
In sum, the United States Constitution is designed to limit the discretion of government
officials—including executive officials—to impose punishments. Under the Constitution, the line
between punishments and non-penal regulations is determined primarily by examining the penal
effect of the government’s action. If the government employs a method that has traditionally been
considered a punishment—such as banishment or imprisonment in a penitentiary—it is to be
classified as a punishment even if the government’s asserted purpose is regulatory rather than
penal. Similarly, if the government imposes conditions that increase the harshness of the offender’s
suffering beyond what could reasonably be considered part of an offender’s sentence, it is a new
punishment, as the Medley Court concluded in the context of solitary confinement.
III. HISTORY AND EFFECTS OF SOLITARY CONFINEMENT
As the above discussion demonstrates, the Supreme Court’s current treatment of the
decision to place an offender in solitary confinement as an administrative matter rather than a
punishment conflicts directly with its previous holding that solitary confinement is a punishment
for constitutional purposes.186 Astonishingly, up to now, the Court has not shown any awareness
of the conflict.
Resolving this conflict is important because most states today impose solitary confinement
without statutory authorization. According to Professor Alexander Reinert, only four states
(Delaware, Washington, South Carolina, and Pennsylvania) have statutes authorizing the use of
solitary confinement outside the context of death row, and three states (Idaho, Pennsylvania, and
Wyoming) have statutes authorizing it for those under a sentence of death. 187 This means that the
vast majority of prisoners subjected to solitary confinement are sent there by prison officials

183

Id. at 168.
Id. at 171.
185
Id. at 173–74.
186
Compare, e.g., Wilkinson v. Austin, 545 U.S. 209, 213 (2005) with In re Medley, 134 U.S. 160, 171 (1890).
Interestingly, several courts have held that judges lack the authority to impose solitary confinement as part of an
offender’s sentence where it is not explicitly authorized by penal statute. See, e.g., Fludd v. Goldberg, 854 N.Y.S.2d
362, 367 (2008) (invalidating sentence of four days per year in solitary confinement due to lack of statutory
authorization); State v. McHenry, 525 N.W.2d 620, 627 (Neb. 1995) (same); State v. Snitzky, Nos. 74706, 74811,
1998 WL 827611, at *2 (Ohio Ct. App. 1998) (same). These cases appear to recognize the same truth the Medley
court did—solitary confinement is an additional punishment above and beyond a mere prison sentence.
187
See Alexander A. Reinert, Solitary Troubles, 93 NOTRE DAME L. REV. 927, 959–60 (2018).
184

24

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
without statutory authorization, based solely on their authority to administer prisons. If solitary
confinement is a punishment, this conduct is unconstitutional.
Before reaching a final conclusion concerning how to resolve this conflict, this essay will
look briefly at the history and effects of solitary confinement to determine whether the Medley
Court’s findings concerning its extreme harshness are correct.
a.

Solitary Confinement in Nineteenth Century America

Solitary confinement has a long and troubled history in the United States. It was introduced
by eighteenth and nineteenth century reformers who wished to reduce use of the death penalty,
corporal punishments, and shaming punishments. The reformers argued that imprisonment and
solitary confinement would operate as more effective and humane engines of rehabilitation. 188
Some reformers believed that solitary confinement could transform offenders into peaceful and
productive members of society by segregating them from harmful outside influences and providing
time for reflection and prayer.189 Others believed that the “terror” induced by solitary confinement
would deter crime.190 Led by Pennsylvania and New York, a number of states created
penitentiaries in the nineteenth century and amended their penal statutes to provide for
imprisonment and solitary confinement. 191
Although solitary confinement enjoyed a period of popularity over several decades in the
nineteenth century, it came into disrepute as its effects became known. 192 Experience showed that
solitary confinement did not serve to rehabilitate prisoners; rather, it destroyed them. 193 Numerous
cases of hallucination, insanity, despair, suicide, illness, and death were reported by prison
administrators and outside observers.194 Writers such as Alexis de Tocqueville, Gustave
Beaumont, and Charles Dickens, who observed prisoners subject to solitary confinement,

