Wednesday, 16 April 2008

Ginsburg and Souter dissent

GINSBURG, J., dissenting
No. 07–5439


[April 16, 2008]

It is undisputed that the second and third drugs used in
Kentucky’s three-drug lethal injection protocol, pancuronium
bromide and potassium chloride, would cause aconscious inmate to suffer excruciating pain. Pancuronium
bromide paralyzes the lung muscles and resultsin slow asphyxiation. App. 435, 437, 625. Potassium
chloride causes burning and intense pain as it circulates
throughout the body. Id., at 348, 427, 444, 600, 626. Use
of pancuronium bromide and potassium chloride on a
conscious inmate, the plurality recognizes, would be “constitutionally
unacceptable.” Ante, at 14.
The constitutionality of Kentucky’s protocol thereforeturns on whether inmates are adequately anesthetized bythe first drug in the protocol, sodium thiopental. Kentucky’s
system is constitutional, the plurality states, because
“petitioners have not shown that the risk of an
inadequate dose of the first drug is substantial.” Ante, at
15. I would not dispose of the case so swiftly given thecharacter of the risk at stake. Kentucky’s protocol lacks
basic safeguards used by other States to confirm that aninmate is unconscious before injection of the second and

GINSBURG, J., dissenting
third drugs. I would vacate and remand with instructions
to consider whether Kentucky’s omission of those safeguards
poses an untoward, readily avoidable risk of inflicting
severe and unnecessary pain.
The Court has considered the constitutionality of aspecific method of execution on only three prior occasions.
Those cases, and other decisions cited by the parties and
amici, provide little guidance on the standard that should
govern petitioners’ challenge to Kentucky’s lethal injectionprotocol.
In Wilkerson v. Utah, 99 U. S. 130 (1879), the Court heldthat death by firing squad did not rank among the “crueland unusual punishments” banned by the EighthAmendment. In so ruling, the Court did not endeavor “to
define with exactness the extent of the constitutional
provision which provides that cruel and unusual punishments
shall not be inflicted.” Id., at 135–136. But it was
“safe to affirm,” the Court stated, that “punishments of
torture . . . , and all others in the same line of unnecessary
cruelty, are forbidden.” Id., at 136.
Next, in In re Kemmler, 136 U. S. 436 (1890), death by
electrocution was the assailed method of execution.1 The
Court reiterated that the Eighth Amendment prohibits“torture” and “lingering death.” Id., at 447. The word
“cruel,” the Court further observed, “implies . . . somethinginhuman . . . something more than the mere extinguishment
of life.” Ibid. Those statements, however, were
made en passant. Kemmler’s actual holding was that theEighth Amendment does not apply to the States, id., at
1Hanging was the State’s prior mode of execution. Electrocution,
considered “less barbarous,” indeed “the most humane” way to administer
the death penalty, was believed at the time to “result in instantaneous,
and consequently in painless, death.” In re Kemmler, 136 U. S.
436, 443–444 (1890) (internal quotation marks omitted).

Cite as: 553 U. S. ____ (2008)
GINSBURG, J., dissenting
448–449,2 a proposition we have since repudiated, see, e.g.,
Robinson v. California, 370 U. S. 660 (1962).
Finally, in Louisiana ex rel. Francis v. Resweber, 329
U. S. 459 (1947), the Court rejected Eighth and Fourteenth
Amendment challenges to a reelectrocution following
an earlier attempt that failed to cause death. The
plurality opinion in that case first stated: “The traditional
humanity of modern Anglo-American law forbids the
infliction of unnecessary pain in the execution of the death
sentence.” Id., at 463. But the very next sentence variedthe formulation; it referred to the “[p]rohibition against
the wanton infliction of pain.” Ibid.
No clear standard for determining the constitutionality
of a method of execution emerges from these decisions.
Moreover, the age of the opinions limits their utility as an
aid to resolution of the present controversy. The EighthAmendment, we have held, “‘must draw its meaning from
the evolving standards of decency that mark the progressof a maturing society.’” Atkins v. Virginia, 536 U. S. 304,
311–312 (2002) (quoting Trop v. Dulles, 356 U. S. 86, 101
(1958) (plurality opinion)). Wilkerson was decided 129
years ago, Kemmler 118 years ago, and Resweber 61 years
ago. Whatever little light our prior method-of-executioncases might shed is thus dimmed by the passage of time.
Further phrases and tests can be drawn from morerecent decisions, for example, Gregg v. Georgia, 428 U. S.
153 (1976). Speaking of capital punishment in the abstract,
the lead opinion said that the Eighth Amendmentprohibits “the unnecessary and wanton infliction of pain,”
id., at 173 (joint opinion of Stewart, Powell, and STEVENS,
JJ.); the same opinion also cautioned that a death sen——————

2The Court also ruled in Kemmler that the State’s election to carry
out the death penalty by electrocution in lieu of hanging encounteredno Fourteenth Amendment shoal: No privilege or immunity of United
States citizenship was entailed, nor did the Court discern any deprivation
of due process. Id., at 448–449.

