dissenting.
Defendant David Cooper was convicted of the purposeful and knowing murder of L.G., a six-year-old girl, whom the jury concluded defendant kidnapped, sexually assaulted and then strangled. The jury found two aggravating factors, that defendant committed the murder to escape detection, N.J.S.A. 2C:ll-3c(4)(f), and that he had done so in the course of committing an aggravated sexual assault and kidnapping, N.J.S.A. 2C:ll-3e(4)(g). The jury rejected the State’s submission of the c(4)(e) aggravating factor, that the murder had involved torture or aggravated assault to the victim.
The jury found numerous mitigating factors pursuant to the catch-all e(5)(h) factor, the sum of which provide a picture of a defendant who suffered throughout his youth from extensive abuse and neglect, as well as exposure to drug and alcohol addiction. See ante at 67-68, 731 A.2d at 1006-07. The jury found, however, that the aggravating factors outweighed the mitigating factors.
The jury sentenced Cooper to death on the capital murder charge. The trial court sentenced defendant to fifty years imprisonment with a twenty-five-year parole disqualifier on the kidnapping count and to twenty-five years imprisonment with a ten-year parole disqualifier on the aggravated sexual assault count. The *162trial judge merged defendant’s felony murder conviction with the purposeful or knowing murder conviction.
On direct appeal, this Court affirmed defendant’s convictions and sentences for capital murder and kidnapping. The Court vacated defendant’s sentence for aggravated sexual assault, merging that conviction with his kidnapping conviction. State v. Cooper, 151 N.J. 326, 406-07, 700 A.2d 306 (1997).
Defendant requested a proportionality review for his death sentence pursuant to N.J.S.A. 2C:ll-3e, which we granted. The Court now finds that defendant’s sentence is proportionate. I disagree, for several reasons.
First, the Court should not delay in declaring the 1992 amendment, which limits the universe of comparison cases in proportionality review to death-sentenced defendants, unconstitutional. Accord State v. Harvey, 159 N.J. 277, 354, 731 A.2d 1121 (1999) (Harvey III) (Handler, J., dissenting); State v. Loftin, 157 N.J. 253, 373, 724 A.2d 129 (1999) (Loftin II) (Handler, J., dissenting). The Court has stated very clearly that a universe consisting only of death-sentenced cases would be “inadequate,” State v. Marshall, 130 N.J. 109, 136, 613 A.2d 1059 (1992) {Marshall II), and that a universe of all death-eligible cases would best serve “the purposes to be achieved by proportionality review[,]” id. at 137, 613 A.2d 1059. This finding has now been validated by The Honorable David S. Baime, whom the Court, in Loftin II, supra, 157 N.J. at 265, 724 A.2d 129, appointed as Special Master to examine our proportionality review methodology and make recommendations for improvements. See- The Honorable David S. Baime, Report to the New Jersey Supreme Court: Proportionality Review Project 10 (Apr. 28, 1999) {Special Master Report) (“I emphasize that a universe limited to cases in which the death sentence was imposed cannot support a coherent proportionality review system.”). Having stated that “[o]n our receipt of [the Special Master’s report], we will be in a position to determine whether the statutory limitation on the proportionality review universe prevents meaningful appellate review[,]” Loftin II, supra, *163157 N.J. at 287, 724 A.2d 129, the Court need not, and should not, wait any longer to declare the 1992 amendment unconstitutional, see ante at 73-74, 731 A.2d at 1009-10 (echoing Loftin II, supra, in which the Court deferred decision on 1992 amendment “until the Court [has] received Judge Baime’s report”).
Second, I disagree with the Court’s application of existing proportionality review methodology to defendant’s ease. Given our acknowledgment that the current methodology is flawed in various important ways, evidenced most compellingly by the Court’s decision to appoint Judge Baime to study the problems,1 we cannot, in good conscience, ignore the Special Master’s recommendations for improvement in our evaluation of defendant’s claim. Accord Harvey III, supra, 159 N.J. at 352, 731 A.2d 1121 (Handler, J., dissenting).
Third, and most importantly, I disagree with the Court’s rejection of defendant’s claim that the death penalty statute must be invalidated because of the unconstitutional risk that race discrimination plays a role in the prosecuting and sentencing of death-eligible defendants in New Jersey. See ante at 115-16, 731 A.2d at 1034. I find that defendant, like those before him who presented the same compelling data, see Harvey III, supra; Loftin II, supra, has documented such an unconstitutional risk. We must no longer ignore the fact that our efforts to cleanse the justice system of race discrimination may well have failed. See Harvey III, supra, 159 N.J. at 366-68, 731 A.2d 1121 (Handler, J., dissenting) (examining Court’s early use of extensive voir dire in racially charged cases to try to. identify and eliminate racist jurors). When the final punishment is death, we cannot, in good conscience, keep postponing our acknowledgment that statistical data, along with what we know about history and our society, see id. at 361-74, 731 A.2d 1121 (Handler, J., dissenting) (detailing evidence *164of race discrimination in CCH Report from capital-sentencing database, as well as in larger society); Loftin II, supra, 157 N.J. at 405-409, 724 A.2d 129 (Handler, J., dissenting) (recounting history of race discrimination in the criminal justice system in New Jersey and United States), require ,us to declare the death penalty statute unconstitutional.