188

See, e.g., Craig Haney & Mona Lynch, Regulating Prisons of the Future: A Psychological Analysis of Supermax
and Solitary Confinement, 23 N.Y.U. REV. L. & SOC. CHANGE 477, 482 (1997) (describing reformers’ arguments for
the institution of solitary confinement); ADAM JAY HIRSCH, THE RISE OF THE PENITENTIARY 19 (1992) (same); see
generally David M. Shapiro, Solitary Confinement in the Young Republic, 133 HARV. L. REV. 542 (2019)
(describing the early history of solitary confinement in America).
189
See Haney & Lynch, supra note 188.
190
See, e.g., WILLIAM PALEY, THE PRINCIPLES OF MORAL AND POLITICAL PHILOSOPHY 436 (London, J. Davis ed.,
7th ed. 1790) (discussing the potential deterrent effect of the “terror” induced by solitary confinement).
191
Ashley T. Rubin & Keramet Reiter, Continuity in the Face of Penal Innovation: Revisiting the History of
American Solitary Confinement, 43 LAW & SOC. INQUIRY 1604, 1612–15 (2018) (describing the nineteenth century
trend toward adopting solitary confinement as a method of punishment); David M. Shapiro, Solitary Confinement in
the Young Republic, 133 Harv. L. Rev. 542 (2019) (providing a detailed historical account of early experiments
with solitary confinement, focusing particularly on Pennsylvania).
192
See e.g., id. at 1617–18 (highlighting critiques of the original models of solitary confinement).
193
See, e.g., John F. Stinneford, Experimental Punishments, 95 NOTRE DAME L. REV. 39, 60–64 (2019) (describing
the history of solitary confinement in the nineteenth century).
194
See Haney & Lynch, supra note 188, at 483–84 (describing nineteenth century reports of the extremely harsh
effects of solitary confinement); Peter Scharff Smith, The Effects of Solitary Confinement on Prison Inmates: A Brief
History and Review of the Literature, 34 CRIME & JUST. 441, 459 (2006) (same).

25

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
described its effects as “fatal”195 and “worse than any torture.”196 Thus solitary confinement fell
out of favor as a method of punishment by the 1860s and largely disappeared from the United
States by the 1870s and 1880s.197
b. Late Twentieth Century Resurgence of Solitary Confinement
Long-term solitary confinement experienced a resurgence starting in the 1960s and 1970s
before drastically accelerating with the rise of “Supermax” prisons in the 1980s and 1990s. 198 By
2004, forty-four states and the District of Columbia had joined the federal government in opening
supermax facilities.199 According to recent estimates by the Liman Center at Yale Law School,
more than 60,000 offenders are currently housed in solitary confinement. 200 Of these,
approximately 11,500 have spent a year or more in solitary, and nearly 3,000 have spent six years
or more in solitary.201
The conditions of solitary confinement vary from facility to facility, but the supermax unit
at the Pelican Bay penitentiary in California, as described by the court in Madrid v. Gomez, is
fairly representative.202 At the time Madrid was decided, most prisoners in the secure housing unit
(SHU) at Pelican Bay were kept in total isolation, twenty-four hours per day, seven days per
195

G. DE BEAUMONT & A. DE TOQUEVILLE, ON THE PENITENTIARY SYSTEM IN THE UNITED STATES AND ITS
APPLICATION IN FRANCE 5 (Francis Lieber, trans., 1833) (“This trial, from which so happy a result had been
anticipated, was fatal to the greater part of the convicts: in order to reform them, they had been submitted to
complete isolation; but this absolute solitude, if nothing interrupt[s] it, is beyond the strength of man; it destroys the
criminal without intermission and without pity; it does not reform, it kills. The unfortunates, on whom this
experiment was made, fell into a state of depression, so manifest, that their keepers were struck with it; their lives
seemed in danger, if they remained longer in this situation; five of them, had already succumbed during a single
year; their moral state was not less alarming; one of them had become insane; another, in a fit of despair, had
embraced the opportunity when the keeper brought him something, to precipitate himself from his cell, running the
almost certain chance of a mortal fall.”)
196
CHARLES DICKENS, AMERICAN NOTES FOR GENERAL CIRCULATION 123-24 (Paris, A. & W. Galignani & Co.
1842) (“I hold this slow and daily tampering with the mysteries of the brain, to be immeasurably worse than any
torture of the body . . . .”). Hans Christian Andersen similarly described solitary confinement he witnessed in a
Swedish prison as “a well-built machine—a nightmare for the spirit.” P ICTURES OF SWEDEN 56 (Richard Bentley,
ed., 2d ed. 1852).
197
See, e.g., Rubin & Reiter, supra note 191, at 1617; Shapiro, supra note 188, at 572. Shorter term solitary
confinement continued to exist at the margins of penal practice to discipline unruly prisoners, and there is evidence
that some prison officials subjected some prisoners to extended periods in solitary confinement. See, e.g.,
Stinneford, Experimental Punishments, supra note 193, at 65–68.
198
Stinneford, Experimental Punishments, supra note 193, at 69–75.
199
See DANIEL P. MEARS, EVALUATING THE EFFECTIVENESS OF SUPERMAX PRISONS 40 (2006).
200
REFORMING RESTRICTIVE HOUSING,supra n. 2, at 4. A recent study indicates that the rapid increase in the
number of prisoners subjected to solitary confinement has been caused largely by increases in the length of time
individual prisoners are subjected to it. See Ryan T. Sakoda & Jessica T. Simes, Solitary Confinement and the U.S.
Prison Boom, CRIM. JUST. POL’Y REV. 18–19 (2019).
201
See REFORMING RESTRICTIVE HOUSING, supra note 2, at 14. These numbers are merely estimates because not all
states reported the duration of current inmates’ time in solitary confinement. Those that did report this data indicated
that just under 19% of inmates currently in solitary confinement had spent a year or more there, and just under 5%
had spent six years or more there. Id.
202
889 F. Supp. 1146, 1228–30 (N.D. Cal. 1995).

