The Constitutional Law of Isolated Confinement:
A Quick and Dirty Review

John Boston

The Legal Aid Society
Prisoners' Rights Project
90 Church Street
New York, N.Y. 10007

Prepared for the conference on
Challenging Supermax Prisons
National Prison Project of the
American Civil Liberties Union

February 23-24, 2000

The Constitutional Law of Isolated Confinement:
A Quick and Dirty Review

John Boston

Prisoners and their advocates have been litigating about isolated confinement1 as long as
there has been prison litigation.2 They have had considerable success in mitigating the worst
excesses of such confinement: filthy conditions, lack of lighting and ventilation, deprivation of
medical care, routine use of "strip cells," etc. They have been far less successful in challenging the
core of such confinement: the deprivation of human contact and other sensory and intellectual
stimulation.
There is a certain irony to this result, since the disastrous consequences of the Nineteenth
Century solitary confinement regimes were so well known and so uncontroversial as to be treated
as common knowledge by the Supreme Court:
A considerable number of prisoners fell, after even a short confinement, into a semifatuous 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.3

1

I use this term to refer to confinement that involves both separation from the general prison
population (i.e., segregation) and a substantial degree of restriction of contact even with other
separated prisoners (though not necessarily as rigorous as the Nineteenth Century regimes of solitary
confinement).
2

See, e.g., Hancock v. Avery, 301 F.Supp. 786, 791-92 (M.D.Tenn. 1969) (condemning
confinement in cell without light or ventilation).
3

In re Medley, 134 U.S. 160, 168 (1890) (striking down a statute retroactively imposing solitary
confinement as an ex post facto law).
1

Modern courts have not denied these consequences. One relatively recent decision observed
that "the record shows, what anyway seems pretty obvious, that isolating a human being from other
human beings year after year or even month after month can cause substantial psychological damage,
even if the isolation is not total."4 The court added that "there is plenty of medical and psychological
literature concerning the ill effects of solitary confinement (of which segregation is a variant). . . ."5
Other courts have made similar observations.6 More recently, the district court in the Pelican Bay
SHU litigation concluded after hearing testimony from experts in corrections and mental health, that
"many, if not most, inmates in the SHU experience some degree of psychological trauma in reaction
to their extreme social isolation and the severely restricted environmental stimulation in SHU."7 I
know of no decision that has rejected this conclusion.
The legal significance of this psychological harm is a different question. In Davenport v.
Derobertis, immediately after acknowledging the "literature concerning the ill effects of solitary
confinement," the court (per Judge Posner, well-known for applying economic analyses to legal
questions) proceeded to minimize it: "Of course, it is highly probable that the experience of being
imprisoned inflicts psychological damage whether or not the prisoner is isolated, so it is only the
marginal psychological damage from segregation that is relevant. And the infliction of disutility,

4

Davenport v. DeRobertis, 844 F.2d 1310, 1313 (7th Cir. 1988), cert. denied, 488 U.S. 908
(1989).
5

Id. at 1316, citing Grassian, Psychopathological Effects of Solitary Confinement, 140
Am.J.Psychiatry 1450 (1983).
6

See Langley v. Coughlin, 715 F.Supp. 522, 540 (S.D.N.Y. 1989) (citing Dr. Grassian's affidavit
re effects of SHU placement on disordered individuals); Baraldini v. Meese, 691 F.Supp. 432, 44647 (D.D.C. 1988) (citing Dr. Grassian's testimony re sensory disturbance, perceptual distortions, and
other psychological effects of segregation), rev'd on other grounds, 884 F.2d 615 (D.C.Cir. 1989);
Bono v. Saxbe, 450 F.Supp. 934, 946 ("[p]laintiffs' uncontroverted evidence showed the debilitating
mental effect on those inmates confined to the control unit."), aff'd in part and remanded in part on
other grounds, 620 F.2d 609 (7th Cir. 1980).
7

