dissenting.*
In Furman v. Georgia, 408 U. S. 238, 314 (1972) (concurring opinion), I set forth at some length my views on the basic issue presented to the Court in these cases. The death penalty, I concluded, is a cruel and unusual punishment prohibited by the Eighth and Fourteenth Amendments. That continues to be my view.
I have no intention of retracing the “long and tedious journey,” id., at 370, that led to my conclusion in Fur-man. My sole purposes here are to consider the suggestion that my conclusion in Furman has been undercut by developments since then, and briefly to evaluate the basis for my Brethren’s holding that the extinction of life is a permissible form of punishment under the Cruel and Unusual Punishments Clause.
In Furman I concluded that the death penalty is constitutionally invalid for two reasons. First, the death penalty is excessive. Id., at 331-332; 342-359. And *232second, the American people, fully informed as to the purposes of the death penalty and its liabilities, would in my view reject it as morally unacceptable. Id., at 360-369.
Since the decision in Furman, the legislatures of 35 States have enacted new statutes authorizing the imposition of the death sentence for certain crimes, and Congress has enacted a law providing the death penalty for air piracy resulting in death. 49 U. S. C. §§ 1472 (i), (n) (1970 ed., Supp. IV). I would be less than candid if I did not acknowledge that these developments have a significant bearing on a realistic assessment of the moral acceptability of the death penalty to the American people. But if the constitutionality of the death penalty turns, as I have urged, on the opinion of an informed citizenry, then even the enactment of new death statutes cannot be viewed as conclusive. In Furman, I observed that the American people are largely unaware of the information critical to a judgment on the morality of the death penalty, and concluded that if they were better informed they would consider it shocking, unjust, and unacceptable. 408 U. S., at 360-369. A recent study, conducted after the enactment of the post-Furman statutes, has confirmed that the American people know little about the death penalty, and that the opinions of an informed public would differ significantly from those of a public unaware of the consequences and effects of the death penalty.1
Even assuming, however, that the post -Furman enactment of statutes authorizing the death penalty renders the prediction of the views of an informed citizenry an *233uncertain basis for a constitutional decision, the enactment of those statutes has no bearing whatsoever on the conclusion that the death penalty is unconstitutional because it is excessive. An excessive penalty is invalid under the Cruel and Unusual Punishments Clause “even though popular sentiment may favor” it. Id., at 331; ante, at 173, 182-183 (opinion of Stewart, Powell, and Stevens, JJ.); Roberts v. Louisiana, post, at 353-354 (White, ' J., dissenting). The inquiry here, then, is simply whether the death penalty is necessary to accomplish the legitimate legislative purposes in punishment, or whether a less severe penalty — life imprisonment — would do as well. Furman, supra, at 342 (Marshall, J., concurring).
The two purposes that sustain the death penalty as nonexcessive in the Court’s view are general deterrence and retribution. In Furman, I canvassed the relevant data on the deterrent effect of capital punishment. 408 U. S., at 347-354.2 The state of knowledge at that point, after literally centuries of debate, was summarized as follows by a United Nations Committee:
“It is generally agreed between the retentionists and abolitionists, whatever their opinions about the validity of comparative studies of deterrence, that the data which now exist show no correlation between the existence of capital punishment and lower rates of capital crime.” 3
The available evidence, I concluded in Furman, was convincing that “capital punishment is not necessary as a deterrent to crime in our society.” Id., at 353.
The Solicitor General in his amicus brief in these cases *234relies heavily on a study by Isaac Ehrlich,4 reported a year after Furman, to support the contention that the death penalty does deter murder. Since the Ehrlich study was not available at the time of Furman and since it is the first scientific study to suggest that the death penalty may have a deterrent effect, I will briefly consider its import.
