dissenting.
Finally, the bell tolls. This Court for the first time affirms both the murder conviction and death sentence of a defendant prosecuted for capital murder. The Court’s decision and judgment serve only to confirm the intractable constitutional infirmities of our capital-murder jurisprudence, its unfathomable incoherence and unmanageable contradictions.
Significant shortcomings surround this prosecution, resulting in a death sentence that under our Constitution can be viewed *215only as arbitrary and capricious, a death sentence that is unmarked by the due process and fundamental fairness demanded in capital causes, a death sentence that constitutes a cruel and unusual punishment. The Court has previously considered and resolved many of the constitutional issues that are directly implicated in this case. See State v. Ramseur, 106 N.J. 123, 524 A.2d 188 (1987); State v. Biegenwald, 106 N.J. 13, 524 A.2d 130 (1987). Given the evolving nature of capital-murder jurisprudence, such issues remain imperatively subject to reconsideration and ripe for redetermination. It is apparent that no guiding legal principle should be considered as definitively settled or controlling when it comes to the death penalty.
This capital conviction and death sentence forcefully draw attention to the weakness of the constitutional support on which the Court has rested the death penalty. The bifurcated-trial scheme in capital prosecutions, with its use of the death-qualified jury, fails to secure genuine impartiality, and in fact may promote the opposite. In this case, the unfairness is peculiarly accentuated: it confronts the defendant with an insoluble dilemma — the choice between a fair trial on guilt or a fair trial on punishment, but not both. Further, identical statutory definitions and simultaneous jury determinations of guilt and punishment create the unacceptable risk that capital prosecutions are overinclusive. These factors, together with the inefficacy of the penalty proceeding, which grossly undermines the reliability of the determination that the defendant deserves to die, exemplify the arbitrary and capricious imposition of the death sentence. Constitutional defects, with other major trial-level errors and deficiencies, impugn this prosecution and demand reversals of the conviction and sentence.
This case augments the formidable doubts concerning the feasibility of the death penalty within a criminal justice system. The enormous resources and extraordinary costs now being poured into capital-murder prosecutions sap our strength in other areas of law enforcement and criminal justice. The distinctive jurisprudence of capital punishment has begun to *216exert an influence on the legal principles and doctrines that govern criminal justice. Our growing absorption, if not obsession, with the death penalty, has generated an institutional compulsion to make the death penalty “work” at any price.
The tolling of the bell is sharply and disturbingly dissonant, given the grave instances of reversible error arising from this prosecution, the serious constitutional infirmities of our capital-murder laws, and the genuine policy considerations militating against the death penalty. Even though it is the defendant who suffers the extreme penalty, constitutional failures penalize everyone. A constitution compromised to convict the guilty gives no comfort to the innocent.
I.
The most profound concern engendered by the imposition of the death sentence in this ease is that defendant was sentenced to die by a jury that was not qualified to render that awesome judgment. Astonishingly, this fact goes virtually unnoted by the Court and, sadly, the Court appears untroubled by it.
It is well-settled that in capital-murder trials only a duly-qualified jury may be impanelled to determine a defendant’s guilt of capital-murder and, if guilty, whether the defendant deserves to die. Indeed, in a capital-murder prosecution the jury must be specially qualified in order to assure that high degree of objectivity, sensitivity, fairness, and impartiality essential to determine criminal guilt and appropriate sentence. Ross v. Oklahoma, 487 U.S. 81, 108 S.Ct. 2273, 101 L.Ed.2d 80 (1988); State v. Bey, 112 N.J. 123, 151-54, 548 A.2d 887 (1988) (Bey II); State v. Ramseur, supra, 106 N.J. 248-54, 524 A.2d 188; State v. Williams, 93 N.J. 39, 61-62, 459 A.2d 641 (1983) (Williams I). The jury’s heavy and special responsibilities in a capital-murder case are unique. In a prosecution in which a defendant can be sentenced to death, the jury is a vital “link between contemporary community values and the penal system — a link without which the determination of punishment could hardly *217reflect ‘the evolving standards of decency that mark the progress of a maturing society.’ ” Witherspoon v. Illinois, 391 U.S. 510, 519 n. 15, 88 S.Ct. 1770, 1775 n. 15, 20 L.Ed.2d 776, 783 n. 15 (1968) (quoting Trap v. Dulles, 356 U.S. 86, 100, 78 S.Ct. 590, 597, 2 L.Ed.2d 630, 642 (1958) (plurality opinion)).
We thus take extraordinary pains in capital-murder cases to insure the impanelling of a jury with special qualifications. All potential jurors must submit to “thorough and searching inquiry by the trial court intp each individual’s attitude concerning the death penalty.” State v. Williams, 113 N.J. 393, 413, 550 A.2d 1172 (1988) (Williams II). We insist that no person may serve on a jury in a capital case whose views concerning the death penalty “would ‘prevent or substantially impair the performance of his duties as a juror in accordance with his instructions and his oath.’ ” Id. at 415, 550 A.2d 1172 (quoting Adams v. Texas, 448 U.S. 38, 45, 100 S.Ct. 2521, 2526, 65 L.Ed.2d 581, 589 (1980)).
We have wrestled with the dilemma posed by the use of a death-qualified jury: whether such a jury qualified with the capacity to sentence a defendant to death can also fairly, and objectively determine the defendant’s criminal guilt. State v. Ramseur, supra, 106 N.J. at 248-54, 524 A.2d 188; see State v. Rose, 112 N.J. 454, 476-77, 548 A.2d 1058 (1988). The Court has resolved that dilemma. It has decided that such a jury can adequately perform both functions, rejecting the notion that death-qualification imperils the jury’s ability to determine guilt. Ibid.; Ramseur, supra, 106 N.J. at 251, 524 A.2d 188; but see Ramseur, supra, 106 N.J. at 428-35, 524 A.2d 188 (Handler, J., dissenting) (the issue of guilt of capital defendants should not be determined by a death-qualified jury); id. at 335-41, 524 A.2d 188 (O’Hern, J., concurring) (death qualification may affect jury’s ability to determine criminal guilt). According to the Court, a defendant may fairly be tried for both guilt and sentence before the same jury in the same proceeding.
*218The basic premise underlying the Court’s reasoning is that death qualification itself is indispensable in securing a jury that can be relied on to decide whether a defendant should be put to death for his crimes. See Williams II, supra, 113 N.J. at 413-27, 550 A.2d 1172. Further, death qualification does not undermine the ordinary juror qualification that suffices to enable a jury to determine only guilt. The Court thus believes that death qualification, in effect, can be superimposed on conventional qualification. Accordingly, such a jury — conventionally qualified and death-qualified — can determine both criminal guilt and punishment. Therefore, in a capital-murder case both the guilt and punishment of the defendant can — and should — be tried by such a jury.
In this case, however, the jury was not death-qualified. Before the jury selection began, defense counsel stated that he preferred to omit all death qualification questions from the jury-selection process because he believed that such questions led to a conviction-prone jury. Counsel obviously reached this position reluctantly. Counsel contended that to exclude prospective jurors opposed to the death penalty from' the guilt phase violated defendant’s right to an impartial jury. He urged that those jurors ordinarily excludable for cause through death qualification nevertheless be qualified for the guilt phase of trial. Further, counsel proposed that if a penalty phase should be necessary those jurors be replaced by separate death-qualified jurors, also selected before the guilt phase, who would sit as alternates during the guilt phase. Defense counsel asserted that this proposed method was the fairest way to try a capital case. (See, e.g., State v. Ramseur, supra, 106 N.J. at 339, 524 A.2d 188 (O’Hern, J., concurring); State v. Bey II, supra, 112 N.J. at 198, 548 A.2d 887 (Handler, J., dissenting)). The trial court denied the motion.
After one full day of jury selection, defense counsel reiterated his view that death qualification of jurors violated his client’s right to an impartial jury. To reduce the asserted prejudicial impact, counsel moved to limit death qualification of *219potential jurors. Specifically, defendant’s counsel, with the concurrence of counsel for co-defendant Thompson, sought to have the trial court limit death qualification to the prospective jurors’ answers to the questionnaire that had been distributed before the individual questioning.1 The State did not object, and the trial court approved this request. Accordingly, during the balance of the voir dire, the trial court conducted additional death qualification only if a potential juror’s questionnaire revealed views on the death penalty that might impair that juror’s ability to follow the court’s instructions. It asked no further questions of jurors whose simple “yes” or “no” answers to the questionnaire contained no such suggestion.
The Court finds nothing untoward in that extraordinary handling of death qualification, explaining that it merely represented a tactical trial decision by defense counsel. Ante at 93, 586 A.2d at 131. Except in the most extreme cases, strategic decisions made by defense counsel will not present grounds for reversal on appeal. Ramseur, supra, 106 N.J. at 281-82, 524 A.2d 188. Our views on the proper administration of the voir dire process, like our views on other death-penalty procedures, are seasoned with a degree of deference to defense counsel’s strategic decisions. Thus, in State v. Hunt, 115 N.J. 330, 354, 558 A.2d 1259 (1989), the Court diluted its commitment to the indispensability of open-ended questions during death-qualification voir dire, noting that defense counsel had “declined the opportunity to request further questioning and did *220not object to the jurors’ qualifications.” In that case, however, voir dire examination went beyond the acceptance of simple answers to written questions, and the Court found that although death-qualification examination “may not have been perfect in all respects, it was sufficient to enable counsel and the court to evaluate the jurors’ fitness to serve.” Ibid. We observed in Rose that trial courts could “take into account a defendant’s concerns about the collateral effects of the death qualification process” and address these concerns in the course of the voir dire. 112 N.J. at 477, 548 A.2d 1058. We never intimated, however, that the trial of a capital-murder case involving both guilt and punishment should be undertaken without duly death-qualifying the jury.
