dissenting.
These petitions ask us to consider whether the Eighth Amendment prohibits as “cruel and unusual punishmen[t3” the execution of prisoners who have spent nearly 20 years or more on death row. Both of these cases involve astonishingly long delays flowing in significant part from constitutionally defective death penalty procedures. Where a delay, measured in decades, reflects the State’s own failure to comply with the Constitution’s demands, the claim that time has rendered the execution inhuman is a particularly strong one. I believe this Court should consider that claim now. See Lackey v. Texas, 514 U.S. 1045 (1995) (Stevens, J., respecting denial of certiorari); Elledge v. Florida, 525 U.S. 944 (1998) (Breyer, J., dissenting from denial of certiorari).
The petitioner in Moore v. Nebraska was sentenced to death on June 20, 1980, more than 19 years ago. By mid-1982, Moore had invoked all his direct appellate remedies and lost. By mid-1984, he had invoked all state collateral remedies and lost. But in 1988 a Federal District Court agreed with Moore that Nebraska’s death sentence procedures violated the Constitution because its standards were too vague, permitting the death penalty’s arbitrary application. See Furman v. Georgia, 408 U. S. 238 (1972) (per curiam). The District Court issued a writ of habeas corpus. The Eighth Circuit affirmed. And in May 1992, this Court denied *994the State’s petition for certiorari, making final the lower federal court decision in Moore’s favor.
In April 1995, after modifying its death sentence procedures and 15 years after Moore’s first sentencing proceeding, the State held a new sentencing proceeding; Moore was again sentenced to death. By April 1997, Moore had invoked all direct appellate remedies and lost. He then invoked state collateral review and lost in the lower courts. See 210 Neb. 457, 316 N. W. 2d 33 (1982). He now seeks certiorari, asking us to review his claim of inordinate delay (among others) — 19 years and 4 months after he was first sentenced to death.
The petitioner in Knight v. Florida was sentenced to death on April 21, 1975, nearly 25 years ago. By mid-1976, Knight had invoked all his direct appellate remedies and lost. By mid-1983, he had invoked all state collateral remedies and lost. But Knight had also filed a petition for habeas corpus in federal court; and in December 1988, the Federal Court of Appeals for the Eleventh Circuit found that Florida’s death penalty sentencing procedure was constitutionally defective because it did not require the jury to take account of an unusually traumatic and abusive childhood as a potentially mitigating factor. See Lockett v. Ohio, 438 U. S. 586 (1978). The Court of Appeals ordered a new sentencing proceeding.
In February 1996, the State held a new proceeding, and Knight was again sentenced to death. In November 1998, the Florida Supreme Court affirmed. 721 So. 2d 287 (1998). Knight now seeks certiorari, asking us to review his claim of inordinate delay (among others) — 24 years and 6 months after he was first sentenced to death.
It is difficult to deny the suffering inherent in a prolonged wait for execution — a matter which courts and individual judges have long recognized. See Lackey, supra, at 1045-1047. More than a century ago, this Court described as “horrible” the “feelings” that accompany uncertainty about whether, or when, the execution will take place. In re Medley, 134 U. S. 160, 172 (1890). The California Supreme Court has referred to the “dehumanizing effects of. .. lengthy imprisonment prior to execution.” People v. Anderson, 6 Cal. 3d 628, 649, 493 P. 2d 880, 894 (1972). In Furman v. Georgia, supra, at 288-289 (concurring opinion), Justice Brennan wrote of the “inevitable long wait” that exacts “a frightful toll.” Justice Frankfurter noted that the “onset of insanity *995while awaiting execution of a death sentence is not a rare phenomenon.” Solesbee v. Balkcom, 339 U. S. 9, 14 (1950) (dissenting opinion). See Strafer, Volunteering for Execution, 74 J. Crim. L. & C. 860, 872, n. 44 (1983) (a study of Florida inmates showed that 35% of those confined on death row attempted suicide; 42% seriously considered suicide). And death row conditions of special isolation may well aggravate that suffering. See Connolly, Better Never Than Late, 23 New England J. on Crim. & Civ. Confinement 101, 121 (1997); Strafer, supra, at 870-871, n. 37.
At the same time, the longer the delay, the weaker the justification for imposing the death penalty in terms of punishment's basic retributive or deterrent purposes: Lackey, supra, at 1046. Nor can one justify lengthy delays by reference to constitutional tradition, for our Constitution was written at a time when delay between sentencing and execution could be measured in days or weeks, not decades. See Pratt v. Attorney General for Jamaica, [1994] 2 A. C. 1, 18, 4 All E. R. 769, 773 (P. C. 1993) (en banc) (Great Britain’s “Murder Act” of 1751 prescribed that execution take place on the next day but one after sentence).
A growing number of courts outside the United States — courts that accept or assume the lawfulness of the death penalty — have held that lengthy delay in administering a lawful death penalty renders ultimate execution inhuman, degrading, or unusually cruel. In Pratt v. Attorney General for Jamaica, supra, for example, the Privy Council considei'ed whether Jamaica lawfully could execute two prisoners held for 14 years after sentencing. The Council noted that Jamaican law authorized the death penalty and that the United Nations Committee on Human Rights has written that “‘capital punishment is not per se unlawful under the [Human Rights] Covenant.’” Id., at 26, 4 All E. R., at 780. But the Privy Council concluded that it was an “inhuman act to keep a man facing the agony of execution over a long extended period of time,” id., at 29, 4 All E. R., at 783, and the delay of 14 years was “shocking,” id., at 33, 4 All E. R., at 786. It held that the delay (and presumptively any delay of more than five years) was “‘inhuman or degrading punishment or other treatment’” forbidden by Jamaica’s Constitution unless “due entirely to the fault of the accused.” Id., at 29, 4 All E. R., at 783.
The Supreme Court of India has held that an appellate court, which itself has authority to sentence, must take account of delay when deciding whether to impose a death penalty. Sher Singh *996v. State of Punjab, A. I. R. 1983 S. C. 465. A condemned prisoner may ask whether it is “just and fair” to permit execution in instances of “[prolonged delay.” Id., at 470-471. The Supreme Court of Zimbabwe, after surveying holdings of many foreign courts, concluded that delays of five and six years were “inordinate” and constituted “‘torture or . . . inhuman or degrading punishment or other such treatment.'” Catholic Commission for Justice and Peace in Zimbabwe v. Attorney-General, [1993] 1 Zimb. L. R. 239, 240, 269 (S) (Aug. 4, 1999), http: //www.law.wits.ac.za/salr/catholic.html. And the European Court of Human Rights, interpreting the European Convention on Human Rights, noted the convention did not forbid capital punishment. But, in the court’s view, the convention nonetheless prohibited the United Kingdom from extraditing a potential defendant to the Commonwealth of Virginia — in large part because the 6- to 8-year delay that typically accompanied a death sentence amounts to “cruel, inhuman, [or] degrading treatment or punishment” forbidden by the convention. Soering v. United Kingdom, 11 Eur. Ct. H. R. (ser. A), pp. 439, 478, ¶111 (1989).
Not all foreign authority reaches the same conclusion. The Supreme Court of Canada, for example, held that Canadian constitutional standards, though roughly similar to those of the European Convention on Human Rights, did not bar extradition to the United States of a defendant facing the death penalty. Kindler v. Minister of Justice, [1991] 2 S. C. R. 779, 838 (joint opinion). And the United Nations Human Rights Committee has written that a delay of 10 years does not necessarily violate roughly similar standards set forth in the Universal Declaration of Human Rights. Views adopted by the United Nations Human Rights Committee, 44th Sess., Mar. 30, 1992, In re: Barrett v. Jamaica (Nos. 270/1988 and 271/1988) §8.4. Given the closeness of the Canadian Court’s decision (4 to 3) and language that the United Nations Human Rights Committee used to describe the 10-year delay (“disturbingly long”), one cannot be certain what position those bodies would take in respect to delays of 19 and 24 years.
Obviously this foreign authority does not bind us. After all, we are interpreting a “Constitution for the United States of America.” Thompson v. Oklahoma, 487 U. S. 815, 868, n. 4 (1988) (Scalia, J., dissenting). And indeed, after Soering, the United States Senate insisted on reservations to language imposing similar standards in various human rights treaties, specifying, for *997example, that the language in question did not “restrict or prohibit the United States from applying the death penalty consistent with the ... Constitution of the United States, including any constitutional period of confinement prior to the imposition of the death penalty.” 136 Cong. Rec. 36192-36199 (1990) (U. S. Senate Resolution of Advice and Consent to Ratification of the Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment).