The Ehrlich study focused on the relationship in the Nation as a whole between the homicide rate and “execution risk” — the fraction of persons convicted of murder who were actually executed. Comparing the differences in homicide rate and execution risk for the years 1933 to 1969, Ehrlich found that increases in execution risk were associated with increases in the homicide rate.5 But when he employed the statistical technique of multiple regression analysis to control for the influence of other variables posited to have an impact on the homicide rate,6 Ehrlich found a negative correlation between changes in the homicide rate and changes in execution risk. His tentative conclusion was that for the period from 1933 to 1967 each additional execution in the United States might have saved eight lives.7
The methods and conclusions of the Ehrlich study *235have been severely criticized on a number of grounds.8 It has been suggested, for example, that the study is defective because it compares execution and homicide rates on a nationwide, rather than a state-by-state, basis. The aggregation of data from all States — including those that have abolished the death penalty — obscures the relationship between murder and execution rates. Under Ehrlich's methodology, a decrease in the execution risk in one State combined with an increase in the murder rate in another State would, all other things being equal, suggest a deterrent effect that quite obviously would not exist. Indeed, a deterrent effect would be suggested if, once again all other things being equal, one State abolished the death penalty and experienced no change in the murder rate, while another State experienced an increase in the murder rate.9
The most compelling criticism of the Ehrlich study is *236that its conclusions are extremely sensitive to the choice of the time period included in the regression analysis. Analysis of Ehrlich’s data reveals that all empirical support for the deterrent effect of capital'punishment disappears when the five most recent years are removed from his time series — that is to say, whether a decrease in the execution risk corresponds to an increase or a-decrease in the murder rate depends on the ending point of the sample period.10 This finding has east severe doubts on the reliability of Ehrlich’s tentative conclusions.11 Indeed, a recent regression study, based on Ehrlich’s theoretical model but using cross-section state data for the years 1950 and 1960, found no support for the conclusion that executions act as a deterrent.12
The Ehrlich study, in short, is of little, if any, assistance in assessing the deterrent impact of the death penalty. Accord, Commonwealth v. O’Neal, -Mass. -, -, 339 N. E. 2d 676, 684 (1975). The evidence I reviewed in Furman13 remains convincing, in my view, that “capital punishment is not necessary as a deterrent to crime in our society.” 408 U. S., at 353. The justification for the death penalty must be found elsewhere.
The other principal purpose said to be served by the death penalty is retribution.14 The notion that retribu*237tion can serve as a moral justification for the sanction of death finds credence in the opinion of my Brothers Stewart, Powell, and Stevens, and that of my Brother White in Roberts v. Louisiana, post, p. 337. See also Furman v. Georgia, 408 U. S., at 394-395 (Burger, C. J., dissenting). It is this notion that I find to be the most disturbing aspect of today’s unfortunate decisions.
The concept of retribution is a multifaceted one, and any discussion of its role in the criminal law must be undertaken with caution. On one level, it can be said that the notion of retribution or reprobation is the basis of our insistence that only those who have broken the law be punished, and in this sense the notion is quite obviously central to a just system of criminal sanctions. But our recognition that retribution plays a crucial role in determining who may be punished by no means requires approval of retribution as a general justification for punishment.15 It is the question whether retribution can provide a moral justification for punishment — in particular, capital punishment — that we must consider.
My Brothers Stewart, Powell, and Stevens offer the following explanation of the retributive justification for capital punishment:
“ ‘The instinct for retribution is part of the nature of man, and channeling that instinct in the administration of criminal justice serves an important purpose in promoting the stability of a society governed *238by law. When people begin to believe that organized society is unwilling or unable to impose upon criminal offenders the punishment they 'deserve/ then there are sown the seeds of anarchy — of self-help, vigilante justice, and lynch law.’ ” Ante, at 183, quoting from Furman v. Georgia, supra, at 308 (Stewart, J., concurring).
This statement is wholly inadequate to justify the death penalty. As my Brother Brennan stated in Furman, “[t]here is no evidence whatever that utilization of imprisonment rather than death encourages private blood feuds and other disorders.” 408 U. S., at 303 (concurring opinion).16 It simply defies belief to suggest that the death penalty is necessary to prevent the American people from taking the law into their own. hands.
