concurring in the judgment.
I
Given my views in Apprendi v. New Jersey, 530 U. S. 466, 555 (2000) (dissenting opinion), and Harris v. United States, *614ante, at 569 (opinion concurring in part and concurring in judgment), I cannot join the Court’s opinion. I concur in the judgment, however, because I believe that jury sentencing in capital cases is mandated by the Eighth Amendment.
II
This Court has held that the Eighth Amendment requires States to apply special procedural safeguards when they seek the death penalty. Gregg v. Georgia, 428 U. S. 153 (1976). Otherwise, the constitutional prohibition against “cruel and unusual punishments” would forbid its use. Furman v. Georgia, 408 U. S. 238 (1972) (per curiam). JUSTICE Stevens has written that those safeguards include a requirement that a jury impose any sentence of death. Harris v. Alabama, 513 U. S. 504, 515-526 (1995) (dissenting opinion); Spaziano v. Florida, 468 U. S. 447, 467-490 (1984) (Stevens, J., joined by Brennan and Marshall, JJ., concurring in part and dissenting in part). Although I joined the majority in Harris v. Alabama, I have come to agree with the dissenting view, and with the related views of others upon which it in part relies, see Gregg, supra, at 190 (joint opinion of Stewart, Powell, and Stevens, JJ.). Cf. Henslee v. Union Planters Nat. Bank & Trust Co., 335 U. S. 595, 600 (1949) (Frankfurter, J., dissenting) (“Wisdom too often never comes, and so one ought not to reject it merely because it comes late”). I therefore conclude that the Eighth Amendment requires that a jury, not. a judge, make the decision to sentence a defendant to death.
I am convinced by the reasons that Justice Stevens has given. These include (1) his belief that retribution provides the main justification for capital punishment, and (2) his assessment of the jury’s comparative advantage in determining, in a particular case, whether capital punishment will serve that end.
As to the first, I note the continued difficulty of justifying capital punishment in terms of its ability to deter crime, to *615incapacitate offenders, or to rehabilitate criminals. Studies of deterrence are, at most, inconclusive. See, e. g., Sorensen, Wrinkle, Brewer, & Marquart, Capital Punishment and Deterrence: Examining the Effect of Executions on Murder in Texas, 45 Crime & Delinquency 481 (1999) (no evidence of a deterrent effect); Bonner & Fessenden, Absence of Executions: A special report, States With No Death Penalty Share Lower Homicide Rates, N. Y. Times, Sept. 22, 2000, p. A1 (during last 20 years, homicide rate in death penalty States has been 48% to 101% higher than in non-death-penalty States); see also Radelet & Akers, Deterrence and the Death Penalty: The Views of the Experts, 87 J. Crim. L. & C. 1, 8 (1996) (over 80% of criminologists believe existing research fails to support deterrence justification).
As to incapacitation, few offenders sentenced to life without parole (as an alternative to death) commit further crimes. See, e. g., Sorensen & Pilgrim, An Actuarial Risk Assessment of Violence Posed by Capital Murder Defendants, 90 J. Crim. L. & C. 1251,1256 (2000) (studies find average repeat murder rate of .002% among murderers whose death sentences were commuted); Marquart & Sorensen, A National Study of the F^rmcm-Commuted Inmates: Assessing the Threat to Society from Capital Offenders, 23 Loyola (LA) L. Rev. 5, 26 (1989) (98% did not kill again either in prison or in free society). But see Roberts v. Louisiana, 428 U. S. 325, 354 (1976) (White, J., dissenting) (“[D]eath finally forecloses the possibility that a prisoner will commit further crimes, whereas life imprisonment does not”). And rehabilitation, obviously, is beside the point.
In respect to retribution, jurors possess an important comparative advantage over judges. In principle, they are more attuned to “the community’s moral sensibility,” Spaziano, 468 U. S., at 481 (Stevens, J., concurring in part and dissenting in part), because they “reflect more accurately the composition and experiences of the community as a whole,” id., at 486. Hence they are more likely to “express the con*616science of the community on the ultimate question of life or death,” Witherspoon v. Illinois, 391 U. S. 510, 519 (1968), and better able to determine in the particular case the need for retribution, namely, “an expression of the community’s belief that certain crimes are themselves so grievous an affront to humanity that the only adequate response may be the penalty of death.” Gregg, supra, at 184 (joint opinion of Stewart, Powell, and Stevens, JJ.).
Nor is the fact that some judges are democratically elected likely to change the jury’s comparative advantage in this respect. Even in jurisdictions where judges are selected directly by the people, the jury remains uniquely capable of determining whether, given the community’s views, capital punishment is appropriate in the particular case at hand. See Harris, 513 U. S., at 518-519 (Stevens, J., dissenting); see also J. Liebman et al., A Broken System, Part II: Why There Is So Much Error in Capital Cases, and What Can Be Done About It 405-406 (Feb. 11,2002) (hereinafter A Broken System) (finding that judges who override jury verdicts for life are especially likely to commit serious errors); cf. Epstein & King, The Rules of Inference, 69 U. Chi. L. Rev. 1 (2002) (noting dangers in much scholarly research but generally approving of Liebman).