Madrid v. Gomez, 889 F.Supp. 1146, 1235 (N.D.Cal. 1995).
2

to borrow a convenient economic term, is one of the objectives of criminal punishment. . . ."8 An
even more dismissive attitude is displayed in a recent decision of the Massachusetts Supreme
Judicial Court, which held that evidence that isolated confinement can cause "serious psychiatric
harm" did not raise an issue of disputed fact requiring a trial. Rather, the court simply observed that
prior federal court decisions evinced a "widely shared disinclination" or a "reluctan[ce]" to strike
down isolated confinement, and that one of its own older decisions upheld confinement in somewhat
harsher conditions. It viewed these decisions as sufficiently authoritative to obviate the need even
to consider the proffered evidence of psychological harm.9
The result in Torres is consistent with the general refusal of courts to find isolated
confinement unconstitutional absent aggravating circumstances. In this regard, the high-water mark
for prisoner advocates was the Supreme Court's decision in the Arkansas prison litigation, which
upheld the lower court's placement of a 30-day limit on punitive segregation, but did so only in light
of the inadequate diet, overcrowding, and misconduct by prison staff demonstrated by the record in
that case.10 Any hope that this decision might ultimately lead to a constitutional limit on the use or
duration of isolated confinement per se was quickly disappointed. This point is best illustrated by
the decision in the Pelican Bay litigation, which represents both the most thorough examination of

8

Id.
The bottom line in the Davenport case is that the appeals court upheld the lower court's
requirement that prisoners in a segregation unit receive a minimum of five hours of out-of-cell
recreation a week, but held that the requirement of a minimum of three showers a week (the
prisoners received one a week) lacked support either in the record or in law. 844 F.2d at 1314-16.
9

Torres v. Commissioner of Correction, 427 Mass. 611, 614-15, 695 N.E.2d 200, 203-04 (1998),
quoting Jackson v. Meachum, 699 F.2d 578, 583 (1st Cir. 1983), and Santana v. Collazo, 714 F.2d
1172, 1179 (1st Cir. 1983), cert. denied, 466 U.S. 974 (1984). The court's reference to its own
earlier decision is a bit disingenuous. That case held only that the Eighth Amendment does not
forbid confinement behind a solid door for no more than 15 days as punishment for disciplinary
offenses by prisoners already in segregation. (By contrast, prisoners could be sentenced to the unit
at issue in Torres for up to ten years.) The trial court judge had held that the degree of sensory
deprivation at issue did not cause psychological harm. Libby v. Commissioner of Correction, 385
Mass. 421, 432 N.E.2d 486, 493-94 (1982).
10

Hutto v. Finney, 437 U.S. 678, 686-87 (1978).
3

these issues, and the one most sympathetic to prisoners:
Here, the record demonstrates that the conditions of extreme social isolation
and reduced environmental stimulation found in the Pelican Bay SHU will likely
inflict some degree of psychological trauma upon most inmates confined there for
more than brief periods. Clearly, this impact is not to be trivialized; however, for
many inmates, it does not appear that the degree of mental injury suffered
significantly exceeds the kind of generalized psychological pain that courts have
found compatible with Eighth Amendment standards. While a risk of a more serious
injury is not non-existent, we are not persuaded . . . that the risk of developing an
injury to mental health of sufficiently serious magnitude . . . is high enough for the
SHU population as a whole, to find that current conditions in the SHU are per se
violative of the Eighth Amendment with respect to all potential inmates.11
As this passage suggests, success in challenging isolated confinement has generally occurred
at the margins. These decisions fall into three broad categories.

1.

Vulnerable populations
The potential mental health consequences of isolated confinement have been recognized at

least to the extent that courts have excluded persons with pre-existing psychiatric illness or
vulnerability from such confinement. The leading case again is the Pelican Bay decision, which
upheld SHU confinement for most prisoners, but excepted
those who the record demonstrates are at a particularly high risk for suffering very
serious or severe injury to their mental health, including overt paranoia, psychotic
breaks with reality, or massive exacerbations of existing mental illness as a result of
the conditions in the SHU. Such inmates consist of the already mentally ill, as well
as persons with borderline personality disordered, brain damage or mental
retardation, impulse-ridden personalities, or a history of prior psychiatric problems
or chronic depression. For these inmates, placing them in the SHU is the mental
equivalent of putting an asthmatic in a place with little air to breathe.12

11

Madrid v. Gomez, 889 F.Supp. 1146, 1265 (N.D.Cal. 1995) (emphasis in original).

12

Madrid v. Gomez, 889 F.Supp. at 1265; see also Casey v. Lewis, 834 F.Supp. 1447, 1548-49
(D.Ariz. 1993) (condemning placement and retention of mentally ill prisoners in lockdown status);
Langley v. Coughlin, 715 F.Supp. 522, 540 (S.D.N.Y. 1988) (holding that psychiatric evidence that
prison officials fail to screen out from SHU "those individuals who, by virtue of their mental
condition, are likely to be severely and adversely affected by placement there" raises a triable Eighth
Amendment issue); Inmates of Occoquan v. Barry, 717 F.Supp. 854, 868 (D.D.C. 1989) (holding
that inmates with mental health problems must be placed in a separate area or a hospital and not in
administrative/punitive segregation area).
4

Thus, the risk that the court found too diffuse to be actionable as applied to the prison population as
a whole was found to constitute an Eighth Amendment violation as applied to populations who could
be shown to have identifiable pre-existing risk factors.
As a practical matter, of course, enforcing such a view requires adequate and unbiased mental
health screening, which is not guaranteed in a prison environment. Indeed, there may be powerful
institutional factors militating against identifying persons at particular risk from isolated
confinement.13

2.