Although defense counsel’s views of the extent and duration of specific questioning may influence the scope of the examination of potential jurors, counsel does not have the prerogative to dispense with death qualification. The process of death qualification is the only way the State and the defendant are assured that the jury harbors no predetermined or preconceived biases and can follow and apply the law in determining whether the defendant is to live or die. The failure to conduct an extensive, individualized death qualification inquiry, even if the result of a strategic and informed decision by defense counsel, negates that assurance. It inescapably constitutes irremediable error. See Williams II, supra, 113 N.J. at 413, 550 A.2d 1172. On the facts before us, the trial court’s decision to grant defense counsel’s request to limit death qualification only to those questions printed on the jury questionnaire was error sufficiently “egregious as to cut mortally into [defendant’s] substantive rights____” Ramseur, supra, 106 N.J. at 282, 524 A.2d 188 (citation omitted) (quoting State v. Harper, 128 N.J.Super. 270, 277, 319 A.2d 771 (App.Div.), certif. denied, 65 N.J. 574, 325 A.2d 708 (1974)).
Because adequate juror qualification is an imperative condition for a valid capital-murder prosecution, extraordinary importance attaches to the voir dire and death-qualification process. *221State v. Bey II, supra, 112 N.J. at 152, 548 A.2d 887. In State v. Biegenwald, supra, 106 N.J. at 29, 524 A.2d 130, we underscored the indispensability of the “searching voir dire interrogation” to obtaining a fair and impartial jury, earlier stressed in Williams I, supra, 93 N.J. 39, 459 A.2d 641. In State v. Hunt, supra, 115 N.J. at 348, 558 A.2d 1259, we stated: “An impartial jury is, of course, a necessary condition to a fair trial, and a voir dire designed to expose potential bias is essential to ensure an impartial jury.”
The majority tacitly but clearly dispenses with the need for a fully death-qualified jury. Its decision exposes, if it does not create, conflict between the jury’s distinctive responsibilities to determine guilt and punishment. The Court’s decision behooves us to revisit the assumptions that required the use of death-qualified juries in a bifurcated-trial scheme for the prosecution of capital causes. We should again ask the original question: is it constitutionally permissible to try the guilt of a capital defendant with a death-qualified jury?
This Court in Ramseur, supra, 106 N.J. at 248-54, 524 A.2d 188, without extended reasoning, chose to follow the majority in Lockhart v. McCree, 476 U.S. 162, 106 S.Ct. 1758, 90 L.Ed.2d 137 (1986). In that case, however, Justice Marshall in dissent persuasively showed that “the very process of death qualification — which focuses attention on the death penalty .before the trial has even begun — has been found to predispose the jurors that survive it to believe that the defendant is guilty.” Id. at 188, 106 S.Ct. at 1772, 90 L.Ed.2d at 157 (citations omitted). A majority of five justices simply rejected the argument in a footnote. Id. at 170 n. 7, 106 S.Ct. at 1763 n. 7, 90 L.Ed.2d at 146 n. 7. I disagreed then with our Court’s acceptance of the Supreme Court’s view on this matter, Ramseur, 106 N.J. at 428-35, 524 A.2d 188 (Handler, J., dissenting), and have continued to disagree with it. See, e.g., State v. Bey II, supra, 112 N.J. at 191-98, 548 A.2d 887 (Handler, J., dissenting). In both Ramseur, supra, 106 N.J. at 431-32, 524 A.2d 188 (Handler, J., dissenting) and Bey II, supra, 112 N.J. at 191-92, 548 A.2d 887 *222(Handler, J., dissenting), I cited the valid studies and authoritative support in ongoing social science research that demonstrate the effect of death qualification on the propensity of jurors to convict. In Bey II, I wrote:
The process [of death qualification] makes paramount the issue of the imposition of the death penalty itself, thus suggesting that the guilt of the defendant is a foregone conclusion. See Haney, “On the Selection of Capital Juries: The Biasing Effects of the Death Qualification Process,” 8 Law Hum. Behav. 121 (1984). One cannot read the transcript of a death-qualifying voir dire without a sense of impending doom. See State v. Ramseur, 106 N.J. at 331, 341 [524 A.2d 188] (O’Hern, J., concurring) (citations omitted).
[Id. 112 N.J. at 192, 548 A.2d 887.]
Knowledgeable defense attorneys in capital causes recognize, as did Marshall’s attorney and the co-defendant’s attorney, that death qualification predisposes jurors to return a guilty verdict. In capital prosecutions, “defense attorneys are forced to consider whether it is better to minimize questioning of prospective jurors in efforts to limit biases created by death qualification.” Hans, “Death by Jury,” in Challenging Capital Punishment 149, 158 (K. Haas & J. Inciardi, eds., 1988) (hereafter “Hans”). This case adds empirical support for the view that death qualification can engender intolerable levels of bias in the trial of a capital case with respect to the determination of criminal guilt. If knowledgeable and experienced persons, and well-prepared, skilled, highly qualified and extremely vigorous attorneys, as in this case, ante at 165, 169, 586 A.2d at 171, 174, understand that a death-qualified jury is predisposed toward believing a defendant charged with capital-murder is guilty, then, how this Court in good conscience can continue to reject that understanding is a mystery.
I remain convinced that the bifurcated trial methodology, with its use of death-qualified juries, is fundamentally flawed. My conviction is strengthened by this case. The Court has consistently rejected the argument that the process of death qualification conditions the jurors to assume the defendant’s guilt, thus violating the defendant’s right to an impartial jury. See, e.g., State v. Rose, supra, 112 N.J. at 476, 548 A.2d 1058. *223And yet here, the Court implicitly acknowledges the validity of the position it has professed to reject. It thus characterizes defense counsel’s request to limit death qualification as “a well-considered strategic attempt to limit juror exposure to questions concerning capital punishment.” Ante at 93, 586 A.2d at 131. What the Court fails to acknowledge is that its decision, in effect, construes our capital-murder statute to give a defendant a choice between two options: a death-qualified jury to determine guilt or a non-death-qualified jury to determine sentence. To try a defendant with a death-qualified jury is to try him with a jury predisposed toward guilt. To try a defendant with a jury that is not death-qualified is to try him with a jury that may be predisposed toward death. Neither choice is constitutionally tolerable.
There are ways out of this dilemma. In State v. Ramseur, supra, Justice O’Hern contended that a capital defendant should be sentenced by a different jury from the jury that found his guilt. He wrote, “The question is this: is trial by a jury that is not more prone to convict a good cause for a truly bifurcated trial? I think that it is. It is hard to state a better cause.” 106 N.J. at 339, 524 A.2d 188 (O’Hern, J., concurring). I repeat here that “I continue ... to agree with Justice O’Hern ... that the inconvenience entailed in providing for a non-death-qualified jury — one that is fair and impartial — in the trial of guilt is not too high a price to pay to vindicate constitutional interests.” Bey II, supra, 112 N.J. at 198, 548 A.2d 887 (Handler, J., dissenting).
A death sentence must reflect the values of the community. Witherspoon v. Illinois, supra, 391 U.S. at 519, 88 S.Ct. at 1775, 20 L.Ed.2d at 783. Society has a critical stake in the integrity of a death-penalty prosecution. See Barefoot v. Estelle, 463 U.S. 880, 913, 103 S.Ct. 3383, 3405, 77 L.Ed.2d 1090, 1117 (1983) (Marshall, J., dissenting). The public’s interest in a fair and reliable determination that a defendant must die overshadows any separate interest of the State or even of the defendant. It is not up to the defendant to decide whether or *224how he or she will be tried for capital murder or sentenced to death. See, e.g., State v. Koedatich, 112 N.J. 225, 329-32, 548 A.2d 939 (1988), cert. denied, 488 U.S. 1017, 109 S.Ct. 813, 102 L.Ed.2d 803 (1989). A defendant cannot be tried for capital murder or be exposed to the death penalty in any manner that departs from the minimal constitutional standards governing capital-murder prosecutions, which assure that the ultimate sentence of death is one that is not capricious, discriminatory or arbitrary, that can be perceived as reliable and accepted as an expression of the conscience of the community. A capital-murder proceeding that fails to reach such standards is unacceptable to a civilized society.
It remains my view that it is unconstitutional to determine the guilt of a capital defendant with a death-qualified jury. But the constitutional offense is infinitely greater if a death sentence is imposed on a capital defendant by a jury that is not death-qualified. The former risks an unfair conviction; the latter risks an unfair sentence of death. Neither a defendant nor defense counsel have the prerogative to decide whether or not to death-qualify a jury in capital cases. The contrary result — a death sentence imposed by an unqualified jury — can be regarded only as arbitrary, unprincipled, and expedient. Nevertheless, that is precisely the result endorsed by the Court. The failure to death-qualify this capital-murder jury cannot be salvaged.
II.
The defendant was subjected to determinations of criminal guilt and, ultimately, the death sentence by a jury that realistically may well have believed that it could reach those decisions if it simply found that the defense witnesses were not credible and were not telling the truth. That serious misdirection of the jury’s proper deliberative responsibility resulted not merely from incorrect instructions of the trial court that advised the jury that it could resolve ultimate issues of guilt and *225sentence by deciding who was telling the truth; but also from the failure of the State to provide critical evidence bearing on the credibility of a key prosecution witness. Those circumstances present the palpable danger that the jury’s ultimate determinations were based on whom the jury believed, not on whether the evidence convinced the jury of defendant’s guilt beyond a reasonable doubt.