26

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
week.203 Each prisoner spent about twenty three hours per day in a cell approximately the size of
a parking spot.204 For about an hour each day, the prisoner would be let into a caged exercise area,
resembling a dog run, where he was permitted to exercise alone. 205 Meals were served through a
slot in the door.206 Prisoners were shackled during all interactions with guards. 207 Any meetings
with visitors were conducted through thick glass windows—but such visits were rare because of
the prison’s isolated location.208 In sum, as the former warden testified, imprisonment in the SHU
amounted to “‘virtual total deprivation, including, insofar as possible, deprivation of human
contact.’”209 Numerous supermax facilities share the characteristics described above. 210
c. Current Evidence Concerning the Effects of Solitary Confinement
For purposes of determining whether solitary confinement is a punishment or a mere
administrative action, the key question is this: How harsh are the effects of solitary confinement
in comparison to the effects of imprisonment generally? Numerous studies of the effects of solitary
confinement have been performed over the past forty years. 211 They show that solitary confinement
has extraordinarily harmful effects on prisoners subjected to it. 212 More importantly for our
purposes, they show that the harmful effects of solitary confinement are extreme, not just as an
absolute matter, but also in comparison to the effects of imprisonment generally.
Suffering in prison seems to follow a curve. Studies have shown that prisoners at the
beginning of their sentence tend to experience very high levels of depression and anxiety. 213 But
these symptoms tend to recede as time goes on, so that several years into a term of incarceration,
prisoners’ happiness levels generally settle somewhere near their pre-incarceration baseline. 214
This curve corresponds to the theory of “hedonic adaptation,” which refers to the general human
tendency to revert back to a baseline level of happiness after a significantly positive or negative
experience.215

203

Id. at 1228–29.
Id.
205
Id. at 1229.
206
Id.
207
Id.
208
Id.
209
Id. at 1230.
210
See generally Judith Resnik, Sarah Baumgartel & Johanna Kalb, Time-in-Cell: Isolation and Incarceration,
125 YALE L.J. F. 212 (2016) (describing the characteristics of American supermax facilities).
211
See infra notes 211–20.
212
See id.
213
See John Bronsteen et al., Happiness and Punishment, 76 U. CHI. L. REV. 1037, 1047–48 (2009). Some groups of
prisoners, such as the mentally ill, may not be able to adapt to prison as well as others. See, e.g., E. Lea Johnston,
Vulnerability and Just Desert: A Theory of Sentencing and Mental Illness, 103 J. CRIM. L. & CRIMINOLOGY 147,
204

159–60 (2013) (“Prison is a toxic environment for individuals with serious mental health problems”).
214
See Bronsteen supra note 209, at 1047.
215
See id at 1040.

27

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
Suffering in solitary confinement follows a very different trajectory. Studies show that
inmate well-being is significantly harmed by solitary confinement, and harms get worse over
time.216 Effects include “extreme forms of psychopathology,” 217 and include suicidal thoughts,
hallucinations, perceptual distortions, violent fantasies, talking to one’s self, overall deterioration,
mood swings, emotional flatness, chronic depression, social withdrawal, confused thought
processes, oversensitivity to stimuli, irrational anger, and ruminations. 218
Compared to the symptoms displayed by prisoners subjected to incarceration generally, the
effects of solitary are extraordinarily grave. For example, a recent study compared the
psychological well-being of prisoners who had been held in the Secure Housing Unit (SHU) of the
Pelican Bay state prison in California for ten years or more to prisoners held in the general
population of the same prison for a comparable period of time.219 The study excluded prisoners
listed on the prison’s mental health caseload,220 and prisoners from both groups were randomly
selected.221 It found that extreme long-term SHU prisoners reported nearly twice the number of
stress-related trauma and isolation-related pathology symptoms as prisoners in the general
population,222 and much greater intensity of stress, trauma, and isolation-related pathology
symptoms.223 Finally, the study found that “the prisoners in long-term solitary confinement were
216