Due process concerns
In upholding isolated confinement notwithstanding the potential injury it may cause, courts

give weight to prison officials' legitimate interest in disciplining prisoners who have broken prison
rules and in preventing disruptive or assaultive actions by prisoners who present risk of such
behavior.14 The obvious next question is whether prison officials are under any constitutional
obligation of care to make sure that the prisoners they place in isolation actually merit such
treatment. That is, are prisoners entitled to procedural protections in connection with placement in
isolated confinement?
Before 1995, the law in most jurisdictions was that punitive segregation of any substantial
duration required a Wolff15 hearing, but that administrative segregation required due process
protections only if state rules or regulations created a "liberty interest" by imposing substantive limits
13

See Madrid, 889 F.Supp. at 1225 (citing evidence of prison staff's concern--described by one
expert witness as "an almost obsessive preoccupation"--that prisoners are malingering or
manipulating in their dealings with the medical and mental health system).
14

See, e.g., Madrid, 889 F.Supp. at 1263. A leading statement of this view appears in a decision
concerning conditions at the high-security federal penitentiary in Marion, Illinois: "The current
conditions, ghastly as they are, testify in a weird way to our nation's aspirations to a humane criminal
justice system, for they result from forbidding murderous inmates to be executed or to be killed or
beat senseless by outraged guards; no inmate has been killed at Marion save by another inmate."
Bruscino v. Carlson, 854 F.2d 162, 166 (7th Cir. 1988).
15

Wolff v. McDonnell,
5

on prison officials' discretion.16 However, in 1995 the Supreme Court decided Sandin v. Conner,17
which held that 30 days of punitive segregation did not call for due process protections because it
did not constitute "atypical and significant hardship . . . in relation to the ordinary incidents of prison
life."18 Sandin overturned both the law of disciplinary due process and liberty interest analysis
generally as applied to prisoners. It has been followed by a large number of decisions holding that
prisoners are not entitled to due process protections in connection with placement in isolated
confinement even for long periods of time.19
In several New York cases prisoners with counsel have cited evidence on the consequences
of isolated confinement in arguing successfully that isolated confinement of sufficient duration is
sufficiently "significant" under Sandin to call for due process protections, irrespective of whether
it violates the Eighth Amendment. In Lee v. Coughlin,20 the court found that 376 days served in
punitive segregation was atypical and significant. It emphasized the atypicality of segregated
confinement of that duration and the harshness of the segregation regime cmpared with the treatment
of prisoners generally in New York. However, it also observed that "[t]he effect of prolonged
isolation on inmates has been repeatedly confirmed in medical and scientific studies."21 In McClary
v. Kelly,22 where the plaintiff complained of being held for four years in administrative segregation
on a pretext, the court held that evidence of psychological harm (both expert evidence and the
16

Hewitt v. Helms, 459 U.S. 460, 468 (1983).

17

515 U.S. 472 (1995).

18

515 U.S. at 484.

19

See, e.g., Jones v. Baker, 155 F.3d 810 (6th Cir. 1998) (holding that two-year placement in
segregation pending investigation did not require due process protections); Bonner v. Parke, 918
F.Supp. 1264, 1269-70 (N.D.Ind. 1996) (holding three years in segregation was not atypical and
significant).
20

26 F.Supp.2d 615 (S.D.N.Y. 1998).

21

26 F.Supp.2d at 637.

22

4 F.Supp.2d 195 (W.D.N.Y. 1998).
6

plaintiff's own testimony) created a triable issue under the Sandin atypical and significant standard.23
Whether such victories have any substantial consequences for prison officials' use of isolated
confinement is debatable. The standard to which prison officials are held in prison disciplinary
hearings is minimal,24 and trumped-up charges do not state a constitutional claim as long as the
procedural rituals are observed.25 The due process standard in connection with administrative
segregation is even less demanding.26

3.