A.
Concededly, the trial court fully explained to the jury the State’s burden of proof. In its instructions, however, the trial court also undermined that explanation by advising that the verdict should constitute the jury’s determination of the “truth”: “So let your verdict declare the truth.” 2
The Court concludes simply that “defendant was not denied his constitutional right to be convicted only on proof beyond a reasonable doubt.” Ante at 136, 586 A.2d at 155. Viewed from a wider perspective, however, the trial court’s instruction did not guarantee that a conviction would be based only on proof beyond a reasonable doubt.
In United States v. Pine, 609 A.2d 106 (1979), a bribery case, the Court of Appeals for the Third Circuit considered the confusion and misdirection likely to result from such a jury charge concerning the “truth.” In Pine, the trial court’s instructions to the jury included the following passage:
Fundamentally, this case, as most eases do, involves the question of fact. The basic question is whether the Government’s witnesses are telling the truth or whether the defendants and their witnesses are telling the truth. Your basic task is to evolve the truth.
*226[Id. at 107-08.]
Observing that such an instruction tends “to dilute and thereby impair the constitutional requirement of proof beyond a reasonable doubt,” the Court of Appeals directed all trial courts in the Third Circuit to desist from using such instructions. Id. at 108. Nevertheless, examining that objectionable instruction in the context of the entire jury charge, the court held that the trial court’s instructions did not violate the constitutional rights of the defendant. Id. at 109.
Of course a single instruction to a jury should not be assessed “in artificial isolation.” Cupp v. Naughten, 414 U.S. 141, 146-47, 94 S.Ct. 396, 400-01, 38 L.Ed.2d 368, 373 (1973); State v. Hunt, supra, 115 N.J. at 373, 558 A.2d 1259. To determine whether the instruction violated defendant’s constitutional rights, the Court should consider the instructions both in the context of the complete jury charge and in the wider perspective of the prosecution.
Here, from the outset of the prosecution, the parties initiated a battle over which side was telling the truth. The prosecutor insinuated in his opening statement that Marshall was a liar and that the credibility of his testimony was a central issue in the case. The opening statement by Marshall’s counsel suggested that Billy Wayne McKinnon was a liar. Marshall’s attorney told the jury, “I ask you to closely listen to [McKinnon], closely scrutinize his demeanor, because you’re going to have to decide in your mind whether you can believe that man beyond a reasonable doubt.” That battle intensified during the evidentiary phase of the trial. McKinnon and Marshall were critical witnesses for the State and the defense, respectively. The jury could not believe the import of either person’s testimony and also believe the other’s story. The defense cross-examined McKinnon at length and vigorously attacked the credibility of his testimony; the State subjected Marshall to similar treatment. McKinnon’s and Marshall’s veracity was challenged through other witnesses as well.
*227The summations sharpened the issue whether McKinnon or Marshall was telling the truth about their relationship, the events leading up to the murder, and what happened at the Oyster Creek Picnic Area. The prosecutor opened his remarks by telling the jury that the case turned on “what witnesses are believable and what witnesses are not believable.” He continued: “[T]he testimony is so diametrically opposed in this case that there can be no question that there has been a conscious effort to deceive you. Make no mistake about that.” He told the jury, “[Y]ou would have not found the truth that you have seen unfold during the course of this trial” if McKinnon had not signed the plea agreement and testified. He repeatedly asserted that the jury would have to find all the State’s witnesses, separately and together, to be liars in order to find the defendant not guilty. Marshall’s attorney’s closing argument also presented the case to the jury as a matter of determining the truth. He told the jury to consider “[w]ho you can believe and who you’re not going to believe, and who comes in here with a motive to lie and who comes in here to tell the truth.”
Thus, prior to the jury charge, the attorneys’ statements and the evidence combined to present the case to the jury as a question of which version of the conspiracy and murder — the State’s or the defense’s — was true. This question was intertwined with the issue of the credibility of the witnesses’ testimony. Against that background the jury instructions augmented the importance of credibility. The court stated that the jury should consider all of the testimony “in determining what is the truth;” and, again, that the jury should “consider anything and everything ... in determining what is the truth.” (Emphasis added.) Finally, any distinction between the search for “truth” and the reasonable-doubt standard became confused when the trial court told the jury that it could consider credibility in deciding whether the State had carried its burden of proof beyond a reasonable doubt. Hence, the references in this jury charge to a duty to find the truth could only obfuscate the correct standard of proof and impair the constitutional require*228ment of proof beyond a reasonable doubt. United States v. Pine, supra, 609 F.2d at 108.
The cumulative emphasis on “truth” throughout the entire guilt phase of this trial created a very real danger that the jury did not render its guilt verdict on the proper legal standard. Most especially in this, a capital case, any legitimate questioning of such error, implicating defendant’s constitutional right to be convicted of capital murder only on proof beyond a reasonable doubt should be resolved in defendant’s favor. This is particularly critical because extraordinarily important evidence bearing directly on the credibility of a key prosecution witness was withheld from the defense, and, ultimately, the jury.
B.
The Court concludes no reversible error inhered in the State’s failure to disclose vital information bearing directly on the credibility of a key prosecution witness, Sarann Kraushaar. Ante at 204, 586 A.2d at 194. I disagree with the Court’s disposition of this issue, particularly in light of the trial court’s strong direction to the jury that it must determine guilt based on who was telling the “truth.”
The State did not dispute that the prosecutor had not disclosed to the defense the documents evidencing the grant of immunity to that witness. Additionally, the State conceded that the material was discoverable and had been improperly withheld. The lower court concluded that the failure to disclose the documents pertaining to the promise of immunity for Kraushaar had not been willful.
The lower court seriously understates the importance of this witness “as being merely a person with-knowledge of relevant facts.” (Emphasis added.) The court acknowledged that Kraushaar had furnished important information. This included “the details of her affair with Robert Marshall, the plan to leave their respective spouses, the financial difficulties that Marshall was having, the fact that insurance on Maria’s life *229would solve these difficulties, [and] the fact that Marshall had inferred a desire to have his wife killed.” This evidence was not only relevant, but also extremely probative of defendant’s motives.
After initially obtaining some of this information on September 7, 1984, the State sought to reconfirm and acquire additional information. Accordingly, it scheduled another interview for September 27, 1984. At that time Kraushaar had an attorney, who insisted on a grant of immunity from prosecution. The State acceded to that request and prepared a letter granting such immunity, which was read to the attorney over the telephone before the interview occurred. Then, a similar letter was prepared and actually signed by the investigating officers at the attorney’s office. The second interview transpired only after the immunity had been secured.
The prosecutor, according to the lower court, simply “placed [the letters] in the Marshall file and, thereafter, assumed that all persons concerned in his office would know of them____ [and] gave no further thought to them.” The lower court then theorized that non-disclosure by the State had been inadvertent. In concluding that the State’s action had not been willful, the court gave significant weight to a “motive” not to withhold this information. That “motive,” however, is pure hypothesis, constructed on the putative strength of the prosecution’s ease.3
*230This Court ordinarily must defer to the fact-finding of the trial judge, who has had the first-hand opportunity to assess the credibility of the witnesses. State v. Johnson, 42 N.J. 146, 161, 199 A.2d 809 (1964). Nevertheless, no facts were adduced to support the conclusion that the State believed its case against defendant was so strong that it had no reason to withhold disclosure of Kraushaar’s grant of immunity. To credit that theory one would also have to impute to the State a belief that the information it obtained from Kraushaar from the second interview was really unnecessary and wholly superfluous and that her trial testimony was either inconsequential or impregnable. The record does not clearly suggest that the prosecution entertained that position or acted on it. In view of the trial court’s reliance on the State’s motive in determining, without adequate support, that nondisclosure was nonwillful, its conclusion should not be credited.
Moreover, “[i]n determining whether prosecutorial nondisclosure violates due process, the good faith or bad faith of the prosecution is irrelevant.” State v. Carter, 91 N.J. 86, 112, 449 A.2d 1280 (1982) (citing Brady v. Maryland, 373 U.S. 83, 87, 83 S.Ct. 1194, 1196, 10 L.Ed.2d 215, 218 (1963)). In the context of a non-capital crime we have recognized that “[wjhere the defendant has made a specific request for information and the prosecution has failed to reveal the requested information, the standard of materiality is whether ‘the suppressed evidence might have affected the outcome of the trial.’ ” Ibid. (quoting United States v. Agurs, 427 U.S. 97, 104, 96 S.Ct. 2392, 2398, 49 L.Ed.2d 342, 350 (1976))., That test “is not translatable into the mere possibility that the undisclosed information might *231have helped the defense____ There must be a real possibility that the evidence would have affected the result.” Id. at 113, 96 S.Ct. at 2402. Because this is a capital case, we should under our State Constitution adopt a more protective standard. See, e.g., United States v. Bagley, 473 U.S. 667, 682, 105 S.Ct. 3375, 3383, 87 L.Ed.2d 481, 494 (1985) (If absence of undisclosed evidence is “sufficient to undermine confidence in the outcome,” nondisclosure is material).
The importance of Kraushaar’s credibility is unquestionable. The undisclosed information relating to her grant of immunity surely would have provided a means to impeach her. On several points, including particularly her testimony about Marshall getting rid of his wife, his financial status, and his anticipation of insurance proceeds, her testimony and the defendant’s were in sharp conflict. Credibility as between those two conflicting witnesses was extraordinarily important. Knowledge of Kraushaar’s immunity from prosecution could have provided a basis for the jury to infer that she was exaggerating her version of the events. In view of the extensive and vigorous cross-examination and sharp attacks on the credibility of other key prosecution witnesses, like McKinnon, it cannot be assumed that defense counsel could not have effectively weakened Kraushaar’s testimony had they known of her grant of immunity. Because the trial court itself, as well as the" prosecutor, put credibility in issue, the result of the trial might have been affected had the defense been able to impeach her credibility. Although none of Kraushaar’s testimony directly placed Marshall in the conspiracy or implicated him in the murder at the Oyster Creek Picnic Area, it clearly furnished evidence of a strong motive to kill his wife. Defendant strenuously challenged this evidence and was denied without just cause the opportunity to challenge the credibility of its author.
C.
In sum, the trial court’s instructions directing the jury to focus on witness credibility and to reach a verdict that reflected *232the “truth” enabled the jury to convict in whole or in part on credibility, aside from the cogency and weight of all of the evidence. There simply is no legal authority that a criminal verdict must mirror the “truth.” That error takes on a special effect in light of the fact that defendant was denied the opportunity fully to challenge the credibility of a key prosecution witness. Particularly, in the context of a capital-murder prosecution, such errors cannot be passed off as innocuous or harmless.
III.
Prosecutorial misconduct was rampant throughout the trial. The Court finds no justification for reversal of defendant’s conviction or death sentence on those grounds. I disagree.
A capital case solemnities and enhances the prosecutorial responsibility “to use every legitimate means to bring about a just [result].” State v. Ramseur, supra, 106 N.J. at 320, 524 A.2d 188 (citations omitted). It also elevates the court’s supervisory duty to assure adherence to the prosecutor’s “particularly stringent ethical obligations in capital cases.” Id. at 324, 524 A.2d 188; see Williams II, supra, 113 N.J. at 456, 550 A.2d 1172.
The .record reveals a pattern of calculation on the part of the prosecutor that underscores the offensiveness of his conduct. The prosecutor constantly made needling remarks obviously designed only to cast aspersions on defendant, defense witnesses, and defense counsel. These gibes, individually and together, do not warrant reversal, but, with other tactics of the prosecutor, they injected a tone of opprobrium and denigration that had no place in this trial. In addition, the prosecutor brought to the attention of the jury, against court instructions and in a manner calculated to prejudice defendant, assorted irrelevant and inflammatory tidbits of information. The prosecutor noted, for example, that Sarann Kraushaar’s father had died of a heart attack a few weeks after the murder, implying *233that it was related to the man’s horror at Kraushaar’s connections to defendant and the murder. In response to pointed admonitions from the court, the prosecutor several times promised the court that he would refrain from speaking about some objectionable topic, but would subsequently broach that topic before defense counsel could respond. Additionally, many of his comments and questions strongly insinuated that defense counsel, as well as defendant and the defense investigator, had conspired to obstruct the investigation in Louisiana. Although the prosecutor often withdrew objectionable remarks, he made them with such frequency that their damage cannot be discounted. In short, to say that the prosecutor was simply spiritedly trying the case or was carried away by the emotional pitch of the trial is naive. Most of those improprieties were calculated.
Against that background, certain instances of the prosecutor’s improper conduct stand out starkly. For example, the trial court ruled that investigator Mahoney could testify only that the September 21 interview with defendant “was terminated,” to avoid disclosure by defendant that he had been advised by counsel not to say anything to law enforcement officials about the case outside the presence of counsel. Yet, the prosecutor persistently placed before the jury, albeit in the form of questions to the witness, the fact that defendant had retained an attorney, in effect, betraying his own guilt.
The Court stresses that defendant introduced evidence of his retention of counsel. Ante at 122-123, 586 A.2d at 147-148. Nevertheless, that does not excuse, justify, or undo the prosecutor’s attempt to penalize defendant’s exercise of his right to counsel by improperly suggesting defendant’s guilt. “Even the most innocent individuals do well to retain counsel.” Bruno v. Rushen, 721 F.2d 1193, 1194-95 (9th Cir.1983), cert. denied sub nom McCarthy v. Bruno, 469 U.S. 920, 105 S.Ct. 302, 83 L.Ed.2d 236 (1984) (citations omitted). “ ‘[Ljawyers in criminal courts are necessities not luxuries.’ ” Gideon v. Wainwright, 372 U.S. 335, 344, 83 S.Ct. 792, 796, 9 L.Ed.2d 799, 805 (1963) *234(citations omitted). Prosecutorial conduct impugning the retention of counsel by a defendant equates with serious trial error. Bruno v. Rushen, supra, 721 F.2d at 1195 (prosecutor’s “insidious” attacks on defendant’s exercise of his right to counsel and on the integrity of defense counsel were errors of constitutional dimension); State v. McDonald, 620 F.2d 559, 564 (5th Cir.1980) (“It is impermissible to attempt to prove a defendant’s guilt by pointing ominously to the fact that he has sought the assistance of counsel”); United States ex rel. Macon v. Yeager, 476 F.2d 613 (3d Cir.1973) (prosecutor’s comment penalized defendant exercising his right to counsel and was thus constitutional error), cert. denied, 414 U.S. 855, 94 S.Ct. 154, 38 L.Ed.2d 104 (1974); see United States v. Daoud, 741 F.2d 478, 481 (1st Cir.1984) (trial court’s refusal of defendant’s request for curative instruction, where prosecutor twice “needlessly” elicited testimony that defendant had requested counsel after issuance of Miranda warnings, was error).
Although the Court attempts to bury this error in other evidence, ante at 126, 586 A.2d at 149, the error cannot be made to disappear. The brevity of the comment in the context of this lengthy trial does not render it harmless. The fact that the reference to defendant’s retention of counsel was in the form of a question rather than a statement does not lessen its prejudicial impact. The trial court itself agreed that the question was argumentative, and the intended effect on the jury was not dependent on whether the question was answered. Most importantly, the prosecutor’s insinuation was calculated and struck at the core of the defense. Bruno v. Rushen, supra, 721 F.2d at 1195. The attempt to prove defendant's guilt by pointing to defendant’s retention of counsel can be harmful per se. State v. McDonald, supra, 620 F.2d at 564. This species of prosecutorial misconduct is reversible error.
The prosecution also implied facts by referring to phantom witnesses who were not called to testify. For example, the financial condition of the defendant’s family at the time of the *235murder was a major issue. The success of either the State’s theory of the case or the defense depended significantly on what the jury believed about defendant’s financial status and planning. The prosecutor asserted, not once but twice, that the figures he had prepared were correct and based on sources in the financial industry, and that if the defendant challenged the prosecutor’s figures, then the prosecutor would bring in witnesses to show that defendant was wrong.
In State v. Rose, supra, 112 N.J. at 518-19, 548 A.2d 1058, this Court found improper “the prosecutor’s statement that he could have produced ‘ten’ experts to testify [concerning defendant’s mental condition] differently from the defense experts.” The prosecutor’s assertions here were no less improper than the statements in Rose, given the importance of defendant’s finances. The prosecutor misled the jury concerning the inferences it could draw. He also improperly pitted his own credibility against defendant’s. Id. at 519, 548 A.2d 1058; see ABA Standards for Criminal Justice § 3-5.8(a) and (b). The trial court’s sustaining of the objections could hardly have made clear to the jurors what they could not infer, nor could it have caused them to erase the implications from their minds.
Furthermore, the prosecutor made remarks of a testimonial character that were unsubstantiated, namely, that defendant had already received six hundred thousand dollars and that in the event of acquittal, “the checks will be in the mail within a week.” Such a blatant misrepresentation of fact is improper. Rose, supra, 112 N.J. at 522, 548 A.2d 1058; ABA Standards for Criminal Justice, § 3-5.8(a) (“It is unprofessional conduct for the prosecutor intentionally to misstate the evidence or mislead the jury as to the inferences it may draw”); New Jersey Rules of Professional Conduct 3.4 (“A lawyer shall not ... (e) in trial, allude to any matter that ... will not be supported by admissible evidence”). Such references divert the jury’s attention from the fact issues in the case. The misrepresentations were not accidental. Indeed, the prosecutor emphatically attested to the truth of his misrepresentations. *236“[I]mproper ... assertions of personal knowledge are apt to carry much weight against the accused when they should properly carry none.” Berger v. United States, 295 U.S. 78, 88, 55 S.Ct. 629, 633, 79 L.Ed. 1314, 1321 (1935); see also Ramseur, supra, 106 N.J. at 321, 524 A.2d 188 (well-established rule that prosecutor may not express personal belief in defendant’s guilt because it diverts jury’s attention from evidence adduced at trial); State v. Thornton, 38 N.J. 380, 398, 185 A.2d 9 (1962) (prosecutor must refrain from expressing personal belief because of danger that his or her official status may improperly influence jury), cert. denied, 374 U.S. 816, 83 S.Ct. 1710, 10 L.Ed.2d 1039 (1963).
Moreover, those comments improperly instructed the jury on the consequences of acquittal, implying that a conviction was the only precaution against defendant’s receipt of insurance proceeds. “The prosecutor should refrain from argument which would divert the jury from its duty to decide the case on the evidence, by injecting issues broader than the guilt or innocence of the accused under the controlling law, or by making predictions of the consequences of the jury’s verdict.” ABA Standards of Criminal Justice § 3-5.8(d). Cf. Darden v. Wainwright, 477 U.S. 168, 179-80, 106 S.Ct. 2464, 2470-71, 91 L.Ed.2d 144, 156-57 (1986) (improper to imply only guarantee against future similar criminal act is death penalty); State v. Koedatich, supra, 112 N.J. at 323-24, 548 A.2d 939 (statement that had defendant not been arrested, “he would have been given a license to kill” constitutes misconduct). In denying defendant’s motions for a mistrial based on those comments, the trial court gave only a very brief jury instruction, which did not refute squarely the prosecutor’s avowal that defendant, if acquitted, would quickly receive insurance money.
In addition, the prosecutor made improper references to his experience, attempting to influence the jury’s perception of the evidence. He added his own testimony to the evidence concerning the size and shape of the supposed mail depository, a factual issue that was disputed throughout the trial, the resolu*237tion of which influenced the determination of many of the., evidentiary issues and inferences. Ramseur, supra, 106 N.J. at 321, 524 A.2d 188. The prosecutor’s reference to his experience in the middle of summation might well have impermissibly placed the prosecutor’s authority behind his summation arguments concerning many other evidentiary disputes. Id. at 320, 524 A.2d 188. “[S]uch statements may add the weight of the prosecutor’s official and personal influence and knowledge to the probative force of the evidence adduced, thus creating the possibility that the jurors consciously or unconsciously might adopt the prosecutor’s view without applying their own independent judgment to the evidence.” State v. Thornton, supra, 38 N.J. at 398, 185 A.2d 9.
Further, in cross-examination, counsel for both Marshall and Thompson combed McKinnon’s plea agreement and statement, attempting to call into question both McKinnon’s motives behind his agreement and the substance of his version of the conspiracy and murder. The prosecutor countered that legitimate defense tactic in his closing argument by vouching for the truth of McKinnon’s testimony and representing to the jury facts — the truth of which cannot be known — about the defense’s cross-examination that had nothing to do with the evidence in the case. He also insinuated that defense counsel was acting improperly in challenging the plea-agreement and the reliability of McKinnon’s testimony, observing, again on his experience, that the defense was a sham.
The Court treats these several circumstances dismissively, ante at 153-155; 156-159, 586 A.2d at 165-166; 166-168, yet diverting the jury’s attention from the facts of the case is clearly improper. Ramseur, supra, 106 N.J. at 322, 524 A.2d 188; Thornton, supra, 38 N.J. at 398, 185 A.2d 9; ABA Standards for Criminal Justice § 3-5.8(a). A prosecutor cannot allude to matters not supported by evidence admitted at trial. See New Jersey Rules of Professional Conduct 3.4.
Further, the prosecutor stated in summation:
*238And he has the audacity to bring in his three boys to testify. That’s obscene. And I’m not being critical of them, because I would probably do the same thing. To put his boys on that witness stand is obscene, and for that there’s a place in hell for him. He will use anybody, he will say anything and he will do anything, including his own family, to get out from under. And that’s Kobert Oakley Marshall. Make no mistake about it.
After the summation, defendant moved for a mistrial based on those remarks. The trial court simply instructed the jury that defendant “has a right to bring in any witnesses ... to testify on his behalf and that no adverse inferences should be drawn against the defendant merely because his sons testified as witnesses on his behalf.” Those instructions hardly palliated the “obscenity” that the prosecutor ascribed to defendant. The prosecutor’s comments were studied, coldly calculated remarks designed to prejudice defendant. Such forceful and vivid remarks carried an overwhelming capacity to divert the jury’s attention from the proper fact issues in the case. See, e.g., Ramseur, supra, 106 N.J. at 322, 524 A.2d 188.
Additionally, in his sentencing-phase statement to the jury, the prosecutor expressed his personal view that there is nothing “more heinous in our society than to ... hire somebody to kill somebody else,” particularly “a family member” such as “your wife.” He added that the victim, “had no prior criminal history” and was “civic-minded” but that “defendant did not give her the option of thirty years.”
Those comments were clearly improper. What the prosecutor thinks is the most heinous crime in our society is not relevant to the issues properly before the jury at sentencing; it has the effect of indirectly adding to the penalty trial an aggravating factor, c(4)(c), that was not before the jury, and in phraseology expressly disapproved by this Court. See Ramseur, supra, 106 N.J. at 321, 524 A.2d 188; Thornton, supra, 38 N.J. at 398, 185 A.2d 9. Further, the victim’s lack of a criminal record and civic-mindedness were not relevant to defendant’s sentence. In effect, the prosecutor rated the victim under the mitigating factors sought by defendant, subtly but impermissibly diverting the jury’s consideration of these mitigating *239factors from defendant’s character to the character of defendant’s wife. It encouraged the jury to find that if his wife was civic-minded and law-abiding, the mitigating factor would not be available to defendant, leaving him unworthy of a life sentence. These insidious remarks, in the context of the unusual, highly abbreviated sentencing phase proceedings, denied defendant the protections required in a fair penalty proceeding.
The question remains whether that stream of misconduct warrants reversals respectively of the murder conviction and the death sentence. “The determination that the prosecutor was guilty of misconduct does not end [the] inquiry.” Ramseur, supra, 106 N.J. at 322, 524 A.2d 188. None of the instances of misconduct that occurred in the guilt trial, as detailed here, can arguably be described as a slip of the tongue, an unavoidable consequence of the charged atmosphere of a criminal trial, or a mistake committed in good faith. To the contrary, those instances were clearly deliberate attempts to place matters not pertinent to defendant’s guilt or innocence before the jury. The few and somewhat meager attempts by the court and counsel to correct the damage wrought by these acts of misconduct were inadequate, and some of that misconduct, and certainly its cumulative effect, was incurable.
The Court greatly misperceives the impact of this misconduct. Ante at 170, 586 A.2d at 174. Prosecutorial misconduct at the guilt phase requires reversal of the conviction of capital-murder. The prejudice resulting from that misconduct was escalated by the prejudice that arose from the prosecutor’s improper penalty-phase remarks, warranting reversal of defendant’s death sentence.
IV.
This case underscores the deficiencies that can, and often do, surround the imposition of the death sentence by invoking the bifurcated trial methodology, which entails the continuous but ostensibly separate trials of guilt and punishment. The use in *240both proceedings of the same jury achieves continuity. The application of different substantive standards and procedural rules in each proceeding achieves separateness. “[Djeath-qualification and death-selection are discrete, successive determinations, each to be made by the jury under appropriate standards.” Ramseur, supra, 106 N.J. at 392-93, 524 A.2d 188 (Handler, J., dissenting). Continuity between the guilt and penalty trials serves only administrative convenience. The purpose of separateness of the penalty hearing is to assure that the penalty determination is independent and objective, and truly focused on the ultimate issue of the death-worthiness of the defendant. In this case, the penalty-phase aspect of the trial was fundamentally unfair because it failed wholly to separate the jury’s responsibilities with respect to guilt and punishment.
A.
Defendant’s death sentence was based solely on one aggravating factor, c(4)(e), which duplicated an element of the crime for which he was convicted. Defendant was charged with capital murder consisting of a murder by contract. N.J.S.A. 2C:ll-3. Defendant claims this statutory structure in which the capital crime, the death-eligible crime, is coextensive with a death-worthy crime is unconstitutional because it fails to narrow the class of murderers eligible for the death penalty. The Court today rejects that contention on the strength of Ramseur, supra, 106 N.J. at 188, 524 A.2d 188. Ante at 137-138, 586 A.2d at 156. I continue to disagree with the Court’s position on constitutional grounds. “[A]s a matter of its basic definition of murder the statute effectively telescopes the definitional narrowing of the class of death-eligible murderers with the jury’s finding of aggravating factors at sentencing.” Ramseur, supra, 106 N.J. at 403, 524 A.2d 188 (Handler, J., dissenting).
*241The shortcoming in the statute in failing to narrow sufficiently the class of capital-murder defendants eligible to receive the death sentence before the penalty trial commences is particularly significant in this case. The jury here was asked to determine in the guilt trial whether defendant was guilty of the murder of his wife through the payment of consideration to another person. That element — the payment of consideration— was the sole aggravating factor that the State relied on in seeking the death penalty. In the wake of the multitudinous circumstances and the resolution of numerous factual issues, as well as the ultimate determination of guilt, at the guilt phase of this trial, the penalty phase did not genuinely require or allow the jury to render an additional, significantly different decision concerning defendant’s death-worthiness. The penalty phase simply asked the jury to reconsider the evidence it had already assessed at the guilt phase. See ante at 61, 586 A.2d at 114.
To think that the jury could be sufficiently objective in its second assessment of such evidence is unrealistic and naive. It has been observed that “[cjapital jurors begin the penalty phase with a story developed during the guilt phase____ [Ojnce the story has been developed, jurors are likely to resist reconstruction of it.” Hans, supra, at 162. We have recognized in other contexts the perils entailed in asking a jury to redetermine evidence already considered and resolved. If a jury must make a second, independent determination of guilt, it may not simply reconsider evidence previously considered, weighed and resolved in reaching a previous verdict without understanding fully that its reconsideration of such evidence must be fresh and independent. State v. Ingenito, 87 N.J. 204, 432 A.2d 912 (1981). “It becomes essential, therefore, that the jury be instructed in no uncertain terms to consider anew the evidence previously admitted but to disregard completely its prior verdict.” State v. Ragland, 105 N.J. 189, 195, 519 A.2d 1361 (1986). When the redetermination is the basis for the imposition of the death penalty, we may reasonably doubt the efficacy *242of simple, pro forma instructions that allow the jury merely to churn previously digested evidence.
In this case, there is little assurance that the instructions of the trial court were at all effective. The guilt and penalty determinations were in effect collapsed into each other because the identical statutory element defined both the crime and the punishment. Also, as noted, the critical evidence that converted this crime into capital murder was the identical — and only— evidence that could justify the death sentence.
Other equally serious deficiencies occurred in the imposition of the death sentence. The fusion of crime and punishment was intensified because the penalty determination was so perfunctory as to be indistinct from the jury’s guilt determination. Even overlooking the problems inherent in the bifurcated trial system, this trial was not, as it should have been, truly bifurcated.
A short sentencing phase will not, in itself, necessarily impugn the ultimate sentencing determination, but here neither the State nor defendant developed any independent evidentiary record in the penalty phase. The only evidence available and used in the penalty phase to seek the death sentence was the evidence that was probative on the issue of guilt. Defense counsel relied solely on evidence presented in defense of the crime charged. There is no reason to suppose that the evidence offered by defendant at the guilt trial by way of a defense to the charges could be more than marginally helpful with respect to whether he deserved only life imprisonment. During the defense case in the guilt trial, the defense witnesses had testified in no particular order and with no particular focus on defendant’s death-worthiness. Such an unstructured and loose approach to the mitigating factors is unacceptable given the “ultimate” decision, see State v. Biegenwald, supra, 106 N.J. at 62, 524 A.2d 130, for which the proceeding is designed to provide guidance.
*243Further, the problem of the scant and, at the time of the sentencing phase, distant mitigating evidence was compounded by the timing of the sentencing phase. The entire sentencing phase took place the same day the jury returned its guilt verdict, with the awareness of its guilt determination still vivid. Indeed, the sentencing phase began only a few hours after the guilt phase had been completed. The record of the sentencing phase including the court’s opening remarks to its closing instructions requires only seventeen pages of transcript. This record conveys no sense of thoroughness and only a superficial sense of gravity.
It has been argued that a capital defendant’s right to a continuance between the guilt and sentencing phases is of constitutional magnitude. See Note, “A Capital Defendant’s Right to a Continuance Between the Two Phases of a Death Penalty Trial,” 64 N.Y.U.L.Rev. 579 (1989). A continuance is advisable partly because the defendant, the jury, and defense counsel need time to recover emotionally from the capital conviction and to focus on the issue of sentencing. Id. at 582. This observation is not an idle one. On return of the guilty verdict at this trial, defendant collapsed, was taken from the court on a stretcher, and then transported to a hospital. Less than two and a half hours later, defendant was returned to court to begin the sentencing phase, which lasted approximately twenty minutes. After deliberating for about an hour and a half, the jury returned a sentence of death.
It has been further noted that on a substantive level, defense attorneys are not capable of adequately preparing for the two trials — guilt and sentencing — simultaneously. Ibid.; Goodpaster, “The Adversary System, Advocacy, and Effective Assistance of Counsel in Criminal Cases,” 14 N.Y.U.Rev.L. & Soc. Change 59, 83-85 (1986) (“Penalties in capital cases ultimately will turn on mitigating evidence and on the advocate’s ability to marshal and present that evidence. As a matter of law and practice, the opportunity to present almost any arguably mitigating evidence crucially distinguishes death penalty trials *244from all other criminal trials.”); see also Proffitt v. Wainwright, 685 F. 2d 1227, 1271 (11th Cir.1982) (Clark, J., concurring and dissenting) (quoting A.B.A. Project on Standards for Criminal Justice, Standards Relating to: The Prosecution Function and the Defense Function 277 (1970): “The lawyer ... has a substantial and very important role to perform in raising mitigating factors both to the prosecutor initially and to the court at sentencing.”), cert. denied, 464 U.S. 1002, 104 S.Ct. 508, 78 L.Ed.2d 697 (1983). It is understandable that some states forbid courts from holding the sentencing hearing immediately following the return of the guilty verdict. Note, “A Capital Defendant’s Right to a Continuance,” supra, 64 N.Y.U.L.Rev. at 623-24.
Newly introduced mitigating evidence might have firmly refocused the jury’s attention and closely attuned the jury to the meaning and function of the mitigating factors. Instead, the absence of such evidence, coupled with the celerity and brevity of the sentencing trial, left open the possibility that the jury did not fully and properly set aside its fixed impressions and the impact of its guilt deliberations. A truly separate and distinct penalty trial theoretically succeeds in infusing guilt-phase evidence with a different hue because it is filtered through the prism of the special protective standards that govern the imposition of the death sentence. The integrity of that process is dubious, even under the best circumstances. See, e.g., State v. Harvey, 121 N.J. 407, 449, 581 A.2d 483 (1991) (Handler, J., dissenting and concurring) (other-crimes evidence marginally admissible in guilt phase may be extremely prejudicial in penalty phase; hence, “it is imperative that the potential for prejudice of such evidence in the penalty-phase trial be considered by the trial court in determining its admissibility in the guilt-phase trial.”); State v. Long, 119 N.J. 439, 525-17, 575 A.2d 435 (1990) (Handler, J., dissenting and concurring) (same); State v. Pennington, 119 N.J. 547, 607-08, 575 A.2d 816 (1990) (Handler, J., dissenting and concurring) (prior murder conviction admitted in guilt phase to affect only credibility gains added prejudicial *245weight when used a second time as an aggravating factor in penalty phase); Hans, supra, at 162. The truncated sentencing proceeding in this case, however, defeats the theory that it is fair to use guilt-phase evidence to determine whether defendant deserves to die. I repeat an observation made in Ramseur:
This is not to deny that the weighing of aggravating and mitigating circumstances “provides the additional restraint on jury discretion that the petitioner in [Zant v. Stephens, 462 U.S. 862, 103 S.Ct. 2733, 77 L.Ed.2d 235 (1983)] argued was constitutionally necessary.” ... The constitutional infirmity is not in a weighing process. Rather, it inheres in the fact that weighing does double duty. It is used in the factfinding necessary to determine whether a murder is capital in the same proceeding and deliberation that determines whether the murderer should be executed. Any distinction between finding an aggravating circumstance and weighing it when the tasks are performed in the same proceeding is at best, in my view, academic; the potential for jury misguidance and arbitrariness is simply too great.
[Ramseur, supra, 106 N.J. at 391-92, 524 A.2d 188 (Handler, J., dissenting) (footnote omitted).]
Pertinent to this matter is the further observation that “the jury’s consideration of aggravating factors serves both to specify which defendants are in the class and, in the same process, to decide their punishment____ [FJrom the defendant’s perspective, the sentence is imposed as and when the offense is defined” and determined. Id. at 393, 524 A.2d 188 (Handler, J., dissenting).
The guilt determination overwhelmed the penalty determination. A death sentence that does not entail a decision that is reliably qualitatively different from its predicate guilt verdict cannot be sustained.
B.
In addition, the trial court’s failure to amplify the meaning and function of mitigating factors exacerbated the prejudice inherent in the abbreviated sentencing phase of this trial. The trial court’s instructions during the sentencing phase were erroneous in that they did not explain adequately the significance and function of mitigating factors. Thus, in its introductory remarks during the sentencing phase, the trial court in*246formed the jury that the defendant need not prove the existence of mitigating factors but, rather, need only introduce evidence of them in order for the jury to consider them. The court simply named the two mitigating factors sought by defendant and then told the jury it had to find one of them, no history of criminal activity. The trial court did not, however, explain the meaning of these mitigating factors. Similarly, in its closing sentencing phase instructions, the trial court repeated that the jury did not have to be convinced of the existence of the mitigating factors, but that if defendant introduced any evidence related to those factors — even though the court was aware that no such evidence would be forthcoming — the jury had to consider the factors and weigh them against any aggravating factors. Again, the court did not discuss the meaning of the mitigating factors in the sentencing determination.
Thus, the court’s only reference to the function of mitigating factors was indirect, a by-product of its explanation of the weighing process. At no time in the sentencing phase did the trial court express or expound the thought that “mitigating factors are those which would tend toward the sentence of life imprisonment.” See Bey II, supra, 112 N.J. at 167, 548 A.2d 887. “[I]n brief, the charge of mitigating factors was essentially a recitation of the statutory language.” Id. at 168, 548 A.2d 887.
The Court has clearly and emphatically imposed a requirement that mitigating factors be fully explained to the jury. Bey II, supra, 112 N.J. at 169-70, 548 A.2d 887. The trial court here did inform the jury that it could consider evidence from the guilt phase in determining sentence, but the court did not specifically say that the jury could consider mitigating evidence adduced at the guilt phase. The fact that defense counsel attempted to explain to the jury the meaning and function of one of the two mitigating factors does not redeem the inadequate instructions of the trial court. Moreover, as previously mentioned, the evidence that the jury considered was offered for a different purpose and in a different context, *247namely, as it bore on guilt. Furthermore, most significantly, the jury had already considered and discounted that evidence, determining obviously that it did not ameliorate defendant’s guilt.
We cannot be confident that the finding of mitigating factors, under those circumstances, demonstrates that the jury understood their meaning or function, absent clarification by the court. This deficiency is especially serious because the jury must not only find the existence of mitigating factors, it must weigh them and further compare their weight to that of aggravating factors. That delicate and complex deliberative process cannot be fulfilled without accurate, clear, and comprehensive instructions. The absence of an explanation of the function and meaning of mitigating factors at the sentencing phase raises the substantive risk that the jury arrived at a sentence in an arbitrary and capricious — and therefore impermissible — manner.
C.
The foregoing errors compound another significant error in the penalty determination. In Ramseur, this Court stated that in capital sentencing it is of paramount importance that “the jury is aware, not simply of the consequences of its actions, but of its total responsibility for the judgment.” 106 N.J. at 316, 524 A.2d 188. The Court stated further that “jury instructions in capital cases should never lead the sentencer to believe that responsibility for determining the appropriateness of defendant’s death rests elsewhere.” Ibid. (citing Caldwell v. Mississippi, 472 U.S. 320, 105 S.Ct. 2633, 86 L.Ed.2d 231 (1985)). Correct jury instructions are critically important in capital eases “because of the jury’s responsibility to decide whether a defendant shall live or die.” Bey II, supra, 112 N.J. at 162, 548 A.2d 887. “[I]n its charge to the jury in the sentencing phase of a capital trial, a trial court must be careful not to dilute the jury’s *248sense of responsibility for determining the appropriateness of the death penalty.” Ibid, (citations omitted).
In its introductory sentencing phase remarks concerning the jury’s determination of sentence, the trial court said: “You have already returned a verdict in which you concluded beyond a reasonable doubt that the defendant is guilty of the murder____ [W]hat is presented here will be concerned with whether or not there are factors which, on balance, lead you to conclude that the defendant should suffer the death penalty.” The court then identified the aggravating and mitigating factors at issue and the burdens of proof with respect to those factors. Arguments before the jury followed.
The court began its final instructions with the remark, “[N]ow you have to decide whether or not the defendant is to be sentenced to death. Quite obviously, your ultimate decision in that regard is extremely important, both from the perspective of the State of New Jersey and from the viewpoint of the defendant.” The court acknowledged the difficulty of the jury’s task, but expressed confidence in the jury’s ability. The court then stated: “The defendant must be sentenced to death if you are satisfied beyond a reasonable doubt that an aggravating factor exists and that such aggravating factor, as you find to exist beyond a reasonable doubt, outweighs all mitigating factors.” After reviewing the reasonable-doubt standard, the court explained that not finding the aggravating factor would result in a prison sentence and that “the defendant will be sentenced to death only if you are convinced beyond a reasonable doubt that the aggravating factor exists and outweighs the mitigating factors, and all of the mitigating factors, beyond a reasonable doubt.” The court concluded by calling the jury’s attention to the special verdict form, which the court described as “self-explanatory.”
The record discloses that the trial court twice indicated that the ultimate issue before the jury was whether defendant should be put to death. The court also recognized the difficulty *249of that decision. Those instructions are deficient, however, because they cast the difficult, ultimate issue largely in terms of a mechanical process. The crucial point that the court failed to communicate anywhere in the sentencing charge was that to arrive at a death verdict would be to make a “normative judgment that death is ‘the fitting and appropriate punishment.’ ” Bey II, supra, 112 N.J. at 162, 548 A.2d 887 (quoting Ramseur, supra, 106 N.J. at 316 n. 80, 524 A.2d 188).
In Ramseur, this Court disapproved of instructions that included language to the effect that the jury was “simply” fact-finding and weighing and “merely” applying the law. 106 N.J. at 315-16, 524 A.2d 188. Although the trial court in this case did not use words such as “simply” or “merely” in its instructions, the effect of the charge was no different from that in Ramseur. In Bey II, the sentencing instructions impermissibly left the impression that “the mechanics of the statute or the ‘law,’ ” not the jury, was “ultimately responsible for the imposition of the death penalty.” 112 N.J. at 164, 548 A.2d 887. The instructions in that case suggested in part that a finding on the verdict form of “no mitigating factors” resulted in a death sentence. Id. at 163-64, 548 A.2d 887. The trial court here gave no such instruction, because the jury was required to find one of the two mitigating factors. As the record indicates, however, the instructions suggested that certain findings by the jury would lead irretrievably to certain sentences. The court should have, instead, expressly instructed the jury that a finding that the aggravating factor outweighed beyond a reasonable doubt all the mitigating factors meant that the jury considered death a fitting and appropriate punishment for the defendant. See id. at 164, 548 A.2d 887. “Such a charge would have ‘suitably directed’ any belief of a juror about the inappropriateness of the death penalty to one or more mitigating factors.” Bey II, supra, 112 N.J. at 164, 548 A.2d 887 (quoting Gregg v. Georgia, 428 U.S. 153, 189, 96 S.Ct. 2909, 2932, 49 L.Ed.2d 859, 883 (1976)).
*250In this case, “it is not clear that the jury understood that it bore the burden of deciding whether to impose the death sentence and that it was not relieved by some statutory scale as the ultimate arbiter of defendant’s life.” Ibid. The trial court’s comment, “Should you find that any aggravating factor exists, and beyond a reasonable doubt that this aggravating factor outweighs all mitigating factors, then it’s the responsibility of the Judge to impose the death penalty in this case,” further diluted the jury’s sense of responsibility in its deliberations. Particularly when coupled with a failure to explain mitigating factors, “any instruction that ‘tend[s] to dilute the jury’s sense of responsibility in passing on the issue of life or death’ is erroneous” and constitutes prejudicial error. Ramseur, supra, 106 N.J. at 316, 524 A.2d 188 (quoting State v. Mount, 30 N.J. 195, 214, 152 A.2d 343 (1959)).
Furthermore, the Court did not sufficiently convey to the jury that a possible and acceptable verdict was a non-unanimous determination. See State v. Ramseur, supra, 106 N.J. at 301, 524 A.2d 188. Finally, because the record and verdict in the guilt phase were determinative in the penalty phase, the trial court’s earlier failure to assure that the jury clearly understood the State's correct burden of proof acquired enormous additional weight in the penalty phase. The jury, as previously discussed, was authorized and encouraged to resolve defendant's guilt by deciding credibility, determining who was telling the truth. It cannot be assumed then that credibility — a belief concerning who was telling the truth — did not also influence the jury in determining punishment.
D.
This case forces us again to confront the arbitrary results engendered by unbounded prosecutorial discretion. The Court has recognized this issue in its decisions. See, e.g., State v. Koedatich, supra, 112 N.J. at 250-58, 548 A.2d 939; State v. Ramseur, supra, 106 N.J. at 329, 524 A.2d 188. In several *251cases I have condemned the “virtually unfettered power of prosecutors to select who is to be prosecuted for capital murder.” State v. Di Frisco, 118 N.J. 253, 303, 571 A.2d 914 (1990) (Handler, J., dissenting and concurring) (quoting State v. Matulewicz, 115 N.J. 191, 207, 557 A.2d 1001 (1989) (Handler, J., concurring)); see also State v. Kiett, 121 N.J. 483, 499, 582 A.2d 630 (1990) (Handler, J., dissenting and concurring) (prosecutor’s decision to allow defendant to seek to avoid imposition of death penalty, even though defendant pled guilty to capital murder, underscores need for uniform standards governing plea agreements). Di Frisco presented a particularly egregious example of the perils of lack of prosecutorial guidance. In that contract-murder case, the State prosecuted the defendant triggerman for capital murder without even seeking an indictment against the higher-up. 118 N.J. at 259, 571 A.2d 914. This case presents a different twist on prosecution in a contract-murder case. Here, the State sought the death penalty against both Marshall and the alleged gunman, Larry Thompson. The record, however, indicates that the prosecutor pursued Marshall with unrelenting fervor, while exercising considerably more restraint with Thompson. This imbalance of zeal is perhaps best dramatized in the prosecutor’s closing argument. After rebutting defense counsels’ closing arguments (during which he admittedly argued for the conviction of both Marshall and Thompson), the prosecutor argued the State’s case. He called strenuously for the conviction of Marshall; not once did he mention Thompson. Most tellingly, the prosecutor informed the jury who he felt was more deserving of conviction:
Don’t be fooled by [Marshall’s] tears. He cries for no one but himself. Don’t be fooled by his “I love you” signs, because he loves no one but himself. I’ll sit across a table from Larry Thompson any day, because with Larry Thompson, you see what you get. What you see is what you bargain for. But this defendant is a coward, he’s self-centered, he’s greedy, he’s desperate, he’s materialistic, and he’s a liar. He is a legend in his own mind. No one has the audacity to question Robert Oakley Marshall as per Robert Oakley Marshall. He is an outstanding citizen in the community, and no one would ever dare to point a finger at him. And if it comes down to Robert Oakley Marshall’s testimony against Billy Wayne McKinnon, no juror in their right mind would *252accept the testimony of Billy Wayne McKinnon over such an outstanding citizen. That is what is sitting in that chair. Make no mistake about it. And that’s the pill that he wants you to swallow.
Thompson was acquitted of all charges. Because the prosecutor targeted Marshall as the more offensive defendant, Marshall’s conviction and death sentence particularly exemplify the arbitrariness and disproportionality resulting from unbounded prosecutorial discretion. Like the decision concerning whom to prosecute, the prosecutor’s decision to target Marshall may be “conscientious, but [is] nonetheless ... highly subjective and speculative.” Matulewicz, supra, 115 N.J. at 207-08, 557 A.2d 1001 (Handler, J., concurring).