Numerous studies demonstrate the harmful effects of solitary confinement on prisoner well-being. As discussed
above, these effects were well-known in the nineteenth century. Modern studies in America and Europe have
replicated these findings. Many such studies are summarized in the following literature reviews: See Elizabeth
Bennion, Banning the Bing: Why Extreme Solitary Confinement Is Cruel and Far Too Usual Punishment, 90 IND.
L.J. 741, 753–63 (2015); Haney & Lynch, supra note 183; Craig Haney, Mental Health Issues in Long-Term
Solitary and “Supermax” Confinement, 49 CRIME & DELINQ. 124, 132 (2003); Stuart Grassian, Psychiatric Effects
of Solitary Confinement, 22 WASH. U. J.L. & POL’Y 325 (2006); Smith, supra note 189; Bruce Arrigo & Jennifer
Bullock, The Psychological Effects of Solitary Confinement on Prisoners in Supermax Units: Reviewing What We
Know and Recommending What Should Change, 52 INT’L. J. OFFENDER THERAPY AND COMP. CRIMINOLOGY 622
(2008). A group of researchers published a study in 2010, known as the “Colorado study,” claiming that prolonged
solitary confinement has few negative effects and may actually be good for mentally ill prisoners. See MAUREEN
O’KEEFE ET AL., ONE YEAR LONGITUDINAL STUDY OF THE PSYCHOLOGICAL EFFECTS OF ADMINISTRATIVE
SEGREGATION 1 (2010). In 2016, a meta-analysis based largely on the Colorado study reached a similar conclusion.
See Robert Morgan et al., Quantitative Synthesis of the Effects of Administrative Segregation on Inmates’ WellBeing, 22 PSYCHOL., PUB. POL’Y, AND L., Aug. 8, 2016, at 18. These studies have been shown to be so deeply
flawed as to be meaningless. See, e.g., Craig Haney, The Psychological Effects of Solitary Confinement: A
Systematic Critique, 47 CRIME & JUST. 365, 369–70 (2018) (flaws include contamination of treatment and
comparison groups, poorly constructed sampling and group composition, uncontrolled differences in conditions of
“general population” and “administrative segregation” imprisonment, failure to control or record treatment dose, and
a number of problems undermining the reliability of data collection); Stuart Grassian & Terry A. Kupers, The
Colorado Study vs. The Reality of Supermax Confinement, 13 CORRECTIONAL MENTAL HEALTH REP. 1, 7–8 (2011)
(noting that, among other problems, the study ignored key data including the fact that the prisons recorded almost 7
times as many psychiatric emergencies per inmate in solitary confinement than among those in the general
population).
217
See Haney, supra note 212, at 134.
218
See id.
219
See Craig Haney, Restricting the Use of Solitary Confinement, 1 ANN. REV. OF CRIMINOLOGY 285, 292–293
(2018).
220
See id. at 292 n. 2.
221
See id at 291.
222
See id. at 292–293.
223
See id. at 293.

28

Electronic copy available at: https://ssrn.com/abstract=3605517

ROUGH DRAFT
not only significantly more lonely than the long-term GP prisoners . . . , but also reported
extremely high levels of loneliness rarely found anywhere in the literature.” 224 Similar results have
been found by other comparative studies.225 There is broad consensus that placing a prisoner in
solitary has a psychological and physical impact well above any that a prisoner would experience
in general population.”
In sum, historical and modern empirical evidence concerning the effects of solitary
confinement strongly supports the Supreme Court’s conclusion in In re Medley that solitary
confinement is an additional punishment within the meaning of the United States Constitution. It
causes suffering and harm well beyond that caused by a general sentence of imprisonment.
Moreover, unlike the suffering caused by imprisonment generally, the suffering continues to
intensify over time. Because solitary confinement is an additional punishment, its imposition does
not fall within the legitimate discretion of prison officials seeking to promote prison security.
Rather, before it may be imposed, it must be authorized by the penal statute governing the offense
of conviction in effect at the time the prisoner committed his crime, and must be imposed as part
of the offender’s sentence.
CONCLUSION
Solitary confinement is a punishment. The Supreme Court was correct when it reached this
conclusion in 1890, and more recent cases ignoring this conclusion are incorrect. When a prison
official inflicts solitary confinement on a prisoner without prior statutory and sentencing
authorization, that official imposes a “punishment without law.” Basic principles of separation of
powers and individual rights dictate that this practice must stop.

224

See id.
See, e.g., Smith, supra note 189 at 477-480 (describing comparative studies concerning the effects of solitary
confinement on Norwegian and Danish remand prisoners).
225

29

Electronic copy available at: https://ssrn.com/abstract=3605517