No evidence to date has demonstrated that race discrimination does not play a role in our prosecution and sentencing of capital cases. We must act now to invalidate the death penalty statute, or, at the very least, to place a moratorium on executions until we have such evidence. Accord Harvey III, 159 N.J. at 374, 731 A.2d 1121 (Handler, J., dissenting); Loftin II, 157 N.J. at 446, 724 A.2d 129 (Handler, J., dissenting).
In addition to my conclusions with regard to systemic issues, I disagree with the Court’s final disposition on defendant’s individual proportionality review claim under existing proportionality review methodology. I believe that defendant’s sentence is disproportionate when his case is compared to those of other similarly situated defendants. I also maintain that the Court’s approach to proportionality review remains unworkable due to its inherent potential for subjectivity and manipulation. The Court’s proportionality review methodology yields results that “respond[ ] simply to considerations that ‘are moral and hopelessly subjective and value-laden’ in nature.” State v. DiFrisco, 142 N.J. 148, 213, 662 A.2d 442 (1995) (DiFrisco III) (Handler, J., dissenting) (quoting Marshall II, supra, 130 N.J. at 265, 613 A.2d 1059 (Handler, J., dissenting)); see also State v. Martini, 139 N.J. 3, 106-07, 651 A.2d 949 (1994) (Martini II) (Handler, J., dissenting) (“Although originally designed as the more objective of the two methods that make up proportionality review, frequency analysis as applied by the Court does little more than set the stage for whatever subjective determinations or moral judgments might be made under the precedent-seeking approach.”) (citing omitted).
*165Although these pitfalls in the majority’s analysis are not as prevalent in the present case as in some other cases, they nevertheless exist, and the Court ought not delay its examination of the Special Master’s recommendations and the construction of a more objective and methodologically sound review. Until we have established such a methodology, we should refrain from conducting proportionality reviews.
I, therefore, dissent.
I
I have already addressed in previous cases the systemic issues before the Court.- See Loftin II, supra, 157 N.J. at 373, 724 A.2d 129 (Handler, J., dissenting) (criticizing Court’s application of faulty methodology to defendant’s case, Court’s decision to defer judgment on 1992 amendment, and Court’s failure to acknowledge presence of race discrimination in capital sentencing scheme); Harvey III, supra, 159 N.J. at 352-55, 731 A.2d 1121 (Handler, J., dissenting) (same); DiFrisco III, supra, 142 N.J. at 212-13, 662 A.2d 442 (Handler, J., dissenting) (citing coding problems, Court’s refusal to quantify disproportionality in its frequency review, and “pervasive ambiguity” in precedent-seeking review); Martini II, 139 N.J. at 82-106, 651 A.2d 949 (Handler, J., dissenting). Here, I will focus solely on my objection to portions of the Court’s analysis of defendant’s individual proportionality claim.
A.
First, the Court has drastically, and without explanation, discarded the fundamental basis of its own original standard for measuring disproportionality. In its first proportionality review, the Court clearly stated:
We believe that the frequency approach will help us to review cases in terms of the substantive principle that we believe should be controlling in these cases, namely, “[a] death sentence is comparatively excessive if other defendants with similar *166characteristics generally receive sentences other than death for committing factually similar offenses in the same jurisdiction.”
[Marshall II, supra, 130 N.J. at 153-54, 613 A.2d 1059 (quoting Tichnell v. State, 297 Md. 432, 468 A.2d 1, 17 n. 18 (1983)).]
This standard, in and of itself, is one that fails to provide much guidance for the Court in the absence of some numerical guideposts to ensure consistency over time. In Marshall II and subsequent cases, however, the injection of a substantively different standard into the analysis, without explanation, only added to the difficulty in establishing a stable benchmark for disproportionality. The Court introduced the “aberrational” standard only in passing at first, id. at 167, 613 A.2d 1059 (“[BJecause the case before the Court is a partial reflection of community values, even if aberrational, it should be considered.”), and then more definitively: “[W]e conclude that capital death sentencing for contract murderers is not random or aberrational.” Id. at 174, 613 A.2d 1059. Even in Marshall II, however, the Court did not engage in any discussion of how to reconcile the imposition of an “aberrational” standard with the Court’s stated goal to measure disproportionality by a general imposition standard.
At the time, I criticized the Court for employing a “subjective intuitive examination of precedent,” which allowed the majority to reach the conclusion that defendant Marshall’s sentence was not disproportionate. Id. at 250, 613 A.2d 1059 (Handler, J., dissenting). I have since echoed this concern by noting the number of unquantifiable and substantively different verbal, rather than numerical standards the Court has employed over time to measure disproportionality. Harvey III, supra, 159 N.J. at 356-57, 731 A.2d 1121 (Handler, J., dissenting) (listing fourteen qualitatively different standards used by Court in its proportionality reviews to date). Unanchored to a quantifiable or numerical base, proportionality review continues to drift away from its goal.