GINSBURG, J., dissenting
tence cannot “be imposed under sentencing procedures
that creat[e] a substantial risk that it would be inflicted inan arbitrary and capricious manner,” id., at 188.
Relying on Gregg and our earlier decisions, the Kentucky
Supreme Court stated that an execution procedure
violates the Eighth Amendment if it “creates a substantial
risk of wanton and unnecessary infliction of pain, tortureor lingering death.” 217 S. W. 3d 207, 209, 210 (2006).
Petitioners respond that courts should consider “(a) the
severity of pain risked, (b) the likelihood of that pain
occurring, and (c) the extent to which alternative means
are feasible.” Brief for Petitioners 38 (emphasis added).
The plurality settles somewhere in between, requiring a
“substantial risk of serious harm” and consideringwhether a “feasible, readily implemented” alternative can
“significantly reduce” that risk. Ante, at 13 (internalquotation marks omitted).
I agree with petitioners and the plurality that the degree
of risk, magnitude of pain, and availability of alternatives
must be considered. I part ways with the plurality,
however, to the extent its “substantial risk” test sets a
fixed threshold for the first factor. The three factors are
interrelated; a strong showing on one reduces the importance
of the others.
Lethal injection as a mode of execution can be expected,
in most instances, to result in painless death. Rare
though errors may be, the consequences of a mistake
about the condemned inmate’s consciousness are horrendous
and effectively undetectable after injection of thesecond drug. Given the opposing tugs of the degree of riskand magnitude of pain, the critical question here, as I see
it, is whether a feasible alternative exists. Proof of “a
slightly or marginally safer alternative” is, as the plurality
notes, insufficient. Ante, at 12. But if readily availablemeasures can materially increase the likelihood that theprotocol will cause no pain, a State fails to adhere to con

Cite as: 553 U. S. ____ (2008)
GINSBURG, J., dissenting
temporary standards of decency if it declines to employ
those measures.
Kentucky’s Legislature adopted lethal injection as amethod of execution in 1998. See 1998 Ky. Acts ch. 220, p.
777, Ky. Rev. Stat. Ann. §431.220(1)(a) (West 2006).
Lawmakers left the development of the lethal injection
protocol to officials in the Department of Corrections.
Those officials, the trial court found, were “given the taskwithout the benefit of scientific aid or policy oversight.”
App. 768. “Kentucky’s protocol,” that court observed, “was
copied from other states and accepted without challenge.”
Ibid. Kentucky “did not conduct any independent scientific
or medical studies or consult any medical professionals
concerning the drugs and dosage amounts to be injected
into the condemned.” Id., at 760. Instead, the trial
court noted, Kentucky followed the path taken in other
States that “simply fell in line” behind the three-drug
protocol first developed by Oklahoma in 1977. Id., at 756.
See also ante, at 4, n. 1 (plurality opinion).
Kentucky’s protocol begins with a careful measure: Only
medical professionals may perform the venipunctures andestablish intravenous (IV) access. Members of the IV
team must have at least one year’s experience as a certified
medical assistant, phlebotomist, emergency medical
technician (EMT), paramedic, or military corpsman. App.
984; ante, at 16 (plurality opinion). Kentucky’s IV team
currently has two members: a phlebotomist with 8 years’
experience and an EMT with 20 years’ experience. App.
273–274. Both members practice siting catheters at ten
lethal injection training sessions held annually. Id., at
Other than using qualified and trained personnel to
establish IV access, however, Kentucky does little to ensure
that the inmate receives an effective dose of sodium