Aggravating circumstances
Isolated confinement is often associated with other forms of abusive treatment, including

physical abuse, deprivation of clothing, deprivation of medical care, unsanitary conditions, lack of
opportunity for personal hygiene, etc. In such instances courts have generally responded by striking
down the extreme practices,27 since they appear to be unnecessary to and readily separable from the
isolation regime itself. The Hutto v. Finney device of restricting the use of segregation itself has
rarely been employed.
23

Id. at 205-07. Mr. McClary was eventually awarded a substantial sum in damages by a jury,
an award that I believe is presently being appealed.
24

See Superintendent v. Hill, 472 U.S. 445, 457 (1988) (holding that disciplinary conviction need
be supported only by "some evidence"); Wolff v. McDonnell, 418 U.S. 539, 563-71 (1974) (holding
due process requires only notice, a written statement of the evidence behind a decision and the
reasons for the punishment imposed, a limited right to call witnesses and present documentary
evidence at a hearing, and in certain cases the assistance of a counsel substitute); People ex rel. Vega
v. Smith, 66 N.Y.2d 130, 495 N.Y.S.2d 332, 485 N.E.2d 997, 1002-04 (N.Y. 1985) (holding that a
staff member's written report alone can be sufficient to support a disciplinary conviction).
25

See, e.g., Freeman v. Rideout, 808 F.2d 949, 951-53 (2d Cir. 1986), cert. denied, 485 U.S. 982
(1988).
26

Hewitt v. Helms, 459 U.S. 460, 472, 474 (1983) (requiring only "an informal nonadversary
review of the information supporting [the prisoner's] administrative confinement"; noting that
decisions may be based on "rumor, reputation, and even more imponderable factors . . . 'purely
subjective evaluations' . . . intuitive judgments.")
27

Thus, in the Pelican Bay litigation, the court held that plaintiffs were entitled to injunctive relief
against excessive force, denial of medical and mental health care, while declining to enjoin the
overall regime of isolation and idleness. Madrid v. Gomez, 889 F.Supp. at 1279-82.
7

In some cases, the aggravating circumstance is an extreme of isolation. The use of solid or
"boxcar" doors that interfere with ventilation and with prisoners' ability to communicate with staff
in emergencies has been held unconstitutional in some cases.28 A particularly interesting decision
is United States v. Koch,29 a criminal case in which the court held that a mere six hours of
confinement in a boxcar cell to obtain confession was unconstitutionally coercive, and suppressed
the confession.30 Other courts have not followed suit. In Tyler v. Black,31 the Eighth Circuit initially
held that the use of boxcar doors by itself did not violate the Eighth Amendment, it concluded that
in the totality of the circumstances (the fact that the doors were closed at all times, prisoners spent
23 hours a day in the cells for several months, and prisoners were double celled).32 On rehearing en
banc, however, the author of the opinion backed down from that holding, stating that double celling
had been ended after the previous opinion and other conditions had changed, mooting the claim.33

28

Hoptowit v. Ray, 682 F.2d 1237, 1257-58 (9th Cir. 1982) (affirming finding that solid doors
that excluded nearly all fresh air and light, limited access to medical care, and caused sanitary
problems violated the Eighth Amendment); LeMaire v. Maass, 745 F.Supp. 623, 636 (D.Or. 1990)
(holding "quiet cells" with steel doors were unconstitutional because they made it impossible to call
for medical attention), vacated and remanded on other grounds, 12 F.3d 1444 (9th Cir. 1993);
Toussaint v. McCarthy, 597 F.Supp. 1388, 1408 (N.D.Cal. 1984) (holding "quiet cells" with closed
solid doors were unconstitutional), aff'd in part and rev'd in part on other grounds, 801 F.2d 1080,
1106-07 (9th Cir. 1986), cert. denied, 481 U.S. 1069 (1987); Bono v. Saxbe, 527 F.Supp. 1187
(S.D.Ill. 1981); Bono v. Saxbe, 450 F.Supp. 934, 946-48 (E.D.Ill. 1978), aff'd in part and remanded
in part on other grounds, 620 F.2d 609 (7th Cir. 1980); Berch v. Stahl, 373 F.Supp. 412, 421
(E.D.N.C. 1974) (limiting solid-door confinement to 15 days).
29

552 F.2d 1216 (7th Cir. 1977).

30

Id. at 1218-19.

31

811 F.2d 424 (8th Cir. 1987), withdrawn, 865 F.2d 181 (8th Cir. 1989).

32

Id. at 434-35.

33

Tyler v. Black, 865 F.2d 181, 183-84 (8th Cir. 1989) (en banc).
8