The general imposition standard will never yield a proportionate death sentence because of our growing database and.the remaining infrequency with which New Jersey juries sentence defendants *167to death. See Loftin II, supra, 157 N.J. at 418 n. 18, 724 A.2d 129 (Handler, J., dissenting). Instead of acknowledging this development, the Court, accommodating the existing data, has changed the standard to produce findings of proportionality. See ante at 107, 731 A.2d at 1029 (“Our primary objective is the detection and prevention of aberrational sentences.”); id. at 115, 731 A.2d at 1034 (“Proportionality review seeks to determine only whether a particular death sentence is aberrational____”) (citation omitted). “[T]he Court’s review seems almost to be guided by an effort to find proportionality at all. costs, rather than to root out disproportionality.” Harvey III, supra, 159 N.J. at 413, 731 A.2d 1121 (Handler, J., dissenting); see also Chew II, supra, 159 N.J. at 268, 731 A.2d 1070 (Handler, J., dissenting) (stating that proportionality review has been reduced to “whatever works” to justify a finding of proportionality).
The Court’s failure over time to directly address the inseparable issues of subjectivity and broad application, as well as its subtle eradication here of all but the most narrow of standards, reveals its inability to grasp the need for a numerical preponderance standard, among other things, so that we can, finally, be sure what we mean by “disproportionality.” See Harvey III, 159 N.J. at 412, 731 A.2d 1121 (Handler, J., dissenting) (“It is time for the Court to articulate what it means by disproportionate, rather than to continue to insist that it knows only when a sentence is not disproportionate.”); Loftin II, 157 N.J. at 419, 724 A.2d 129 (Handler, J., dissenting) (calling for numerical majority or preponderance that can be equated with proportionality) (citing DiFrisco III, supra, 142 N.J. at 212, 662 A.2d 442 (Handler, J., dissenting); Martini II, supra, 139 N.J. at 90-91, 651 A.2d 949 (Handler, J., dissenting); Bey IV, supra, 137 N.J. at 408, 645 A.2d 685 (Handler, J., dissenting)).
If the Court is to ignore precedent in this dramatic way, it should do so in an overt manner by providing an explanation for its narrowing of the Court’s standard for determining when a *168sentence is disproportionate. It may be that after review of the Special Master’s recommendations, the Court will want to start with a clean slate, clarifying not only our methodology, but also the very definition of disproportionality. At that point, we would be well-served to attack the issue of subjectivity head-on by establishing a realistically applicable standard by which to measure disproportionality, which the Court has so far failed to do.
In the meantime, we must ask, as the Nebraska Legislature recently did in its own state, why the individuals who now sit on death row — only thirteen out of a death-eligible universe of 433— are being treated differently from the others. See Dirk Johnson, Legislature of Nebraska Votes Pause In Executions, N.Y. Times, May 21, 1999, at A14 (Reporting Republican Nebraska Senator Kermit A. Brashear’s statement, “[W]hen I see 165 murderers in Nebraska prisons and only 10 are on death row, I want to know, Why are some of these people treated differently?”).2 While our proportionality reviews have been more extensive than those in Nebraska, we have often been unable to articulate why certain defendants are sentenced to prison and others to death. See ante at 107, 111, 114-15, 731 A.2d at 1029, 1032, 1033 (stating that the AOC’s summaries provide no explanation for the prosecutors’ decisions to forego capital punishment in six cases in comparison group). As long as this is the case, we should take the Nebraska Legislature’s lead in placing a moratorium on executions (and proportionality reviews) until we have an adequate system in place to gather the information that is necessary to prevent grave injustice.
B.
Second, the Court, in my view, once again misinterprets the results in the salient-factors test. Here, as in past cases, see *169Harvey III, supra, 159 N.J. at 301-03, 731 A.2d 1121; Loftin II, supra, 157 N.J. at 328-29, 724 A.2d 129; DiFrisco III, supra, 142 N.J. at 173-74, 662 A.2d 442; Martini II, supra, 139 N.J. at 33-38, 651 A.2d 949; Bey IV, supra, 137 N.J. at 353-58, 645 A.2d 685, the majority compares the death-sentencing rate of defendant’s salient-factors C-l subeategory, murders involving “sexual assault with particular violence/terror” and that of his composite C category, -with the overall death-sentencing rates for death-eligible defendants. See ante at 77-80, 731 A.2d at 1012-13.
The salient-factors test is designed to assign defendants to categories based on the essential elements of their crimes in order to identify the sentencing rate for similarly situated defendants. “Using the average death-sentencing rate as the benchmark of proportionality would defeat the purpose of dividing the death-eligible cases into subcategories at all ... [T]he court might as well be comparing defendant to all other death-eligible defendants who were not sentenced to death.” Harvey III, supra, 159 N.J. at 380, 731 A.2d 1121 (Handler, J., dissenting). I believe, therefore, that we should simply evaluate the results of the groupings independently of the overall statistics.
In defendant’s C-l category, nineteen percent of the death-eligible defendants were sentenced to death and thirty-nine percent of the penalty-trial eases resulted in death sentences.3 In the broader C category, sixteen percent of death-eligible defendants were sentenced to death, while thirty-four percent of penalty-trial cases resulted in death sentences. See ante at 78, 731 A.2d at 1012.
These numbers do not, in and of themselves, suggest “general imposition” of the death penalty in either the C category, or even *170the C-l subcategory. However, we must use these findings to-inform our precedent-seeking review, which attempts to add a more qualitative element to the analysis. See Harvey III, supra, 159 N.J. at 360, 731 A.2d 1121 (Handler, J., dissenting) (“Without linking the two analyses, the salient-factors test fads to inform the Court at all____”). There, we must look at whether defendant is among those whom we would expect to be sentenced to death within his salient-factors category. Accord id. at 359-60, 731 A.2d 1121 (Handler, J., dissenting).
C.
The Court’s index-of-outcomes test is also meant to inform our judgment of defendant’s disproportionality. As the majority points out, however, the test is problematic because of the inconsistency of results across regressions and the large confidence intervals. See ante at 72-73, 731 A.2d at 1009-10. These two problems have plagued our index-of-outcomes model since its inception. See Harvey III, supra, 159 N.J. at 303-04, 731 A.2d 1121 (noting inconsistency in culpability scores in Court’s preceding proportionality reviews and detailing wide range of confidence intervals in defendant’s results); Chew II, supra, 159 N.J. at 205-06, 731 A.2d 1070 (urging caution in relying on results of index-of-outcomes analysis); Loftin II, supra, 157 N.J. at 422 n. 21, 724 A.2d 129 (Handler, J., dissenting) (noting wide range of confidence intervals). Therefore, I find the results suspect, at best. The Special Master agrees, recommending that for these reasons and others, the test be eliminated from our proportionality review. See Special Master Report, supra, at 76-107.
Defendant’s results confirm these concerns, though not as starkly as in previous cases. His culpability level ranges from one to three and the confidence interval for the model examining both statutory and non-statutory factors with a penalty-trial universe ranges from nine to seventy-five percent. Ante at 83, 731 A.2d at *1711015-16. It is worth noting that defendant’s culpability score is lowest in the two models reflecting the smallest range of confidence intervals. In the model measuring statutory and non-statutory factors with a death-eligible universe, the test reveals that we are ninety-five percent certain that defendant’s death sentencing rate falls between three percent and thirty percent. Cooper is assigned to culpability level one based on his ten percent predicted probability of death, giving his group a death-sentencing rate of five percent. In the model measuring only statutory factors with a death-eligible universe, the confidence interval is similarly low (and the range small): we are ninety-five percent certain that defendant’s death-sentencing rate falls between eight and thirty-eight percent, and again he is assigned to culpability level one with a death-sentencing rate of five percent. Id. at 83, 731 A.2d at 1015-16. 4These low scores, indications of disproportionality, are among the most reliable we have seen to date5 because of the small range of the confidence intervals and the fact that they occur in the two models measuring both juries’ and prosecutors’ decisions — namely, those that employ the death-*172eligible universe. The other two models incorporate only defendants whose cases proceeded to a penalty trial, so prosecutors’ charging decisions are not taken into account.
These results, almost identical to those received by defendant Harvey, see Harvey III, supra, 159 N.J. at 805, 731 A.2d 1121, suggest that Cooper is among the least blameworthy of all death-eligible defendants. The results, however, should make us question the viability of the test to accurately measure true culpability given the severity of the victimization in Cooper’s case. The Court correctly points to the fundamental flaw most likely to skew the results here, namely, that the models “do not take into account the fact that the victim of defendant’s homicide was a six-year-old girl.” Ante at 82, 731 A.2d at 1014. This factor, likely to play a significant role in the jury’s assessment of defendant’s culpability, does not figure into the AOC’s attempt to classify cases according to levels of victimization.
Given the methodological flaws evident in the index-of-outeomes test, both generally and with regard to defendant’s cases specifically, as well as the Special Master’s recommendation to eliminate the test, I hesitate to make much of the results. I therefore rely primarily on the salient-factors and precedent-seeking review results, to the extent possible, in finding defendant’s sentence disproportionate. The index-of-outcomes test findings merely support this conclusion.
D.
The Court’s precedent-seeking review, while comprehensive, suffers from some of the same inadequacies of prior reviews. Following its litany of cases, in which the Court is often hard-pressed to find meaningful distinctions between life-sentenced defendants and Cooper, the majority concludes by finding defendant’s sentence proportionate. See id. at 115-16, 731 A.2d at 1034. Pursuant to the general imposition standard on which our proportionality review was founded and, even, pursuant to the *173Court’s myopic “aberrational” standard, I cannot abide by such a conclusion.
In nine cases, I have been unable to find any adequate distinguishable feature to justify the disparity between the defendants’ life sentences and Cooper’s death sentence.6 These cases account for over one-quarter of the thirty-eight life-sentenced cases in defendant’s comparison group. Only six defendants in defendant’s group have been sentenced to death and not one of them has received a proportionality review, so we have no basis on which to judge these sentences as valid. Contrary to the Court’s assertion that the death-sentenced cases present “no issue of disproportionality” because they resulted in death, see id. at 93, 731 A.2d at 1022, these cases must be reviewed because they have not yet been declared proportionate. In addition, they have some value in helping us to better understand the reasons juries and prosecutors might opt for death instead of life. At their most instructive, these cases may call into question some of the Court’s assumptions about why certain life-sentenced defendants might have been spared, thereby eviscerating the distinctions between those defendants and Cooper.
Marko Bey, for example, suffered from emotional disturbance; he was only seventeen when he committed his first murder; and his relationship with his victim was a consensual one (the victim actually had consensual sex with him first, and he did not rape and kill her until she refused to do so a second time). The Court, in justifying life sentences for several cases in Cooper’s comparison group, cites the consensual or ambiguous nature of the relationship between the victim and the defendant in some of its discussions. See id. at 98-99, 103, 112-13, 114-15, 731 A.2d at 1024-25, 1027, 1032-33, 1033. Quite apart from the fact that I find offen*174sive any standard that would warrant diminished punishment for a defendant who sexually assaults and murders a woman, simply because he has a consensual relationship with her, I find that the Court’s attempt to distinguish other life-sentenced defendants from Cooper on that basis cannot be squared with Marko Bey’s death sentence. These cases may be distinguished on other bases, so are not discussed here. The Court, however, should be mindful of the facts in the death-sentenced cases when examining distinctions between Cooper and other defendants.
Joseph Harris, a death-sentenced defendant, is another example. Harris had a long history of psychiatric problems and reported hearing voices as early as age nine or ten. Some of his school teachers suggested that he have a psychological evaluation because they believed he was mentally disturbed. He never had an evaluation. Harris claimed he suffered from lapses of consciousness in his adult years. While in the Navy, he had three psychiatric evaluations in which he was diagnosed as being “schizoid” and as having an “inadequate personality.” He received no treatment. While employed at a post office, Harris occasionally reported to work dressed in an imitation of “Ninja garb” or in a camouflage military outfit, performing martial-arts maneuvers in front of his co-workers. He believed he was the target of race discrimination and that he needed to defend himself in this manner. State v. Harris, 141 N.J. 525, 538, 662 A.2d 333 (1995). In spite of these facts, the jury failed to find the diminished capacity or mental disease or defect aggravating factors. Harris later committed suicide on death row.
In light of Harris’s capital sentence, the Court’s attempt to distinguish defendants Michael Prater, Sharob Clowney, Alphonso Timpson and David Collins from defendant by pointing out their emotional problems, see ante at 95, 99-100, 103, 731 A.2d at 1023, 1025, 1027, carries less weight. Harris is proof that juries are willing to sentence to death individuals who suffer from mental and emotional problems. Indeed, statistics demonstrate that even *175defendants for whom juries have found the N.J.S.A. 2C:11-3c(5)(a) mental or emotional disturbance factor are more likely to be prosecuted capitally than others.7 See Chew II; supra, 159 N.J. at 269, 731 A.2d 1070.
Harris’s ease warrants close analysis for another reason. The State offered Harris a plea agreement before the trial began, which the defendant turned down. The fact that the State would have declined capital prosecution of Harris in exchange for a life sentence undermines the usefulness of Harris’s case as justification of Cooper’s death sentence. The prosecutor, at least, believed capital prosecution of Harris was not warranted.
Finally, only three of the forty-five defendants who were capitally tried are currently on death row. Joseph Harris committed suicide, but the others’ sentences have been reversed for varying reasons. These defendants were either not capitally prosecuted the second time around or were not sentenced to death by juries in their subsequent trials. Their death sentences, therefore, hold only minimal value.
I.
A close examination of the life-sentenced defendants confirms the disproportionality of defendant’s death sentence. Many of the juries for C-l, life-sentenced defendants found the mitigating factors indicating that the defendants suffered from mental disease or defect, N.J.S.A. 2C:11-3c(5)(a) or e(5)(d) (extreme emotional disturbance or diminished capacity). In addition, the fact that Cooper’s victim was a young child undoubtedly played a significant role in his sentence. See Marshall II, supra, 130 N.J. at 157, 613 A.2d 1059 (noting that Court has indicated that such. factors as vulnerability of victim, not a statutorily-defined factor, might help explain jury judgments even though they might not make defendant death eligible). Cooper’s case, however, is not *176the only one involving the sexual assault and murder of a young child. In addition, some of the other crimes in defendant’s category involve similar degrees of victimization, even though a young victim was not involved. Many of the cases before the Court are difficult to square with defendant’s. I discuss them here.
Defendant Rivera raped, attempted to rob, and murdered his next-door neighbor, a seventy-eight-year-old widow who often took care of defendant’s children. The victim was badly beaten all over. Her face and neck were covered with bruises; hemorrhaging was found under her tongue, behind one of her eyes and underneath her cheek; she had bruises on her forearms and back, abrasions on her lips and two fractured ribs; and she had been strangled to death. The autopsy also revealed that the victim had been sexually assaulted. Her vagina was tom back and oozing blood. She suffered bruising of the mucous membrane and the area near the urethra, injuries believed to be caused by an object at least three inches in diameter, such as the victim’s cane or the defendant’s hand.
Rivera had a history of alcohol and drug abuse and was seen drank on the day of the murder. Rivera was charged capitally and sentenced to life. The Court states, “Except for the age of defendant’s victim, reconciliation of defendant’s death sentence with Rivera’s life sentence is difficult.” Ante at 97, 731 A.2d at 1024. While an elderly woman cannot be compared to a six-year-old in many ways, does the case not present only another of many faces of evil? The age of Rivera’s victim clearly and significantly increases his blameworthiness given the vulnerability of the elderly victim, especially in light of the extreme violence committed by the defendant. We should require more to distinguish Rivera from Cooper.
Defendant Lester Allan Wilson strangled and sexually assaulted a fourteen-year-old girl who lived in the same hotel as he. He attempted to escape from police custody after being arrested for the murder. Defendant’s I.Q. qualifies him as mildly retarded. *177He was sentenced to life. The Court states, “Other than defendant’s mental retardation, no apparent circumstances concerning this homicide and aggravated sexual assault explain the determination to forego capital prosecution of Wilson.” Id. at 101, 731 A.2d at 1026. Here, however, no evidence of emotional disturbance or diminished capacity was present — defendant suffers from mild retardation based solely on a somewhat low I.Q.
Leroy Taylor sexually assaulted and strangled his girlfriend’s thirteen-year-old niece. When she was found, the victim’s underwear was torn and contained blood and semen. The victim had a nylon stocking around her neck and froth and blood were coming from her nose and mouth. Defendant fled to California and was eventually picked up for violating his parole. He was charged with numerous offenses, including felony murder, sexual assault, hindering prosecution and witness tampering. Taylor was not charged capitally and was permitted to plead guilty. He was sentenced to life.
The Court states, “The prosecutor’s decision to accept Taylor’s guilty plea could have been influenced by uncertainty over whether Taylor’s prior murder conviction as a juvenile could constitute an aggravating factor in a murder prosecution.” Id. at 102, 731 A.2d at 1027. The Court’s justification here is mere speculation, which crumbles when we realize that Taylor’s death-eligibility does not rest on his prior murder conviction alone — indeed, most of the defendants in the C category do not have prior murder convictions. Taylor’s sexual assault makes him death eligible. The Court is then left with no apparent reason to explain the prosecution’s decision to forego capital prosecution.
Robert Bolinger, when his victim came home to find him burglarizing her apartment and attempting to leave, stabbed her, then tied her up on the bed, gagged her and sexually assaulted her. When arrested in connection with another rape and attempted murder, Bolinger confessed not only to the previous offense and the one discussed here, but also to a third rape and murder *178that he committed in between the two others. Defendant pled guilty to the rape and attempted- murder and was ultimately acquitted of the second rape and murder charge, but not until after he was prosecuted for the rape and sexual assault reviewed here. Defendant was not tried capitally and was permitted to plead guilty. He was sentenced to life.
The Court states, “The decision to accept [Bolinger’s] guilty plea may have been influenced by the impulsive rather than premeditated nature of the homicide and by defendant’s drug and alcohol addiction.” Id. at 103, 781 A.2d at 1027. Any standard based solely on impulsivity, especially in the face of Bolinger’s confession, is insupportable here. Had the defendant simply stabbed the victim and run away, one could argue that his murder was more impulsive than Cooper’s, thereby perhaps making Bolinger less culpable. However, given that the defendant took the trouble to tie up his victim and rape her while she was still alive and suffering from a stab wound to the chest, the impulsivity argument loses all weight — is it really so meaningful to distinguish one defendant who lured his victim without having plotted his crime from another who, when surprised by his victim, carries out an equally sinister and prolonged brutal attack? Further, in light of the defendant’s confession to two other rapes and attempted murders (one successful), Bolinger’s apparent impulsiveness here cannot possibly make up for his severe culpability.
Gary Lippen and his co-defendant beat, raped, strangled, stabbed, and tortured a seventeen-year-old girl. After the two took turns raping the victim, Lippen beat her in the head with a stick, punched her in the jaw, and kicked her. The two co-defendants then hoisted her into a tree and broke her legs. Lippen had no history of emotional disturbance, but he had a history of drug and alcohol abuse and the jury found the diminished capacity mitigating factor. Although Lippen eventually confessed, he first denied all involvement in the murder and then, even in his confession, lied to the police. Both defendant and his co-defendant accused the other of playing a dominant role in the *179crime. In addition to his capital charges, Lippen was charged with hindering apprehension, for which he was sentenced to five years. The State accepted Lippen’s guilty plea and the defendant was sentenced to twenty years on his aggravated manslaughter conviction.
The Court states, “Lippen’s cooperation with law enforcement authorities, combined with the likelihood that Henderson was the dominant actor, probably accounts for the prosecutor’s decision to forgo capital prosecution of Lippen.” Id. at 104, 731 A.2d at 1028. This conclusion is suspect given that Lippen was charged with and convicted of hindering apprehension. The AOC narrative states clearly that Lippen lied to law enforcement officers at first. Lippen eventually aided police in securing a conviction against his co-defendant, but the notion that this cooperation (which he clearly rendered in order to reduce his own sentence) formed the basis for his reduced culpability is feeble.
Frank Masini (I and III) raped and murdered his eighty-five-year-old aunt and an eighty-year-old woman within a two-week period. Both victims had been stabbed multiple times in the neck and had been sexually assaulted. Both women were found in a pool of blood, naked from the breast down. They also had stab wounds on their hands. Approximately one year later, Masini murdered an elderly couple for whom he had done carpentry work. Masini was not capitally charged in either sexual assault case.
The Court states, “Other than the reference in the AOC’s summary to Masini’s experiencing ‘detachments from reality,’ no other factors surrounding those four homicides suggests an explanation for the prosecutorial decision to proceed non-capitally against Masini.” Id. at 105, 731 A.2d at 1028. This defendant murdered four elderly people, raping two of them. The prosecutor was aware of all of the murders when determining whether or not to capitally prosecute Masini for his first and third death-eligible crimes. Defendant’s own unproven contention that he suffers “detachments from reality,” in the face of these crimes, is *180the only available information the AOC offers the Court to distinguish Masini’s guilty plea from Cooper’s death sentence. This case surely emphasizes the inability of our current system to answer the very question that proportionality review demands— are similarly situated defendants being treated differently?
Jerome Dennis (I and II) presents a similar scenario, only the inequity is even more glaring. Dennis committed five homicides in four months, two of which involved sexual assaults. Defendant abducted a fourteen-year-old girl at knife-point, sexually assaulted her and stabbed her three times in the neck, eleven times in the chest and ten times in the abdomen. Her body was found at the bottom of a hill four months later, tied up and naked under a pine tree. The second sexual assault victim was a thirty-year-old woman. She was also found naked and tied up at the bottom of a hill with ten stab wounds in her neck. Defendant had several prior convictions, including three for sexual assault. Defendant was not prosecuted capitally and was sentenced to life. The AOC summary makes no mention of any emotional or mental problems suffered by defendant. As the Court recognizes, “No explanation for the prosecutor’s decision to forego capital prosecution can be inferred from the AOC summaries.” Id. at 106, 731 A.2d at 1029.
James Williams (IB), at age twenty-one, raped and murdered a twenty-three-year-old woman who worked in a Trenton nursing home. The victim was found face down and naked, her clothing strewn about the room. There was blood on the floor, the walls, and the furniture. She had thirty-six stab wounds, bruises contusions and abrasions all over her body. Her throat had been slashed. The medical examiner testified that the victim remained conscious after seven frontal stab wounds were inflicted, and that she had lived several minutes after sustaining twenty-one fatal stab wounds to her back. In Williams’s capital trial, the jury found that defendant had committed his murder with torture and depravity, but the jury rejected the emotional disturbance, diminished capacity, and reduced age mitigating factors and sentenced *181defendant to death. Defendant’s sentence was reversed by this Court because of the inadequacy of the voir dire. On retrial, the prosecutor declined to prosecute Williams capitally, and defendant was permitted to plead guilty in exchange for a life sentence. The Court states,
“No factors other than the mitigating evidence offered in the penalty phase, but substantially rejected by the jury, explain the prosecutor’s decision to forego a capital prosecution on remand. That mitigating evidence included testimony reflecting an unstable childhood, numerous foster care placements, emotional instability, intoxication and, at age nine, the accidental shooting of his younger brother.”
[Id. at 106-07, 731 A.2d at 1029.]
Although worth noting, this is precisely the kind of mitigating evidence presented at trial by defendant Nathaniel Harvey and ignored by the Court in companion case Harvey III, supra, 159 N.J. at 390-91, 731 A.2d 1121 (Handler, J., dissenting). The majority in Harvey fails to point out the importance of the mitigating evidence that defendant Harvey was emotionally scarred from having lit a match and accidentally set his sister on fire, killing her, when he was a child. See id. at 391, 731 A.2d 1121 (Handler, J., dissenting). Harvey suffered other childhood abuse, including exposure to alcohol and drug addiction, also not considered by the Court. See ibid. Here, the Court’s willingness to distinguish Williams’s case — which involved a sexual assault with particular violence and terror — from Cooper’s on the basis of mitigating evidence presented at trial regarding his traumatic youth cannot be reconciled with the Court’s failure to acknowledge such evidence in Harvey’s proportionality review, where the defendant’s crime was significantly less egregious.
Carlos Vasquez, a defendant with a prior murder conviction in Puerto Rico, sexually assaulted and asphyxiated a thirteen-year-old girl who had gone out to buy bread for her family. The girl’s body was found in yellow plastic garbage bags in a cardboard box. The victim’s hands had been tied and pulled together behind her back with electrical cord and clothesline. Her spine was fractured and she had been strangled. Vasquez was not prosecuted capital*182ly and pled guilty, receiving a sentence of life. The Court offers' no justification for the disparity between Vasquez’s and Cooper’s sentences.
2.
Despite the possible perceived difficulty in finding the death sentence for a crime so heinous as Cooper’s to be disproportionate, we are obliged to compare Cooper to other similarly situated defendants, not to the group of death-eligible defendants as a whole. As long as this Court has no basis for distinguishing defendant from such a large group of similarly situated defendants, especially when the frequency analysis results confirm our expectation that defendants similar to Cooper will not be sentenced to death, we must consider Cooper’s sentence to be an aberration and vacate it.
II.
Pursuant to existing proportionality review methodology, I find defendant’s sentence to be disproportionate. I also maintain my contention that the current methodology is too inherently subjective to be reliable in cases that are not clear-cut. I find the Court’s application of what it readily admits is a faulty system of proportionality review to defendant’s case irresponsible in light of the availability of Judge Baime’s report containing several recommendations for improvement.8 Further, the Court’s refusal to declare unconstitutional the 1992 amendment redefining our proportionality review universe of cases is unacceptable in light of the Court’s own precedent and the Special Master’s report.
*183Finally, and most importantly, we must heed the warnings of the data before us indicating a risk that racism has infected our capital punishment scheme. When we supplement the consistent statistical models (all of which indicate that race is playing a role) with our common experience and knowledge, we can no longer conclude that such a risk is not present. We must declare the death penalty statute unconstitutional and vacate defendant’s sentence until such time as we can be certain that we are not executing individuals on the basis of their race.
I, therefore, dissent.
For affirmance — Chief Justice PORITZ, and Justices POLLOCK, O’HERN, GARIBALDI, STEIN and COLEMAN — 6.
For reversal — Justice HANDLER — 1.
Judge Baime was asked to explore eight substantive areas in his evaluation. See Loftin II, supra, 157 N.J. at 454-56, 724 A.2d 129.
Republican Nebraska Governor Mike Johanns vetoed the Legislature's bill five days after it was passed, stating that "it would allow death row inmates to 'advance further unnecessary appeals.' " Nebraska Leader Vetoes Suspension of Executions, N.Y. Times, May 27, 1999, at A21.
Defendant's case is not included for analysis. "Placing defendant's case on both sides of the comparison, gauging the proportionality of a defendant's sentence by comparing it to a group of which he is a member, skews the analysis.” Loftin II, supra, 157 N.J. at 420 n. 19, 724 A.2d 129 (Handler, J., dissenting).
maintain my objections to the majority’s use of predicted probabilities as a basis for comparison of Cooper with other defendants, ante at 85-88, 731 A.2d at 1017-19. Our focus should .be on actual, not predicted, sentencing rates. Accord Harvey III, 159 N.J. at 383, 731 A.2d 1121 (Handler, J., dissenting); Loftin II, supra, 157 N.J. at 423, 724 A.2d 129 (Handler, J., dissenting). Further, the Court’s comparison of Cooper to defendants in different culpability levels, ante at 85-88, 731 A.2d at 1017-19, defeats the purpose of the index-of-outcomes test. Accord Harvey III, 159 N.J. at 383 n. 11, 731 A.2d 1121 (Handler, J., dissenting); Loftin II, supra, 157 N.J. at 423-24, 724 A.2d 129 (Handler, J„ dissenting). As in the salient-factors test, the index-of-outcomes results should be used merely to tell us what level of culpability to attribute to defendant, and from that, what death-sentencing rate applies to similarly situated defendants.
Only defendants Harvey and Martini have comparable confidence interval ranges. See Harvey III, supra, 159 N.J. at 305, 731 A.2d 1121 (reporting confidence interval ranges of .05 to .32 for model 2 and .08 to .38 for model 4); Martini Report, tbls. 14 and 15 (reporting confidence interval ranges of .01 to .30 for model 2 and .02 to .27 for model 4).
This group does not include some of the cases the majority characterizes as difficult to square with defendant's. See ante at 96-98, 731 A.2d at 1023-24. Although I find that those cases can, in fact, be distinguished from defendant's, they only add to the difficulty in making a finding on proportionality in this case.
The c(5)(a) factor was not found by the juries in the trial for defendants Prater, Clowney, Timpson and Collins.
One of these improvements is, notably, to require prosecutors to provide this Court with more detailed reasons for their decisions to forgo capital prosecution. See Special Master Report, supra, at 6. This improvement would provide the very information the Court here acknowledges so often that it lacks. See ante at 107, 111, 114, 731 A.2d at 1029, 1032, 1033. Without such information, our proportionality review is inadequate and any conclusions regarding proportionality must therefore be postponed.