GINSBURG, J., dissenting
thiopental. After siting the catheters, the IV team leavesthe execution chamber. Id., at 977. From that point
forward, only the warden and deputy warden remain withthe inmate. Id., at 276. Neither the warden nor the deputy
warden has any medical training.
The warden relies on visual observation to determine
whether the inmate “appears” unconscious. Id., at 978. In
Kentucky’s only previous execution by lethal injection, the
warden’s position allowed him to see the inmate best from
the waist down, with only a peripheral view of the inmate’s
face. See id., at 213–214. No other check for consciousness
occurs before injection of pancuronium bromide.
Kentucky’s protocol does not include an automaticpause in the “rapid flow” of the drugs, id., at 978, or any ofthe most basic tests to determine whether the sodium
thiopental has worked. No one calls the inmate’s name,
shakes him, brushes his eyelashes to test for a reflex, or
applies a noxious stimulus to gauge his response.
Nor does Kentucky monitor the effectiveness of the
sodium thiopental using readily available equipment, even
though the inmate is already connected to an electrocardiogram
(EKG), id., at 976. A drop in blood pressure orheart rate after injection of sodium thiopental would not
prove that the inmate is unconscious, see id., at 579–580;
ante, at 20–21 (plurality opinion), but would signal thatthe drug has entered the inmate’s bloodstream, see App.
424, 498, 578, 580; 8 Tr. 1099 (May 2, 2005). Kentucky’sown expert testified that the sodium thiopental should
“cause the inmate’s blood pressure to become very, verylow,” App. 578, and that a precipitous drop in blood pressure
would “confir[m]” that the drug was having its expected
effect, id., at 580. Use of a blood pressure cuff and
EKG, the record shows, is the standard of care in surgeryrequiring anesthesia. Id., at 539.3
3The plurality deems medical standards irrelevant in part because

Cite as: 553 U. S. ____ (2008) 7
GINSBURG, J., dissenting
A consciousness check supplementing the warden’svisual observation before injection of the second drug iseasily implemented and can reduce a risk of dreadful pain.
Pancuronium bromide is a powerful paralytic that prevents
all voluntary muscle movement. Once it is injected,
further monitoring of the inmate’s consciousness becomes
impractical without sophisticated equipment and training.
Even if the inmate were conscious and in excruciatingpain, there would be no visible indication.4
Recognizing the importance of a window between thefirst and second drugs, other States have adopted safeguards
not contained in Kentucky’s protocol. See Brief for
Criminal Justice Legal Foundation as Amicus Curiae 19–
23.5 Florida pauses between injection of the first andsecond drugs so the warden can “determine, after consultation,
that the inmate is indeed unconscious.” Light-
bourne v. McCollum, 969 So. 2d 326, 346 (Fla. 2007) (per
curiam) (internal quotation marks omitted). The warden
“drawn from a different context.” Ante, at 21. Medical professionals
monitor blood pressure and heart rate, however, not just to save lives,
but also to reduce the risk of consciousness during otherwise painful
procedures. Considering that the constitutionality of Kentucky’sprotocol depends on guarding against the same risk, see supra, at 1;
ante, at 14–15 (plurality opinion), the plurality’s reluctance to considermedical practice is puzzling. No one is advocating the wholesaleincorporation of medical standards into the Eighth Amendment. But
Kentucky could easily monitor the inmate’s blood pressure and heart
rate without physician involvement. That medical professionals
consider such monitoring important enough to make it the standard ofcare in medical practice, I remain persuaded, is highly instructive.
4Petitioners’ expert testified that a layperson could not tell from visual
observation if a paralyzed inmate was conscious and that doing sowould be difficult even for a professional. App. 418. Kentucky’s warden
candidly admitted: “I honestly don’t know what you’d look for.” Id.,
at 283.
5Because most death-penalty States keep their protocols secret, acomprehensive survey of other States’ practices is not available. See
Brief for American Civil Liberties Union et al. as Amici Curiae 6–12.

GINSBURG, J., dissenting
does so by touching the inmate’s eyelashes, calling hisname, and shaking him. Id., at 347.6 If the inmate’s
consciousness remains in doubt in Florida, “the medical
team members will come out from the chemical room and
consult in the assessment of the inmate.” Ibid. Duringthe entire execution, the person who inserted the IV linemonitors the IV access point and the inmate’s face onclosed circuit television. Ibid.
In Missouri, “medical personnel must examine theprisoner physically to confirm that he is unconscious using
standard clinical techniques and must inspect the catheter
site again.” Taylor v. Crawford, 487 F. 3d 1072, 1083
(CA8 2007). “The second and third chemicals are injected
only after confirmation that the prisoner is unconscious
and after a period of at least three minutes has elapsed
from the first injection of thiopental.” Ibid.
In California, a member of the IV team brushes
the inmate’s eyelashes, speaks to him, and shakes himat the halfway point and, again, at the completion of
the sodium thiopental injection. See State of Califor-
nia, San Quentin Operational Procedure No. 0–770, Execution
by Lethal Injection, §V(S)(4)(e) (2007), online at
In Alabama, a member of the execution team “begin[s]
by saying the condemned inmate’s name. If there is no
6Florida’s expert in Lightbourne v. McCollum, 969 So. 2d 326 (Fla.
2007) (per curiam), who also served as Kentucky’s expert in this case,
testified that the eyelash test is “probably the most common
first assessment that we use in the operating room to determine . . .
when a patient might have crossed the line from being conscious
to unconscious.” 4 Tr. in Florida v. Lightbourne, No. 81–170–CF
(Fla. Cir. Ct., Marion Cty.), p. 511, online at
LightbourneRecord.pdf (all Internet materials as visited Apr. 14, 2008,
and in Clerk of Court’s case file). “A conscious person, if you touch theireyelashes very lightly, will blink; an unconscious person typically will
not.” Ibid. The shaking and name-calling tests, he further testified,
are similar to those taught in basic life support courses. See id., at 512.