Nonetheless, the treaty reservations say nothing about whether a particular “period of confinement” is “constitutional.” And this Court has long considered as relevant and informative the way in which foreign courts have applied standards roughly comparable to our own constitutional standards in roughly comparable circumstances. In doing so, the Court has found particularly instructive opinions of former Commonwealth nations insofar as those opinions reflect a legal tradition that also underlies our own Eighth Amendment. Thompson v. Oklahoma, supra, at 830-831 (opinion of Stevens, J.) (considering practices of Anglo-American nations regarding executing juveniles); Enmund v. Florida, 458 U. S. 782, 796-797, n. 22 (1982) (noting that the doctrine of felony murder has been eliminated or restricted in England, India, Canada, and a “number of other Commonwealth countries”); Coker v. Georgia, 433 U. S. 584, 596, n. 10 (1977) (observing that only 3 of 60 nations surveyed in 1965 retained the death penalty for rape); Trop v. Dulles, 356 U. S. 86, 102-103 (1958) (noting that only 2 of 84 countries surveyed imposed denationalization as a penalty for desertion). See also Washington v. Glucksberg, 521 U. S. 702, 710, n. 8, and 718-719, n. 16 (1997) (surveying other nations' laws regarding assisted suicide); Culombe v. Connecticut, 367 U. S. 568, 583-584, n. 25, and 588 (1961) (considering English practice concerning police interrogation of suspects); Kilbourn v. Thompson, 103 U. S. 168, 183-189 (1881) (referring to the practices of Parliament in determining whether the House of Representatives has the power to hold a witness in contempt). Willingness to consider foreign judicial views in comparable cases is not surprising in a Nation that from its birth has given a “decent respect to the opinions of mankind.”
In these cases, the foreign courts I have mentioned have considered roughly comparable questions under roughly comparable legal standards. Each court has held or assumed that those *998standards permit application of the death penalty itself. Consequently, I believe their views are useful even though not binding.
Further, the force of the major countervailing argument is diminished in these two cases. That argument (as set out by the Human Rights Commission) recognizes that there must be an “element of delay between the lawful imposition of a sentence of death and the exhaustion of available remedies.” Barrett, supra, § 8.4. It claims that “even prolonged periods of detention under a severe custodial regime on death row cannot generally be considered to constitute cruel, inhuman or degrading treatment if the convicted person is merely availing himself of appellate remedies.” Ibid. As the Canadian Supreme Court noted, “a defendant is never forced to undergo the full appeal procedure, but the vast majority choose to do so. It would be ironic if delay caused by the appellant’s taking advantage of the full and generous avenue of the appeals available to him should be viewed as a violation of fundamental justice.” Kindler, supra, at 838; see also Richmond v. Lewis, 948 F. 2d 1473, 1491-1492 (CA9 1990).
The eases before us, however, involve delays which resulted in large part from the States’ failure to apply constitutionally sufficient procedures at the time of initial sentencing. They also involve extensive delays of close to two decades or more. Petitioners argue that the state-induced portion of the delay, perhaps up to 12 years in Moore’s ease, up to 15 years in Knight’s, should not be charged against them in any constitutional calculus. Cf. Pratt, 2 A. C., at 29, 4 All E. R., at 783 (counting against the prisoner only that portion of the delay caused by “escape ... or frivolous and time wasting resort to legal procedures”). Twenty years or more could not be necessary to provide a “reasonable time for appeal and consideration of reprieve.” Id., at 33, 4 All E. R., at 786. For these reasons, I think petitioners’ argument cannot be rejected out of hand.
Nor do I agree with Justice Thomas that the lower courts have “resoundingly rejected” petitioners’ claim. Ante, at 992. I have found about two dozen post-1995 lower court cases in which prisoners have raised Lackey claims. Most involve procedural failings that in part or in whole determined the outcome of the case. Of the eight cases (other than the two cases below) that decided Lackey claims solely on the merits, only four involve lengthy delays for which the State arguably bears responsibility. See Bell v. State, 938 S. W. 2d 35 (Tex. Crim. App. 1996) (20 years; *999conviction overturned once); Ex parte Bush, 695 So. 2d 138 (Ala. 1997) (16 years; conviction overturned twice); State v. Smith, 280 Mont. 158, 931 P. 2d 1272 (1996) (13 years; sentence overturned once); People v. Massie, 19 Cal. 4th 550, 967 P. 2d 29 (1998) (16 years; sentence overturned once). Neither the opinions in these four cases, nor those in any other of the lower court cases that I have found, discuss the potential significance of that state responsibility at any length. Thus, although the experiment may have begun, it is hardly evident that we “should consider the experiment concluded.” Ante, at 993.
Finally, the constitutional issue, even if limited to delays of close to 20 years or more, has considerable practical importance. Available statistics indicate that as of two years ago, December 1997, 24 prisoners sentenced to death had been on death row for more than 20 years. At that time 125 prisoners on death row had been sentenced in or before 1980 and therefore may now fall within the relevant category. U. S. Dept, of Justice, Bureau of Justice Statistics Bulletin, Capital Punishment 1997, p. 13 (Dec. 1998). Given these figures and the nature of the question, despite the absence of a division among the lower federal courts, this Court should consider the issue.
I would grant the petitions for certiorari in these two eases.