The importance of trying to translate a community’s sense of capital punishment’s appropriateness in a particular case is underscored by the continued division of opinion as to whether capital punishment is in all circumstances, as currently administered, “cruel and unusual.” Those who make this claim point, among other things, to the fact that death is not reversible, and to death sentences imposed upon those whose convictions proved unreliable. See, e. g., Weinstein, The Nation’s Death Penalty Foes Mark a Milestone Crime: Arizona convict freed on DNA tests is said to be the 100th known condemned U. S. prisoner to be exonerated since executions resumed, Los Angeles Times, Apr. 10, 2002, p. A16; G. Ryan, Governor of Illinois, Report of Governor’s Commis*617sion on Capital Punishment 7-10 (Apr. 15, 2002) (imposing moratorium on Illinois executions because, post-Fwrraan, 13 people have been exonerated and 12 executed); see generally Bedau & Radelet, Miscarriages of Justice in Potentially Capital Cases, 40 Stan. L. Rev. 21, 27 (1987).
They point to the potentially arbitrary application of the death penalty, adding that the race of the victim and socioeconomic factors seem to matter. See, e. g., U. S. General Accounting Office, Report to Senate and House Committees on the Judiciary: Death Penalty Sentencing 5 (Feb. 1990) (synthesis of 28 studies shows “pattern of evidence indicating racial disparities in the charging, sentencing, and imposition of the death penalty”); Baldus, Woodworth, Zuekerman, Weiner, & Broffitt, Racial Discrimination and the Death Penalty in the Post-Furman Era: An Empirical and Legal Overview, With Recent Findings from Philadelphia, 83 Cornell L. Rev. 1638, 1661 (1998) (evidence of race-of-victim disparities in 90% of States studied and of race-of-defendant disparities in 55%); McCleskey v. Kemp, 481 U. S. 279, 320-345 (1987) (Brennan, J., dissenting); see also, e.g., D. Baldus, G. Woodworth, G. Young, & A. Christ, The Disposition of Nebraska Capital and Non-Capital Homicide Cases (1973— 1999): A Legal and Empirical Analysis 95-100 (Oct. 10, 2001) (death sentences almost five times more likely when victim is of a high soeio-ecónomic status).
They argue that the delays that increasingly accompany sentences of death make those sentences unconstitutional because of “the suffering inherent in a prolonged wait for execution.” Knight v. Florida, 528 U. S. 990, 994 (1999) (Breyer, J., dissenting from denial of certiorari) (arguing that the Court should consider the question); see, e.g., Lackey v. Texas, 514 U. S. 1045 (1995) (Stevens, J., respecting denial of certiorari); Bureau of Justice Statistics, Capital Punishment 2000, pp. 12, 14 (rev. 2002) (average delay is 12 years, with 52 people waiting more than 20 years and some more than 25).
*618They point to the inadequacy of representation in capital cases, a fact that aggravates the other failings. See, e. g., Bright, Counsel for the Poor: The Death Sentence Not for the Worst Crime but for the Worst Lawyer, 103 Yale L. J. 1835 (1994) (describing many studies discussing deficient capital representation).
And they note that other nations have increasingly abandoned capital punishment. See, e. g., San Martin, U. S. Taken to Task Over Death Penalty, Miami Herald, May 31, 2001, p. 1 (United States is only Western industrialized Nation that authorizes the death penalty); Amnesty International Website Against the Death Penalty, Facts and Figures on the Death Penalty (2002), http://www.web.amnesty.org/ rmp/dplibrary.nsf (since Gregg, 111 countries have either abandoned the penalty altogether, reserved it only for exceptional crimes like wartime crimes, or not carried out executions for at least the past 10 years); DeYoung, Group Criticizes U. S. on Detainee Policy; Amnesty Warns of Human Rights Fallout, Washington Post, May 28, 2002, p. A4 (the United States rates fourth in number of executions, after China, Iran, and Saudi Arabia).
Many communities may have accepted some or all of these claims, for they do not impose capital sentences. See A Broken System, App. B, Table 11A (more than two-thirds of American counties have never imposed the death penalty since Gregg (2,064 out of 3,066), and only 3% of the Nation’s counties account for 50% of the Nation’s death sentences (92 out of 3,066)). Leaving questions of arbitrariness aside, this diversity argues strongly for procedures that will help assure that, in a particular case, the community indeed believes application of the death penalty is appropriate, not “cruel,” “unusual,” or otherwise unwarranted.
For these reasons, the danger of unwarranted imposition of the penalty cannot be avoided unless “the decision to impose the death penalty is made by a jury rather than by a single governmental official.” Spaziano, 468 U. S., at 469 *619(Stevens, J., concurring in part and dissenting in part); see Solem v. Helm, 463 U. S. 277, 284 (1983) (Eighth Amendment prohibits excessive or disproportionate punishment). And I conclude that the Eighth Amendment requires individual jurors to make, and to take responsibility for, a decision to sentence a person to death.