In a related vein, it may be suggested that the expression of moral outrage through the imposition of the death penalty serves to reinforce basic moral values— that it marks some crimes as particularly offensive and therefore to be avoided. The argument is akin to a deterrence argument, but differs in that it contemplates the individual’s shrinking from antisocial conduct, not because he fears punishment, but because he has been told in the strongest possible way that the conduct is wrong. This contention, like the previous one, provides no support for the death penalty. It is inconceivable that any individual concerned about conforming his conduct to what society says is “right” would fail to realize that murder is “wrong” if the penalty were simply life imprisonment.
The foregoing contentions — that society’s expression of moral outrage through the imposition of the death penalty pre-empts the citizenry from taking the law into its *239own hands and reinforces moral values — are not retributive in the purest sense. They are essentially utilitarian in that they portray the death penalty as valuable because of its beneficial results. These justifications for the death penalty are inadequate because the penalty is, quite clearly I think, not necessary to the accomplishment of those results.
There remains for consideration, however, what might be terme'd the purely retributive justification for the death penalty — that the death penalty is appropriate, not because of its beneficial effect on society, but because the taking of the murderer’s life is itself morally good.17 Some of the language of the opinion of my Brothers Stewart, Powell, and Stevens in No. 74-6257 appears positively to embrace this notion of retribution for its own sake as a justification for capital punishment.18 They state:
“[T]he decision that capital punishment may be the appropriate sanction in extreme cases is 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.” Ante, at 184 (footnote omitted).
*240They then quote with approval from Lord Justice Denning’s remarks before the British Royal Commission on Capital Punishment:
“ 'The truth is that some crimes are so outrageous that society insists on adequate punishment, because the wrong-doer deserves it, irrespective of whether it is a deterrent or not.’ ” Ante, at 184 n. 30.
Of course, it may be that these statements are intended as no more than observations as to the popular demands that it is thought must be responded to in order to prevent anarchy. But the implication of the statements appears to me to be quite different — namely, that society’s judgment that the murderer “deserves” death must be respected not simply because the preservation of order requires it, but because it is appropriate that society make the judgment and carry it out. It is this latter notion, in particular, that I consider to be fundamentally at odds with the Eighth Amendment. See Furman v. Georgia, 408 U. S., at 343-345 (Marshall, J., concurring). The mere fact that the community demands the murderer’s life in return for the evil he has done cannot sustain the death penalty, for as Justices Stewart, Powell, and Stevens remind us, “the Eighth Amendment demands more than that a challenged punishment be acceptable to contemporary society.” Ante, at 182. To be sustained under the Eighth Amendment, the death penalty must “eompor[t] with the basic concept of human dignity at the core of the Amendment,” ibid.; the objective in imposing it must be “[consistent] with our respect for the dignity of [other] men.” Ante, at 183. See Trop v. Dulles, 356 U. S. 86, 100 (1958) (plurality opinion). Under these standards, the taking of life “because the wrongdoer deserves it” surely must *241fall, for such a punishment has as its very basis the total denial of the wrongdoer’s dignity and worth.19
The death penalty, unnecessary to promote the goal of deterrence or to further any legitimate notion of retribution, is an excessive penalty forbidden by the Eighth and Fourteenth Amendments. I respectfully dissent from the Court’s judgment upholding the sentences of death imposed upon the petitioners in these cases.
[This opinion applies also to No. 75-5706, Proffitt v. Florida, post, p. 242, and No. 75-5394, Jurek v. Texas, post, p. 262.]
Sarat & Vidmar, Public Opinion, The Death Penalty, and the Eighth Amendment: Testing the Marshall Hypothesis, 1976 Wis. L. Rev. 171.
See e. g., T. Sellin, The Death Penalty, A Report for the Model Penal Code Project of the American Law Institute (1959).
United Nations, Department of Economic and Social Affairs, Capital Punishment, pt. II, ¶ 159, p. 123 (1968).
1. Ehrlich, The Deterrent Effect of Capital Punishment: A Question of Life and Death (Working Paper No. 18, National Bureau of Economic Research, Nov. 1973); Ehrlich, The Deterrent Effect of Capital Punishment: A Question of Life and Death, 65 Am. Econ. Rev. 397 (June 1975).
Id., at 409.
The variables other than execution risk included probability of arrest, probability of conviction given arrest,' national aggregate measures of the percentage of the population between age 14 and 24, the unemployment rate, the labor force participation rate, and estimated per capita income.
Id., at 398, 414.
See Passell & Taylor, The Deterrent Effect of Capital Punishment: Another View (unpublished Columbia University Discussion Paper 74-7509, Mar. 1975), reproduced in Brief for Petitioner App. E in Jurek v. Texas, O. T. 1975, No. 75-5844; Passell, The Deterrent Effect of the Death Penalty: A Statistical Test, 28 Stan. L. Rev. 61 (1975); Baldus & Cole, A Comparison of the Work of Thorsten Sellin & Isaac Ehrlich on the Deterrent Effect of Capital Punishment, 85 Yale L. J. 170 (1975); Bowers & Pierce, The Illusion of Deterrence in Isaac Ehrlich’s Research on Capital Punishment, 85 Yale L. J. 187 (1975); Peck, The Deterrent Effect of Capital Punishment: Ehrlich and His Critics, 85 Yale L. J. 359 (1976). See also Ehrlich, Deterrence: Evidence and Inference, 85 Yale L. J. 209 (1975); Ehrlich, Rejoinder, 85 Yale L. J. 368 (1976). In addition to the items discussed in text, criticism has been directed at the quality of Ehrlich’s data, his choice of explanatory variables, his failure to account for the interdependence of those variables, and his assumptions as to the mathematical form of the relationship between the homicide rate and the explanatory variables.
See Baldus & Cole, supra, at 175-177.
Bowers & Pierce, supra, n. 8, at 197-198. See also Passell & Taylor, supra, n. 8, at 2-66—2-68.
See Bowers & Pierce, supra, n. 8, at 197-198; Baldus & Cole, supra, n. 8, at 181, 183-185; Peck, supra, n. 8, at 366-367.
Passell, supra, n. 8.
See also Bailey, Murder and Capital Punishment: Some Further Evidence, 45 Am. J. Orthopsychiatry 669 (1975); W. Bowers, Executions in America 121-163 (1974).
In Furman, I considered several additional purposes arguably served by the death penalty. 408 U. S., at 314, 342, 355-358. The only additional purpose mentioned in the opinions in, these cases is specific deterrence — preventing the murderer from com*237mitting another crime. Surely life imprisonment and, if necessary, solitary confinement would fully accomplish- this purpose. Accord, Commonwealth v. O’Neal, — Mass. —, —, 339 N. E. 2d 676, 685 (1975); People v. Anderson, 6 Cal. 3d 628, 651, 493 P. 2d 880, 896, cert. denied, 406 U. S. 958 (1972).
See, e. g., H. Hart, Punishment and Responsibility 8-10, 71-83 (1968); H. Packer, Limits of the Criminal Sanction 38-39, 66 (1968).
See Commonwealth v. O’Neal, supra, at -, 339 N. E. 2d, at 687; Bowers, supra, n. 13, at 135; Sellin, supra, n. 2, at 79.
See Hart, supra, n. 15, at 72, 74-75, 234-235; Packer, supra, n. 15, at 37-39.
Mr. Justice White’s view of retribution as a justification for the death penalty is not altogether clear. “The widespread reenactment of the death penalty,” he states at one point, “answers any claims that life imprisonment is adequate punishment to satisfy the need for reprobation or retribution.” Roberts v. Louisiana, post, at 354. (White, J„ dissenting). But Mr. Justice White later states: “It will not do to denigrate these legislative judgments as some form of vestigial savagery or as purely retributive in motivation; for they are solemn judgments, reasonably based, that imposition of the death penalty will save the fives of innocent persons." Post, at 355.
See Commonwealth v. O’Neal, supra, at —, 339 N. E. 2d, at 687; People v. Anderson, 6 Cal. 3d, at 651, 493 P. 2d, at 896.