Cite as: 553 U. S. ____ (2008)
GINSBURG, J., dissenting
response, the team member will gently stroke the condemned
inmate’s eyelashes. If there is no response, theteam member will then pinch the condemned inmate’sarm.” Respondents’ Opposition to Callahan’s Application
for a Stay of Execution in Callahan v. Allen, O. T. 2007,
No. 07A630, p. 3 (internal quotation marks omitted).
In Indiana, officials inspect the injection site after administration
of sodium thiopental, say the inmate’s name,
touch him, and use ammonia tablets to test his response to
a noxious nasal stimulus. See Tr. of Preliminary Injunction
Hearing in 1:06–cv–1859 (SD Ind.), pp. 199–200,
online at
LethalInjection/Public/MoralesTaylorAmicus/20.pdf (hereinafter
Timberlake Hearing).7
These checks provide a degree of assurance—missing
from Kentucky’s protocol—that the first drug has beenproperly administered. They are simple and essentially
costless to employ, yet work to lower the risk that theinmate will be subjected to the agony of conscious suffocation
caused by pancuronium bromide and the searing paincaused by potassium chloride. The record contains no
explanation why Kentucky does not take any of these
elementary measures.
The risk that an error administering sodium thiopental
would go undetected is minimal, Kentucky urges, becauseif the drug was mistakenly injected into the inmate’stissue, not a vein, he “would be awake and screaming.”
Tr. of Oral Arg. 30–31. See also Brief for Respondents 42;
Brief for State of Texas et al. as Amici Curiae 26–27. That
argument ignores aspects of Kentucky’s protocol that
render passive reliance on obvious signs of consciousness,
such as screaming, inadequate to determine whether the
inmate is experiencing pain.
7In Indiana, a physician also examines the inmate after injection of
the first drug. Timberlake Hearing 199.

GINSBURG, J., dissenting
First, Kentucky’s use of pancuronium bromide to paralyze
the inmate means he will not be able to scream after
the second drug is injected, no matter how much pain he is
experiencing. Kentucky’s argument, therefore, appears torest on the assertion that sodium thiopental is itself painful
when injected into tissue rather than a vein. See App.
601. The trial court made no finding on that point, and
Kentucky cites no supporting evidence from executions inwhich it is known that sodium thiopental was injected intothe inmate’s soft tissue. See, e.g., Lightbourne, 969 So. 2d,
at 344 (describing execution of Angel Diaz).
Second, the inmate may receive enough sodium thiopental
to mask the most obvious signs of consciousness without
receiving a dose sufficient to achieve a surgical plane
of anesthesia. See 7 Tr. 976 (Apr. 21, 2005). If the drug isinjected too quickly, the increase in blood pressure cancause the inmate’s veins to burst after a small amount of
sodium thiopental has been administered. Cf. App. 217
(describing risk of “blowout”). Kentucky’s protocol does
not specify the rate at which sodium thiopental should beinjected. The executioner, who does not have any medicaltraining, pushes the drug “by feel” through five feet of
tubing. Id., at 284, 286–287.8 In practice sessions, unlike
in an actual execution, there is no resistance on the catheter,
see id., at 285; thus the executioner’s training maylead him to push the drugs too fast.
“The easiest and most obvious way to ensure that an
inmate is unconscious during an execution,” petitionersargued to the Kentucky Supreme Court, “is to check forconsciousness prior to injecting pancuronium [bromide].”
Brief for Appellants in No. 2005–SC–00543, p. 41. See
8The length of the tubing contributes to the risk that the inmate willreceive an inadequate dose of sodium thiopental. The warden and
deputy warden watch for obvious leaks in the execution chamber, see
ante, at 6 (plurality opinion), but the line also snakes into the neighboring
control room through a small hole in the wall, App. 280.

Cite as: 553 U. S. ____ (2008)
GINSBURG, J., dissenting
also App. 30 (Complaint) (alleging Kentucky’s protocol
does not “require the execution team to determine that the
condemned inmate is unconscious prior to administering
the second and third chemicals”). The court did not address
petitioners’ argument. I would therefore remand
with instructions to consider whether the failure to include
readily available safeguards to confirm that the
inmate is unconscious after injection of sodium thiopental,
in combination with the other elements of Kentucky’sprotocol, creates an untoward, readily avoidable risk of
inflicting severe and unnecessary pain.

No comments: