dissenting.
The federal and state constitutions instruct that the death penalty must not be applied in an arbitrary, inconsistent, or discriminatory manner. Because no proportionality review, including ours, can sufficiently ensure that the death penalty is applied rationally and consistently, I dissent and would vacate Richard Feaster’s death sentence.
*442I.
For thirteen years, we have worked diligently to devise procedures to determine whether a death sentence is proportionate. We have commissioned the best members of our judicial family to employ criminal justice experts and statisticians to improve our proportionality tests. Currently, we are in the process of another effort to design effective procedures for proportionality review, In re Proportionality Review Project, 165 N.J. 206, 757 A.2d 168 (2000) (Proportionality Review II), and we anticipate ongoing deliberations about how to improve those procedures as more information becomes available.
Despite the enormous effort we have expended, it has become apparent that proportionality review does not provide “a standard of due process protection commensurate with the gravity of the sentence to be imposed.” State v. Bey, 137 N.J. 334, 427, 645 A.2d 685 (1994) (Bey IV) (Handler, J., dissenting). Because “death is different” in its severity and finality, Gregg v. Georgia, 428 U.S. 153, 188, 96 S.Ct. 2909, 2932, 49 L.Ed.2d 859, 883 (1976), merely identifying justifications for a death sentence under review is inadequate. What is required is a comparative sense of rationality and consistency. Although we pride ourselves on the fact that our Constitution provides a “more expansive source of protections against the arbitrary and nonindividualized imposition of the death penalty” than does the United States Constitution, State v. Ramseur, 106 N.J. 123, 190, 524 A.2d 188 (1987), I believe our proportionality review falls short of guaranteeing that the death penalty is administered in a fair and consistent manner.
One reason for the failure is that, despite our efforts to improve proportionality review, we have not yet settled on consistent, meaningful standards for our two tests: salient-factors and comparative culpability. Close scrutiny of our opinions reveals that the standards change from case to case. Since our first proportionality review opinion, State v. Marshall, 130 N.J. 109, 613 A.2d 1059 (1992) (Marshall II), we have employed fourteen different descriptions of the standard for disproportionality. See State v. *443Harvey, 159 N.J. 277, 355-58, 731 A.2d 1121 (1999) (Harvey III) (Handler, J., dissenting).
Equally important is the fact that the standards we employ are applied in a haphazard and arbitrary fashion. Precedent-seeking review has devolved into a litany of pejorative but generic descriptions of defendants accompanied by a catalogue listing of distinctions from comparison cases; there is never any meaningful discussion of why those distinctions are significant. See State v. Loftin, 157 N.J. 253, 439-40, 724 A.2d 129 (1999) (Loftin II) (Handler, J., dissenting), cert. denied, 528 U.S. 897, 120 S.Ct. 229, 145 L.Ed.2d 193 (1999). There is also little consistency among our opinions with regard to which objective factors are most significant, which make a case more or less aggravated, and how, proeedurally, we are to perform comparisons. It is not that we generally offer irrational analyses, but simply that we have defined the process so loosely that almost any negative description of a defendant or any rationalization for why he was sentenced to death can lead to a conclusion of proportionality.
Many of the deficiencies in proportionality review have been discussed extensively elsewhere. See In re Proportionality Review Project, 161 N.J. 71, 99, 735 A.2d 528 (1999) (Proportionality Review I) (Handler, J., concurring in part and dissenting in part) (criticizing Court’s standard for assessing disproportionality); State v. DiFrisco, 142 N.J. 148, 224, 662 A.2d 442 (1995) (DiFrisco III) (Handler, J., dissenting) (pointing to principle of unique assignment); State v. Martini, 139 N.J. 3, 90-91, 651 A.2d 949 (1994) (Martini II) (Handler, J., dissenting) (discussing lack of statistical standard to measure disproportionality under frequency review); Marshall II, supra, 130 N.J. at 249-50, 263-65, 613 A.2d 1059 (Handler, J., dissenting) (criticizing coding of reversed death sentences as death sentences; inconsistency and inherent subjectivity of proportionality tests; inclusion of defendant’s own case in frequency analysis; abandonment of generally-imposed standard for proportionality). Rather than repeat those arguments, I simply note that I concur in them and add two additional concerns.
*444First, the standards used in the salient-factors and comparative culpability tests continue to be so flaccid as to allow the Court to conclude that any death sentence is proportional without engaging in a searching and stringent review. Cf. State v. Davis, 116 N.J. 341, 356, 561 A.2d 1082 (1989) (affirming that heightened concerns and responsibilities in capital eases require searching and stringent review of record). Second, the process of comparative review, by its very nature, has proven to be so subjective and unreliable, that it can no longer be counted on as a guarantee of the fairness demanded by our Constitution. Cf. State v. Bey, 112 N.J. 45, 95, 548 A.2d 846 (1988) (Bey II) (noting that constitutional violations cast “tremendous doubt” on fairness of proceedings) (quoting Satterwhite v. Texas, 486 U.S. 249, 250, 108 S.Ct. 1792, 1794, 100 L.Ed.2d 284, 293 (1988)).
II.
The salient-factors test was conceived as a measure of the correlation between the single most significant aggravating factor in a particular case, the “salient factor,” and the death-sentencing rate in a group of cases with the same salient factor. We have said that if the ratio of death sentences to penalty-trial or death-eligible cases is high, then the relatively high rate of death sentencing constitutes “strong evidence of the reliability of [the] defendant’s death sentence.” Bey TV, supra, 137 N.J. at 358, 645 A.2d 685. A relatively low rate of death sentencing indicates that we should “scrutinize the ease for the possible influence of impermissible factors.” Martini II, supra, 139 N.J. at 30, 47, 651 A.2d 949. That standard fails in both theory and practice.
By omitting identification of any range of sentencing rates that would constitute strong evidence of the un reliability of a defendant’s death sentence, the salient-factors test cannot serve as a “coefficient of consistency.” Marshall II, supra, 130 N.J. at 153, 613 A.2d 1059. The test is supposed to determine what kinds of murders society regards as most reprehensible, and thus give us more confidence in a death sentence that is in a category deemed *445particularly reprehensible. Id. at 168, 613 A.2d 1059. However, even if the salient-factors test reveals that, in a particular category, the defendant alone was singled out for death, we do not utilize that fact to suggest disproportionality.1 Indeed, we use salient factors only to determine whether there is a societal consensus for the death penalty for certain categories of murders, never to determine whether there is a societal consensus against the death penalty. Such a standard is at odds with the growing skepticism in our state about the fairness and propriety of the death penalty. See A. 1817, 106 Sess. (N.J. Jan. 11, 2000) (proposing bill abolishing the death penalty in New Jersey); A. 1853, 106 Sess. (N.J. Jan. 24, 2000) (proposing three-year moratorium on imposition of death penalty and creating study commission in interim); Kathy Barrett Carter, 63% of Jerseyans Favor Death Penalty, The Star-Ledger, Oct. 10, 1999 (reporting poll results that 63% of New Jersey residents approve of death penalty, down from 72% in 1994, although 56% believe it is disproportionately imposed on the poor, 42% believe it is disproportionately imposed on African-Americans, and only 44% support death penalty if life without parole available).
Since 1987, when a record nine death sentences were handed down, no more than four death sentences have been imposed in any year. In 1998-99, out of thirty-three death-eligible eases and twelve capital trials, juries handed down only two death sentences. Only one death sentence has been imposed this year after six penalty-phase trials. To inquire whether there is a societal consensus that the death penalty is an appropriate punishment for a certain category of homicides, see State v. Cooper, 159 N.J. 55, 72, 731 A.2d 1000 (1999) (Cooper II), while ignoring the fact that our *446juries are increasingly reluctant to impose the death penalty, is difficult to rationalize.
A more practical problem is that we do not relate differing salient-factors test outcomes to precedent-seeking review to ensure that the two proportionality tests are “complementary, can confirm each other, and can be compared to each other.” Harvey III, supra, 159 N.J. at 315, 731 A.2d 1121. Indeed, it is apparent that we do not link the level of scrutiny we accord to precedent-seeking review at all to the outcome of the salient-factors test. Rather, our opinions reveal that we conduct precedent-seeking review with the same degree, or an even lesser degree, of scrutiny in cases with relatively low death-sentencing rates. For example, the defendants in Marshall and Martini were classified in subcategories with the lowest possible death sentencing rates (i.e., zero percent excluding those defendants). A reasonable person would expect a concomitantly enhanced precedent seeking review. On the contrary, the opinions in those cases provide the briefest discussion of comparative culpability — three and six pages, respectively. Marshall II, supra, 130 N.J. at 181-82, 185-86, 187-88, 613 A.2d 1059; Martini II, supra, 139 N.J. at 74-79, 651 A.2d 949. Ironically, the cases with more in-depth comparative culpability analyses are those with higher frequencies of death sentencing. See, e.g., State v. Chew, 159 N.J. 183, 211-20, 731 A.2d 1070 (Chew II), cert. denied, — U.S. —, 120 S.Ct. 593, 145 L.Ed.2d 493 (1999); Loftin II, 157 N.J. at 336-45, 724 A.2d 129.
The salient-factors test should be fully integrated with precedent-seeking review. Where similar cases result in a relatively high rate of death sentencing, precedent-seeking review should be used to determine whether the subject case is more analogous to the large group of death-sentenced cases or the smaller group of life-sentenced cases. See David T. Baldus, Special Master, Death Penalty Proportionality Review Project Final Report to the New Jersey Supreme Court at 34 (Sept. 24, 1991). Conversely, in a case where comparison cases generally result in life sentences, we should use precedent-seeking review to determine whether the *447case under review is among the most culpable of the comparison cases.
As it presently plays out, the salient-factors test reveals nothing about the comparative appropriateness of a particular death sentence and does nothing more than “set the stage for whatever subjective determinations or moral judgments might be made under the precedent-seeking approach.” Martini II, supra, 139 N.J. at 106-07, 651 A.2d 949 (Handler, J., dissenting) (citations omitted).
III.
After completing the salient-factors test, the Court selects similar cases from a defendant’s salient factor subcategory, and compares the defendant’s case with AOC summaries of similar cases in that subcategory. Specifically, we are directed to examine the particular facts of the defendant’s ease relative to other cases, compare both the cases’ statutory aggravating and mitigating factors, and objective non-statutory factors indicative of culpability such as gruesomeness of the offense, mental capacity of the defendant, and prior criminal record. See Harvey III, supra, 159 N.J. at 294, 731 A.2d 1121; Chew II, supra, 159 N.J. at 198, 731 A.2d 1070; DiFrisco III, supra, 142 N.J. at 183, 662 A.2d 442. Taking into account the different facts, defendants, juries, and legal issues in each case, we must determine whether similarly situated defendants are generally given life sentences. See Bey TV, supra, 137 N.J. at 369, 645 A.2d 685. Theoretically, the death sentence under review will be overturned if it does not follow a typical sentencing pattern. Ibid. In practice, it has never happened.
The most significant problem with the standard for the comparative culpability test is that it is not truly comparative. Although we believe, in theory, that “no matter how heinous the crime, we focus, finally, on individual defendants; their acts and their lives,” Loftin II, supra, 157 N.J. at 279, 724 A.2d 129, we have not done so in practice. Realistically, our analyses more closely resemble *448traditional, non-comparative proportionality review because they primarily discuss whether a particular death sentence is appropriate as an abstract issue, separate from whether it is disparate relative to results in other cases.2 Instead of questioning whether a particular death sentence is aberrant when compared to the sentences imposed on defendants in factually similar cases, Harvey III, supra, 159 N.J. at 307, 731 A.2d 1121, we, in effect, merely ask whether there exists any possible justification at all for the death sentence and if satisfied, accept the sentence as proportional.
A telling example of our turn toward traditional proportionality review may be seen in our most recent proportionality opinions. Starting with Martini, we began the practice of prefacing our case comparisons with an analysis of the defendant’s culpability. 139 N.J. at 74-76, 651 A.2d 949. In that analysis, we assess the defendant on an undefined scale of low, average, and high culpability. The assessments are made with no reference to comparison cases, only to the Court’s intrinsic sense of morality.
More important, in the actual comparison section of our opinions, no true comparison occurs. We consistently validate death sentences despite similar cases in which the defendant received a life sentence. The way we achieve that result is by focusing all our attention on a single feature. See, e.g., Ante at 412-13, 757 A.2d at 280 (distinguishing defendant’s case from life-sentenced eases based on lack of single mitigating circumstance); Harvey III, supra, 159 N.J. at 317-18, 731 A.2d 1121 (same). There is rarely, if ever, an explanation of why the distinction is significant. In my view, by honing in on the presence or absence of just one or a few factors in a defendant’s case, the Court fails to honor its duty to evaluate a defendant’s culpability by an entire set of *449objective factors relevant to blameworthiness, victimization, and character.3 Martini II, supra, 139 N.J. at 50, 651 A.2d 949.
Worse, it has been our practice to focus almost exclusively on the aggravating circumstances of the case under review while underscoring the mitigating circumstances of comparison cases, and thus distorting the comparison process altogether. See, e.g., Harvey III, supra, 159 N.J. at 316-19, 731 A.2d 1121 (highlighting ugly facts and minimizing mitigating evidence in defendant’s case and using opposite focus in comparison cases). The results of such an analysis will inevitably be biased toward the State’s position and lead us to render inaccurate conclusions.
The Court’s typical precedent-seeking review suffers not only from its superficiality and predetermined result, but from the inconsistency of its approach in labeling a defendant more or less culpable. Some of the factors that the Court considers in categorizing a defendant as highly culpable are so vague and universally applicable as to be meaningless. See, e.g., Harvey III, supra, 159 N.J. at 319, 731 A.2d 1121 (finding defendant’s culpability heightened because he was “cold and calculating”); DiFrisco III, supra, 142 N.J. at 205, 662 A.2d 442 (determining defendant’s culpability was heightened because “at the end of the day there is still a victim”); Bey TV, supra, 137 N.J. at 384, 645 A.2d 685 (concluding defendant’s culpability heightened because mitigating evidence was contested).
With other factors, defendants are proverbially damned if they do and damned if they don’t. For example, in Chew II, supra, 159 *450N.J. at 213, 731 A.2d 1070, we found that the defendant’s relative culpability was high because he failed to confess. On the contrary, in Cooper II, supra, 159 N.J. at 102, 731 A.2d 1000, we said that a comparison case was less deathworthy because the defendant claimed a co-defendant had committed the murder. Cf. DiFrisco III, supra, 142 N.J. at 206, 662 A.2d 442 (explaining defendant’s relative culpability was high because he confessed but for reasons unknown to Court); Bey IV, supra, 137 N.J. at 385, 645 A.2d 685 (finding defendant’s relative culpability high because he waited until penalty phase to confess).
Execution-style killing is likewise a double-edged sword. The Court has determined, by turns, that it made one defendant more deathworthy and another less deathworthy. Compare Loftin II, supra, 157 N.J. at 338, 724 A.2d 129 (noting execution-style killing indicative of low victimization) with DiFrisco III, supra, 142 N.J. at 205, 662 A.2d 442 (noting execution-style killing indicative of high victimization). Killing for money is also Janus-like. Compare DiFrisco III, supra, 142 N.J. at 209, 662 A.2d 442 (determining pecuniary motive indicates high culpability) with Chew II, supra, 159 N.J. at 217, 731 A.2d 1070 (determining defendant in comparison case less culpable because, out of loyalty and friendship, he killed for money).
Other factors that the Court has used to upgrade or downgrade a defendant’s culpability are similarly unjustifiable. In DiFrisco III, we said that it was worse to perform a murder for hire and thus kill for purely pecuniary gain, than it was to be the hirer and employ someone to kill a “destructive” family member. 142 N.J. at 208-09, 662 A.2d 442. In Loftin II, we said that being the victim of only “sporadic” child abuse makes a defendant more deathworthy than one who was subject to “chronic” child abuse. 157 N.J. at 340, 724 A.2d 129. In Harvey III we said that a defendant who intended to inflict a bodily injury that he knew would probably result in death was less deathworthy than a defendant who intended to cause death, despite the fact that the Legislature has deemed both mental states equally culpable. 159 *451N.J. at 317, 731 A.2d 1121. Inexplicably, in Cooper II, we found two murders less deathworthy because, in one case, the fatal events began as a consensual social encounter, and in the other, because the victim bought drugs from the defendant. 159 N.J. at 99, 112, 731 A.2d 1000. In Bey IV, we determined that a defendant who lacked much-needed mental health care as a child was more culpable than another defendant because the other defendant was institutionalized as a child. 137 N.J. at 384, 645 A.2d 685.
Those value judgments are not only purely subjective, they are, for the most part, objectively unreasonable. Giving our stamp of approval to a sentence of death based on analyses such as those contradicts our long-standing commitment to “ensure that the death penalty is being administered in a rational, non-arbitrary, and evenhanded manner, fairly and with reasonable consistency.” Marshall II, supra, 130 N.J. at 131, 613 A.2d 1059.
Despite our heavy, almost exclusive reliance on precedent-seeking review, see, e.g., Cooper II, supra, 159 N.J. at 88, 731 A.2d 1000, it is, in practice, as meaningless as frequency analysis. Although we are clear in our mission, namely to determine whether a particular death sentence is aberrant when compared to the sentences received by defendants in factually similar cases, we have offered no guidance about how to actually conduct the comparison except in the grossest of terms. See Harvey III, supra, 159 N.J. at 307, 731 A.2d 1121; Chew II, supra, 159 N.J. at 210, 731 A.2d 1070. In practice, comparative culpability analysis has become free-form, sloppy and inconsistent from case to case. Unsurprisingly, it has injected an impermissible degree of randomness into proportionality review.
In the past, we have recognized problems with the subjective evaluations inherent in precedent-seeking review. Cooper II, supra, 159 N.J. at 92-93, 731 A.2d 1000. Despite the obvious deficiencies in the process, we have nonetheless stayed the course, declaring that it is useful to go through the mechanics of comparative review, and that, “properly applied,” it can help the court *452identify disproportionate death sentences. Id. at 93, 731 A.2d 1000. I do not share the Court’s optimism that simply going through the motions of comparative review satisfies our constitutional duty to distinguish between proportionate and disproportionate death sentences. We have not fulfilled our promise to make our subjective value judgments “explicit, so that they can be analyzed and tested against whatever objective measurements are applicable.” Marshall II, supra, 130 N.J. at 120, 613 A.2d 1059. Our personal moral assessments, thinly disguised as objective conclusions, are not subject to any meaningful guidelines, measurements, or criteria. Simply put, our Constitution demands more.
IV.
Even with better standards, I doubt that, in light of the inescapable subjectivity of proportionality review, any form of it will be workable. Again, that is not to say that the Court has failed to invest extraordinary resources to create a fair, impartial, and objective proportionality review process. See Proportionality Review I, supra, 161 N.J. at 71, 735 A.2d 528; In re Proportionality Review, 122 N.J. 345, 585 A.2d 358 (1990). The dangerous result of that investment, however, has been to imbue our flawed process with an aura of scientific accuracy. That is pseudo-science at its core: the application of mathematical methods to an inherently subjective effort to give the observer unwarranted confidence in an outcome. State v. J.Q., 252 N.J.Super. 11, 40, 599 A.2d 172 (App.Div.1991), aff'd, 130 N.J. 554, 617 A.2d 1196 (1993).
Through our procedures for proportionality review, we have tried to insulate ourselves from the kind of subjective decisions that capital juries face. The quantitative salient-factors test and the comparison of objective factors in precedent-seeking analysis appear to narrow our discretion. That appearance is deceiving. Without the benefit of witnessing a live defendant and his mitigating evidence, it is almost impossible not to be overwhelmed by the stark and often bloody description of a murder scene and the *453defendant’s gruesome acts. Indeed, “[f]aeed with a horrific crime and overwhelming evidence of guilt, reviewing courts are often unable to imagine that a jury would have imposed any sentence but death.” State v. Marshall, 148 N.J. 89, 248, 690 A.2d 1 (1997) (Marshall III). Thus, although much of our death penalty jurisprudence struggles with limiting jury discretion, even more questionable is our own exercise of discretion.
It is time to acknowledge that the reasoned comparison of objective factors is not merely a difficulty, but a practical impossibility. While recognizing that precedent-seeking review “depends almost exclusively on imprecise, subjective reactions by the Court to the comparison cases,” we have forged ahead because it impels us “to examine, evaluate, and compare” similar crimes and thereby determine whether a particular sentence is aberrational. Cooper II, supra, 159 N.J. at 92-93, 731 A.2d 1000. The latter goal cannot be achieved in light of the former problem.
The determination that a history of drug abuse suggests comparatively less blameworthiness, or that shooting a victim at close range rather than from afar suggests more blameworthiness, may be constant among the subjective views of a majority of the Court, but that does not make the determination objectively reasonable. By making such comparisons, we not only reach untoward results, thinly veiled as “objectivity,” but we do a disservice to the victims of terrible crimes characterized as “low victimization” incidents, as well as to defendants whose lawyers are simply less skilled in describing their traumatic life histories in such a way to make them relatively more “compelling” to us. Loftin II, supra, 157 N.J. at 340, 724 A.2d 129. Although the criminal law often crudely translates human behavior and thought processes into legalistic categories, “the imposition of death by public authority is ... profoundly different from all other penalties....” See Ramseur, supra, 106 N.J. at 326, 524 A.2d 188 (quoting Lockett v. Ohio, 438 U.S. 586, 605, 98 S.Ct. 2954, 2965, 57 L.Ed.2d 973, 990 (1978)). With ten to twenty gruesome murder cases to review, we cannot hope to impose order on varying expressions of human suffering *454and human evil. In the end, despite our good efforts, an impermissible degree of randomness will persist. See Martini II, supra, 139 N.J. at 100, 651 A.2d 949 (Handler, J., dissenting). In New Jersey, “there is no meaningful basis for distinguishing the few cases in which [the death penalty] is imposed from the many cases in which it is not.” Furman v. Georgia, 408 U.S. 238, 313, 92 S.Ct. 2726, 2765, 33 L.Ed.2d 346, 392 (1972) (White, J., concurring). Although we may try to explain why certain defendants are sentenced to death and others to life, in the end our conjecture is only that.
In Ramseur, we stated that the paramount purpose of proportionality review is consistency — achieving similar results in similar cases. 106 N.J. at 330-31, 524 A.2d 188. Stated another way by former Special Master Baldus, proportionality review was established “to insure that the cases in which death sentences are carried out can be meaningfully distinguished from those cases in which lesser penalties are normally imposed,” and to limit death sentences to only the most aggravated cases for which “death sentences are the usual routine practice.” Baldus Report at 25 n. 23.
It is obvious to me from reviewing the AOC case summaries that similarly situated defendants who have committed similar crimes often receive vastly different sentences. Cf. Pulley v. Harris, 465 U.S. 37, 68, 104 S.Ct. 871, 889, 79 L.Ed.2d 29, 51-52 (1984) (Brennan, J., dissenting) (describing unpredictability of capital sentencing nationwide). New Jersey juries infrequently impose the death penalty and we assume that juries reserve the death penalty for the worst crimes. Loftin II, supra, 157 N.J. at 322, 724 A.2d 129 (quoting Marshall II, supra, 130 N.J. at 153, 613 A.2d 1059). Based on the fact that the death-sentencing rate for all death-eligible cases is thirteen percent and the total death-sentencing rate at penalty trial is thirty percent, we have concluded that “[t]he low rates show a proper reservation of the death sentence for only the truly worst murderers.” Martini II, supra, 139 N.J. at 80, 651 A.2d 949. It is tempting to believe the *455argument that the infrequent imposition of death sentences in this state is the result of careful funneling of cases, leaving death row for only the most “deathworthy.” However, experience shows that that is not true. Numbers only reflect the quantity of death sentences meted out, not the deathworthiness of the crimes for which it has been imposed.
Currently, there are fourteen men on death row.4 From August 1982, when the new death penalty statute was enacted, until December 1999, only about ten percent of the 555 death-eligible cases have resulted in a death sentence. Juries imposed a death sentence in only fifty-seven cases — one fifth of the 276 capitally-prosecuted cases. Of those fifty-seven death sentences, this Court has vacated thirty-nine death sentences because of constitutional errors that deprived the defendant of a fair trial.5
A review of those cases reveals that there is no indication that the men on death row are the most culpable among the state’s murderers or that they committed the most gruesome crimes. Indeed, it is much more likely that a host of factors other than culpability, moral blameworthiness and character are determinative of who is sentenced to life and who is sentenced to death. The competence (or incompetence) of defendant’s trial counsel, the resources of the county in which the crime was committed, the popularity of the death penalty in that county, the notoriety of the victim, the defendant’s financial resources, the particular policy of the county prosecutor, the race of the defendant and/or victim(s), and the publicity surrounding the crime are probably better predictors of a jury’s life and death decision. See, e.g., American Bar Association Death Penalty Moratorium Resolution (Feb. 3, 1997) (citing incompetent counsel and racial discrimination as *456barriers to fair administration of death penalty); Stephen B. Bright, Counsel for the Poor: The Death Sentence Not For the Worst Crime But For the Worst Lawyer, 103 Yale L.J. 1835 (1994); David S. Baime, Systemic Proportionality Review Project: Report to the New Jersey Supreme Court at 54 (Dec. 1, 1999) (reporting that although Caucasian-victim cases constitute only 46% of death-eligible cases, 62% of death sentences imposed were for Caucasian-victim cases); Leigh B. Bienen et al, The Re-Imposition of Capital Punishment in New Jersey: The Role of Prosecutorial Discretion, Rutgers L.Rev. (1988) (tracking higher risk of capital prosecution in particular group of pro-death penalty counties, especially for African-American defendants with Caucasian victims).
Although we have concluded that no defendant, as yet, has been able to supply “relentless documentation” of the effect of those factors on a particular death sentence, the State has never shown that those factors do not affect sentencing decisions. Moreover, any savvy observer of the death penalty over the past two decades can not help but be at least skeptical about the insulation of death sentencing decisions from those impermissible factors. David Weisburd, Good for What Purpose?: Social Science, Race and Proportionality Review in New Jersey, Hebrew University of Jerusalem (visited July 21, 2000) (http://mishpatim.mscc.huji. ae.il/newsite/ CrimeGroup/weisburd/workpap.htm) (printed in Social Science, Social Policy and the Law, Russell Sage, R. Kagan, P. Ewick and A. Sarat eds. (1999)).
The outcome of penalty phase decisions may also be explained in part by the common grant of mercy by our state’s juries. However, it appears that mercy is not apportioned rationally or based on factors similar to those examined by the Court for precedent-seeking review.6 See California v. Brown, 479 U.S. *457538, 561-62, 107 S.Ct. 837, 849-50, 93 L.Ed.2d 934, 952-53 (1987) (Blackmun, J., dissenting) (“While the sentencer’s decision to accord life to a defendant at times might be a rational or moral one, it also may arise from the defendant’s appeal to the sentencer’s sympathy or mercy, human qualities that are undeniably emotional in nature.”). Mercy can depend on a multitude of factors: a defendant’s good looks, the enthusiasm and support of his family members, the jury’s identification with his background, or the status of the victim (an essayist of note once observed that our society is more outraged at crimes perpetrated against college girls than those against cocktail waitresses). Such factors are infinite, personal, and often inexplicable. See Saffle v. Parks, 494 U.S. 484, 493, 110 S.Ct. 1257, 1262-1263, 108 L.Ed.2d 415, 427-28 (1990) (rejecting instruction permitting jury to consider sympathy because it would “allow[ ] the fate of a defendant to turn on the vagaries of particular jurors’ emotional sensitivities”).
Proportionality review is at once an essential element of the appellate process and, as presently constituted, inadequate to the task. It simply fails to meet the constitutional mandate of providing meaningful appellate procedures to ensure against an arbitrary and capricious death penalty system.
V.
Separate and apart from my reservations about proportionality review in general, it is clear that the way the Court applied it in Richard Feaster’s case was unjust and that, even if its methodology is accepted, Feaster’s sentence is disproportionate, considering both the crime and the defendant.
A. SalienP-Factors Test
To begin, I note the anomalous fact that Feaster’s own case is included in the salient-factors test’s statistics. It is ineomprehen-*458sible to me that in determining whether a particular death sentence is in conformity with a sentencing pattern, the pattern includes the case under review. In such a scenario, a death sentence confirms its own propriety. Marshall II, supra, 130 N.J. at 263, 613 A.2d 1059 (Handler, J., dissenting).
Equally troubling is the fact that one of the death-sentenced F-2 cases included in Feaster’s category is that of Robert Morton, whose case we also decided today. In that case the Court upheld Morton’s death sentence finding it not disproportionate.7 There is a tautological problem with upholding the validity of Feaster’s sentence based on the assumed validity of Morton’s sentence and simultaneously upholding the validity of Morton’s sentence based on the assumed validity of Feaster’s sentence.
As problematic is the fact that, of the four other cases in the F-2 subcategory that resulted in a death sentence, none has yet been affirmed after proportionality review. The two sentences imposed on Jacinto Hightower have since been vacated because they resulted from constitutionally deficient trials. Ronald Long’s death sentence has also been overturned and he has since pled guilty to non-capital murder. Morton’s death sentence has been deemed not disproportionate only today.
Excluding Feaster’s case under the salient-factors test, thirteen percent of death-eligible cases in the F-2 category result in the death penalty, compared to eleven percent overall. Excluding both Feaster and Morton, as I believe we should, the death sentencing rate among all death-eligible cases in the F-2 subcategory is only ten percent, and the death sentencing rate for those proceeding to the penalty phase is only nineteen percent. From the group of F-2 eases, it is impossible to find a “societal consensus” that the death penalty is an appropriate punishment *459for the F-2 subcategory of homicides. See Cooper II, supra, 159 N.J. at 72, 731 A.2d 1000.
B. Comparative Culpability Test
1. Defendant’s Culpability
Concluding from the salient-factors test that, at the very least, the vast majority of F-2 cases result in life sentences, precedent-seeking analysis should inform us whether Feaster was “singled out unfairly” for the death penalty, or whether he was truly among the most culpable of the F-2 defendants A searching comparison of F-2 cases demonstrates that Feaster is far from the most culpable of the F-2 defendants. His case more closely resembles the large majority of cases that resulted in life sentences and his death sentence should accordingly be vacated.
The precedent seeking analysis in the Court’s opinion, as in all prior proportionality review opinions, is a subjective moral evaluation of Richard Feaster as opposed to the comparative analysis that is promised. The bulk of the precedent-seeking analysis is devoted to recounting defendant’s crime and setting forth comparison case summaries. Scant attention is paid to the actual comparison of objective factors in each case. That analysis is exactly the kind of traditional, offense-oriented proportionality review that we have directly rejected. See, e.g., Marshall II, supra, 130 N.J. at 129-30, 613 A.2d 1059.
Without reference to comparison cases, the Court characterizes Feaster’s level of moral blameworthiness as “average to high” by weighing his youth and lack of premeditation against the victim’s vulnerability, lack of justification or excuse, and “complete callousness.” Ante at 404-06, 757 A.2d at 275-76. The factors supporting the Court’s “average to high” rating are illusory because they are factors present in nearly all death-eligible cases. The description of a “completely callous crime” committed against a “vulnerable victim” without justification or excuse can be applied to nearly every ease in the death-eligible universe and certainly to all the cases in Feaster’s comparison group. Calling a defendant a “cold *460and calculating murderer” may be fitting, but the descriptor is useless insofar as it is applicable to every person who commits a death-eligible crime. As such, it has no place in proportionality review. Harvey III, supra, 159 N.J. at 319, 731 A.2d 1121. The Court’s discussion of victimization and character suffer from similar problems of subjective, standardless evaluation. Ante at 406, 757 A.2d at 276. Its determinations that victimization was “average to low” in Feaster’s case and that Feaster’s character demonstrated “average to high” culpability, ante at 406, 757 A.2d at 276, are based solely on visceral reactions that imply nothing about Feaster’s relative death-worthiness. The Court’s discussion of the degree of victimization, in fact, directly contravenes our opinion in Loftin II. 157 N.J. at 338, 724 A.2d 129. There, we analyzed a very similar “execution-style” shooting, but rather than labeling it as “average to low” victimization, we stated that “[i]n comparison to other murder cases we have examined, this was not a particularly violent or brutal killing.” Ibid. The Court also fails to explain how Feaster’s lack of cooperation with authorities and allegedly callous comments to the jailhouse snitch8 are so negatively probative in light of mitigating circumstances such as his minimal criminal record, capacity for rehabilitation, and the testimony that he cried and expressed remorse and disbelief when driving home on the night of the murder.
Even more disturbing is the fact that in Loftin II, we characterized the facts of Feaster’s case in a good light relative to moral blameworthiness to demonstrate that Loftin was more culpable than Feaster. See Loftin II, supra, 157 N.J. at 340-41, 724 A.2d *461129. In that opinion, we said that Feaster presented “uncontroverted evidence” that he suffers from a mental disease or defect, namely encephalopathy, “an injury to the left frontal lobe region making him more violence prone.” Id. at 340, 724 A.2d 129. We also recognized in Loftin II that Feaster was relatively young at the time of the offense and still leading the life of a juvenile (e.g., living with his parents and lacking employment). Id. at 341, 724 A.2d 129. We highlighted the facts that Feaster’s parents were both alcoholics and that his father was abusive. That we can spotlight those facts in one case to portray Feaster as having relatively low culpability, and virtually ignore them in Feaster’s own case demonstrates the hopelessly unstructured and unreliable nature of our precedent-seeking review.
2. Comparative Culpability
Only eighteen of the thirty-three eases in Feaster’s F-2 subeategory proceeded to the penalty phase. In those eighteen penalty trials, the jury sentenced three people other than Feaster to death: Morton (affirmed today), Hightower (two death sentences vacated), and Long (death sentence overturned, entered guilty plea to non-capital murder). Our duty in this portion of proportionality review is “to ensure that the defendant has not been ‘singled out unfairly for capital punishment.’ ” Cooper II, supra, 159 N.J. at 88, 731 A.2d 1000 (quoting Martini II, supra, 139 N.J. at 47, 651 A.2d 949); accord Chew II, supra, 159 N.J. at 210, 731 A.2d 1070. For the purpose of precedent-seeking review, therefore, we determine whether the comparison life-sentenced cases render Feaster’s death sentence disproportionate. It is important to note that that process is not a justifying one. It does not, in any way, attempt to deflect from the notion of murder as evil or imply that the death of the victim is anything other than horrific. Its focus is only to place a defendant’s terrible crime on a continuum of other terrible crimes.
That said, Feaster’s case is notable for its relatively low victimization. The only aggravating factor in Feaster’s case was that the murder was committed in the course of a robbery, and that *462factor is present in every F-2 case. Thus, it does not, in any way, distinguish Feaster as more culpable than any other F-2 defendant. Feaster is somewhat unusual, however, in that he fired a single shot and did not cause prolonged suffering to either the victim or any non-decedent victims. Donaghy died instantly, and there is no evidence that he was ever threatened, forced to take any action at gunpoint, or injured in any way other than the shooting itself. Again, to distinguish Feaster on the basis of relatively low aggravation is not meant to diminish the horrible impact of the crime on the victim, but to place Feaster on a scale of culpability relative to other murderers.
The Court acknowledges that four life-sentenced cases are sufficiently similar to Feaster’s to raise serious questions about the proportionality of Feaster’s sentence. Ante at 410-12, 757 A.2d at 278-79. I agree. The life sentences for Larry Jones, Rafael Slaughter, Aaron Stamps and Charles Williams are, as the Court states, “more difficult” to reconcile with Feaster’s death sentence. Ante at 411, 757 A.2d at 279.
The Court also recognizes that Abdel Jaber Saleh’s life sentence “cannot readily be reconciled with [Feaster’s] death sentence.” Ibid. That is an understatement. The degree of victimization caused by Abdel Jaber Saleh is among the most serious in the F-2 subcategory. Saleh strangled his victim with a metal chain, bound and gagged him with duct tape and plastic ties, dragged his body across the floor, doused him with charcoal fluid, then set him on fire. The autopsy report indicated that the victim was still alive when Saleh set him on fire. In addition to the 4(g) aggravating factor, Saleh’s jury found aggravating factor 4(f) — that Saleh committed the murder to escape detection or apprehension. Like Feaster, Saleh was only 22 at the time of the offense, had no significant criminal history, and was uncooperative with law enforcement officers. Unlike Feaster, however, Saleh presented no evidence of any mental disease or defect, childhood trauma, brain damage, or intoxication at the time of the offense. Still, Saleh was given a life sentence.
*463Jones, Williams, and Slaughter all harmed multiple victims— five, four, and three, respectively. Although the 4(b) aggravating factor was only found in Jones’s case, all three cases involved a serious risk of death to innocent bystanders as well as the use of threats and violence against non-decedent victims. Neither Jones nor Slaughter presented mitigating evidence as compelling as Feaster’s mitigating evidence. Williams’s evidence of sexual abuse and history of emotional disturbance are unusually tragic, but his criminal history is much more extensive than Feaster’s; Williams spent all but 93 days of his adult life incarcerated. Williams’s culpability is heightened by the fact that one of his surviving victims sustained serious brain damage while another was left disabled.
Stamps’s case resembles Feaster’s quite closely. Stamps, who shot a security guard twice in the chest during a bank robbery, acted with co-defendants and did not cooperate with the authorities. Like Feaster, Stamps grew up in an unstable home, has a substance abuse problem, and has borderline intelligence. Stamps, however, had three prior armed robbery convictions. The major differences between the two cases — Feaster’s minimal criminal history, organic brain damage, and abuse at the hands of his alcoholic father — underscore Feaster’s relatively lower culpability. Viewed side by side with those five life sentences that are either more aggravated, less mitigated, or both, Feaster’s death sentence stands out as an anomaly.
The Court curiously finds five other F-2 penalty phase cases that resulted in life sentences “readily” distinguishable from Feaster’s case, despite their close similarity. Ante at 408, 757 A.2d at 277. In distinguishing the life sentences given to John Downie, Craig Hart, Roger Hoyte, Joseph Wilson, and David Russo, the Court states that, unlike Feaster, those cases contain evidence of, (1) extreme mental or emotional disturbance, (2) youth, (3) diminished capacity at the time of the offense, and (4) lack of a significant criminal history. Ante at 409, 757 A.2d at 278. Those distinctions cannot be reconciled with, (1) Feaster’s uncon-*464troverted evidence of organic brain damage and psychological problems; (2) Feaster’s youth (he was as young as or younger than Downie, Hart, Russo, and Hoyte at the time of the offense, and only a few years older than Wilson); (3) Feaster’s alcohol and cocaine intoxication on the night of the offense; and (4) his minimal prior record consisting of a single, unindictable conviction.
The Court’s conclusions are also difficult to reconcile with our opinion in Loftin II, in which we found Feaster to be relatively less culpable than similar cases specifically because of evidence that he suffered from a mental disease or defect and that his crime was mitigated by his youth and immaturity. 157 N.J. at 340-41, 724 A.2d 129. In comparing Feaster to Downie, Hart, Hoyte, Wilson, and Russo, ante at 408-10, 424-25, 757 A.2d at 277-78, 286-87, the Court also ignores Feaster’s evidence of chronic child abuse that we found “compelling” in Loftin II. 157 N.J. at 340, 724 A.2d 129. None of the comparison defendants suffered from the same pattern of abuse that Feaster did. Only Downie and Hoyte offered comparable evidence of unstable, violent, and dysfunctional childhood environments, but the offenses committed by Downie (shooting at a police officer as well as murdering a gas station attendant) and Hoyte (shooting and stabbing three taxicab drivers) are much more aggravated than Feaster’s.
Regarding the failure of the jury to find the 5(c)(age) and 5(d) (mental disorder or intoxication) mitigating factors in Feaster’s case that were found for Downie, Hart and Wilson, “we are free to go beyond a jury’s conclusion in respect of a mitigating factor.” Martini II, supra, 139 N.J. at 3, 651 A.2d 949. Although we cannot say that the jury’s findings regarding those factors are wrong, they are inexplicable in comparison to other penalty phase cases. Why would a jury, and the Court, find age a mitigating factor for Hart (age 25), Downie (age 24), and Hoyte (age 22), but not Feaster (age 22)? Ante at 409, 757 A.2d at 278. In any case, Feaster’s relative youth and immaturity should be considered in *465the case comparisons. The same logic applies to the 5(d) mitigating factor for intoxication. Although Feaster’s jury did not find the 5(d) factor, the uncontroverted evidence was that he had consumed alcohol and cocaine on the night of the offense and that lowers his relative culpability in precedent-seeking review.
In comparison with Hoyte and Downie, Feaster’s case is not particularly aggravated. To be sure, Feaster did not cooperate with authorities as Hoyte did. On the other hand, Hoyte’s cooperation was fueled by the opportunity he had to plead to a non-capital offense in return for testimony against accomplices. Thus, it is not revelatory, in any meaningful way, of better character. Moreover, because Downie and Hoyte shared other similar characteristics with Feaster, such as a lack of a significant criminal record, youth, and intoxication during the offense, there is no logical explanation for the discrepancy in sentencing.
Russo’s mitigating evidence (a minor criminal record, a history of substance abuse, no serious psychiatric history, a diagnosis of personality disorders, and evidence that he succeeded in controlled environments such as employment) is also similar to Feaster’s, but his shooting of three gas station attendants is more aggravated than Feaster’s offense. In comparison with Russo’s life sentence, Feaster’s death sentence is aberrant.
Hart and Wilson are theoretically distinguishable from Feaster because there was a jury finding of the 5(c) mitigating factor for age and 5(d) mitigating factor for mental disease or defect or intoxication in each case. However, that distinction does not hold water. Feaster’s age and intoxication at the time of the offense should be considered in precedent-seeking review despite the absence of a jury finding. In reality, Feaster’s evidence of youth, immaturity, and intoxication was similar to Hart’s and Wilson’s. What is truly different about Hart and Wilson is that those defendants both threatened victims at gunpoint and Wilson’s victim suffered a protracted death. Neither Hart nor Wilson grew up in the kind of violent and unstable home that Feaster did, *466and neither was affected by the brain damage that impaired Feaster’s mental functioning. Viewing the cases in their entirety, Hart and Wilson are more culpable than Feaster and this their life sentences raise serious questions about the proportionality of Feaster’s death sentence.
Other comparison cases that did not proceed to a penalty trial, either because the defendant pled to a non-capital offense or was not capitally prosecuted, provide further support for the dispro-portionality of Feaster’s death sentence. Four such cases involve serious injuries to non-decedent victims in addition to a homicide victim. Corey Washington shot an elderly store clerk in the head; Emmanuel Charles shot a restaurant manager three times; Harold Rodriguez shot a gas station owner six times; and Jose Soto shot a restaurant owner after his wife told Soto there was no money to steal. Except for Rodriguez, who is married with three children and suffers from AIDS, not one of those cases contains mitigating circumstances as significant as Feaster’s. None suffered from child abuse, organic brain damage, severe depression, or the constraints of borderline intelligence. Rodriguez was much older than Feaster when he committed his offense, and Soto, Charles, and Washington were only a few years younger. Like Feaster, all have minimal prior records except for Charles, who has an extensive juvenile record including numerous thefts and assaults. It is true that we do not know as much about the mitigating circumstances of those defendants because they all pled to non-capital offenses (Washington and Rodriguez to murder; Charles to felony murder; Soto to aggravated manslaughter). However, unlike Feaster, those defendants have also benefited from the AOC’s coding procedures, under which Feaster would have been coded with the 5(d) and 5(f) mitigating factors. See Administrative Office of the Courts, Instructions for Screening Cases: Mens Rea, Ozon Conduct and Factors (Feb. 5, 1999) (directing 5(d) to be coded where record indicates serious head injury, brain damage or drug addiction, and 5(f) to be coded where defendant has no convictions for an indictable offense and less than four disorderly persons convictions).
*467Timothy Harris presents yet another stark example of how anomalous Feaster’s sentence is. Harris, acting alone, followed two people into a store, robbed them at gunpoint, and shot one victim in the head because she did not hand over all her money. The victim survived for hours in the emergency room before dying as a result of the gunshot wound. The only mitigating evidence in the record about Hands is that he was nineteen at the time of the offense, and that that was his first indictable offense as an adult. Accordingly, the AOC coded the 5(c) and 5(f) mitigating factors present. However, Harris had essentially led a life of crime as reflected in thirty-one juvenile complaints beginning when he was thirteen, for a variety of offenses including robbery, assault, and weapons offenses.
Of the twenty-three comparison cases to Feaster, at least sixteen can be easily classified as more culpable than Feaster. Their offenses were either more aggravated, presented fewer mitigating circumstances, or both, yet resulted in a life sentence.
To me, this entire exercise underscores the randomness of the death sentence process and reveals that Feaster’s death sentence is disproportionate because it cannot be reconciled on a consistent, reasoned basis with the majority of life sentences imposed on similarly situated defendants.
Accordingly, I dissent.
For affirmance — Chief Justice PORITZ and Justices O’HERN, STEIN, COLEMAN, and LaVECCHIA— 5.
For vacating and remandment — Justice LONG — 1.
For instance, Robert Marshall and John Martini were both in salient factor subcategories with death sentencing rates of zero percent, excluding their own case. Marshall II, supra, 130 N.J. at 168, 613 A.2d 1059; Martini II, supra, 139 N.J. at 33, 651 A.2d 949. In Marshall's subcategory, none of the three other death-eligible cases resulted in the death penalty, and in Martini's subcategory, none of the five other death-eligible cases resulted in the death penalty. Ibid. Still, their death sentences were upheld as not disproportionate.
Traditional proportionality review requires a court to determine whether, on an abstract level, a crime is heinous enough to justify the imposition of the death penalty. In contrast, our statute directs us to consider a defendant's sentence only as it compares to other sentences for similar crimes. Marshall II, supra, 130 N.J. at 129-30, 613 A.2d 1059.
The complete list of objective factors that the Court is supposed to use to evaluate a defendant's relative culpability is as follows: motive, premeditation, justification or excuse, evidence of mental disease, defect, or disturbance, knowledge of victim's helplessness, knowledge of effects on nondecedent victims, defendant's age, defendant's involvement in planning the murder, violence and brutality of the murder, injury to nondecedent victim, prior record, other unrelated acts of violence, cooperation with authorities, remorse, capacity for rehabilitation. Marshall II, supra, 130 N.J. at 155, 613 A.2d 1059.
Two death row inmates died of natural causes, including Joseph Harris who received two death sentences; one died as a result of a homicide.
Recalculating death sentencing rates in light of reversals shows that only seven percent of penalty phase cases and three percent of all death-eligible cases have resulted in the death penalty.
The factors are: motive, premeditation, justification or excuse, evidence of mental disease, defect, or disturbance, knowledge of victim's helplessness, knowledge of effects on nondecedent victims, defendant's age, defendant's in*457volvement in planning the murder, violence and brutality of the murder, injury to nondecedent victim, prior criminal record, other unrelated acts of violence, cooperation with authorities, remorse, and capacity'for rehabilitation.
Testimony by jailhouse informants is especially problematic in and of itself. Steve Mills and Ken Armstrong, The Jailhouse Informant, Chicago Tribune, Nov. 16, 1999 (describing "jailhouse snitch" testimony as leading cause of wrongful capital convictions); Samuel R. Gross, Lost Lives: Miscarriages of Justice in Capital Cases, 61 Law and Contemp. Problems 125, 138-39 (1998) (citing examples of informant testimony resulting in capital conviction of innocent persons); Mark Curriden, No Honor Among Thieves, ABA Journal (1989) (same).