As I have noted in other cases, the absence of guidance for prosecutors closely relates to the importance of thorough, mandatory proportionality review. See, e.g., Kiett, supra, 121 N.J. at 511, 582 A.2d 630 (Handler, J., dissenting and concurring); Di Frisco, supra, 118 N.J. at 302-05, 571 A.2d 914 (Handler, J., dissenting and concurring); Matulewicz, supra, 115 N.J. at 206-09, 557 A.2d 1001 (dissenting and concurring); State v. Gerald, 113 N.J. 40, 153-67, 549 A.2d 792 (1988) (Handler, J., dissenting and concurring). The Court’s affirmance of this death sentence renders such review critical. As I stated in Kiett, “there are indications that the prosecution of [death penalty] cases in our State is arbitrary____ Certainly there is no evidence on which to conclude with confidence that the selection and prosecution of death penalty cases in this State is consistent and uniform.” 121 N.J. at 511-12, 582 A.2d 630 (Handler, J., dissenting and concurring).
E.
In sum, the pervasive flaws of the penalty phase of the trial rob the death sentence of any vestige of reliability. Because the critical element defining capital murder was identical to the aggravating factor warranting the death sentence, the statute could not and did not serve effectively to separate the determination of guilt from the determination that death was the *253appropriate sentence. That statutory defect was compounded by the absence of a separate and distinct sentencing trial. No . additional evidence was introduced at the penalty proceeding; the same evidence was used to prove guilt and to obtain the death sentence; and the court did not attempt to reorient the jury and to explain fully the significance of mitigating factors. Further, the jury did not fully understand that the imposition of the death sentence cannot entail simply a mechanical weighing of evidence. We cannot under such circumstances suffer this death sentence.
V.
Overall, the majority’s opinion acknowledges almost a dozen errors in the course of this trial. Remarkably, it finds none of the conceded errors, separately or cumulatively, to be reversible. It follows what has become a familiar appellate tactic: divide and discount the errors. See, e.g., State v. Hightower, 120 N.J. 378, 577 A.2d 99 (1990); State v. Koedatich, supra, 112 N.J. 225, 548 A.2d 939. It also invokes another, more subtle tactic: it characterizes some incorrect trial rulings as “discretionary” rather than as error, e.g., the admission of inflammatory post-murder conduct of defendant and prosecutorial comment on such evidence, ante at 127-128, 586 A.2d at 150-151. It thus avoids consideration of the prejudicial impact of such rulings either singly or cumulatively. I strongly disagree with the Court’s approach and particularly with its conclusion that cumulatively those errors and “improper” rulings did not taint the conviction and death sentence.
The Court expressed its standard of appellate review of trial errors in capital cases in State v. Bey, 112 N.J. 45, 548 A.2d 846 (1988) (Bey I):
Thus, in assessing the impact of error in either the guilt or penalty phase of a capital case, we shall continue to determine reversibility on the basis of a qualitative determination that considers, in the context of the entire case, whether the error was clearly capable of affecting either the verdict or the sentence. The only exception involves “constitutional violations * * * [that] by *254their very nature cast so much doubt on the fairness of the trial process that, as a matter of law, they can never be considered harmless.” Satterwhite v. Texas, 486 U.S. 249, 108 S.Ct. 1792, 1795, 100 L.Ed.2d 284 (1988).
[Id. 112 N.J. at 94-95, 548 A.2d 846.]
It is clear that the Court’s standard for determining reversible error is no different from that governing ordinary cases. Ibid. However, even under the Court’s conventional standard of review, its conclusion that cumulative error in this case is not reversible is incomprehensible. For example, the Court finds the prosecutor’s remarks constituted violations of defendant’s privilege against self-incrimination but, “in the context of the entire trial and the abundant evidence of guilt,” the remarks were “harmless beyond a reasonable doubt.” Ante at 121, 586 A.2d at 147. Because “evidence of defendant's guilt was so persuasive,” the Court believes it is “virtually impossible” that the prosecutor’s references to defendant’s retention of counsel, “however reprehensible^] could have contributed significantly to the jury’s determination of guilt.” Ante at 125, 586 A.2d at 149. The Court claims that “any significantly-prejudicial impact” from prosecutorial remarks that were “improper” or “that exceeded the bounds of permissible argument” was “ameliorated by the trial court’s curative instructions.” Ante at 159, 586 A.2d at 168.
The Court, ante at 169, 586 A.2d at 174, notes the observation in State v. Orecchio, 16 N.J. 125, 129, 106 A.2d 541 (1954):
The sound administration of criminal justice in our democracy requires that both the end and the means be just. The accused, no matter how abhorrent the offense charged nor how seemingly evident the guilt, is entitled to a fair trial surrounded by the substantive and procedural safeguards which have stood for centuries as bulwarks of liberty in English speaking countries. This, of course, does not mean that the incidental legal errors, which creep into the trial but do not prejudice the rights of the accused or make the proceedings unfair, may be invoked to upset an otherwise valid conviction; under these circumstances it would be grossly unjust to the State and its people to grant a new trial____
The errors here did not “creep into the trial”; they were repeatedly bootlegged into the trial. The errors were not “incidental legal errors,” but were calculated to generate prose*255cutorial leverage, and created unremittent prejudice and enormous unfairness to defendant.
The Court’s standard of review does not suffice to meet the high constitutional requirements of due process or the protections of fundamental fairness that attend capital-murder prosecutions. I am convinced that in capital cases we cannot rely on the conventional tests for determining reversible error. Because we are dealing with a capital crime, which more fundamentally addresses through the jury the conscience of the community, it is appropriate to invoke a test that is more strict than one that asks only whether an error affected or contributed to a guilty verdict, or whether the error produced an “unjust result.” The Court should resort to an enhanced standard that focuses attention on the jury’s deliberations rather than on whether the error contributed to the verdict or the result reached was unjust. Williams II, supra, 113 N.J. at 460, 550 A.2d 1172 (Handler, J., concurring).
Thus, my differences with the Court in this case, as in prior cases, reflect the fact that the Court has employed a more indulgent substantive standard for determining reversibility than should be permitted in the direct appeal of capital-murder convictions. We should focus not on results in capital cases but on the integrity of the procedures that produce them. Result-oriented tests such as “harmless error” call essentially for a quantification of evidence, a comparing and weighing of proper evidence with improper evidence, to determine simply whether the former can satisfactorily account for and explain the jury’s determination without the latter. The assumption of that test, perhaps well-grounded for the generality of criminal appeals, is that the jury’s determination of guilt can be viewed as primarily the consideration and assessment of evidence and only tangentially as the exercise of conscience. That assumption must be discarded in a capital-murder prosecution, the only proceeding in which the jury is not told that it should find the facts with no thought for the ultimate punishment.
*256The majority’s result illustrates how spongy and elastic the standard for harmless error can become, and how easily courts can succumb, in capital cases, to the temptation to defer passively to trial-level judgments. Even under the Court’s more relaxed standard of review, reversal is required. It cannot be concluded beyond a reasonable doubt that the errors, each separately discounted by the majority, did not collectively contribute to the jury’s verdict of guilty and its ultimate death sentence.
VII.
As I stated at the outset, this case exposes starkly the irredeemable and irremediable deficiencies in our capital-murder jurisprudence. The result reached by the Court today — the affirmance of this death sentence — is not defensible. It is explainable only on the unacknowledged thesis that standards governing capital-murder prosecutions do not in practice maximize or heighten the protections of a defendant. The standards, as applied in this case, do not in any way surpass those that surround a defendant in the prosecution of an ordinary criminal case. We give the lie to the axiom that “death is different.”
Although the Court professes to grant capital-murder defendants enhanced protections under the Constitution and statute, it fails to deliver such protection. Consider in this case alone that an unqualified jury convicted defendant of capital murder and sentenced him to death; the jury was allowed to find defendant guilty and deserving of the death penalty by considering whether the defense was telling the truth, while at the same time defendant was denied the real opportunity to challenge fully the truthfulness of a key prosecution witness; and the entire trial was distorted by pervasive prosecutorial misconduct. The prosecution was attended by numerous conceded errors and near-errors, which in human reason and common experience, would in the aggregate surely color the jury’s *257deliberations. Finally, the death sentence itself was totally unreliable because the capital-murder conviction and the sentence were governed essentially by identical factors, based on the same evidence, and determined virtually at the same time. Further, the penalty-phase instructions were so inadequate that the jury could not sufficiently appreciate the meaning of mitigating factors, and, more tellingly, did not fully understand that the jury alone would be responsible for the imposition of the death sentence.
Errors of this kind flow from the systemic deficiencies in our capital-murder law as enacted, interpreted, and applied. This case exemplifies the irreconcilable conflicts inherent in the bifurcated-trial scheme, which professes to rely on the process of death qualification despite its prepotency toward guilt; and which insists that the same jury determine, with consistency, objectivity, and impartiality, both criminality and punishment. This prosecution exposes the incoherence of capital-murder prosecutions, which incomprehensibly can require that a jury be death qualified in order to impose the death sentence; and, concomitantly, allow a death sentence to be imposed by a jury that is not death-qualified. The Court’s decision countenances a proceeding in which the guilt conviction effectively predetermines the ultimate punishment. Further, as this case shows, although the statute mandates that a defendant sentenced to death have the right of appeal, the amorphous nature of the appellate standard of review can sanction a death sentence based on a trial riddled with errors from beginning to end. The case points up how hollow and illusory the constitutional standards of due process and the strictures of fundamental fairness can become in capital-murder prosecutions.
To those misgivings must be added a deeper skepticism about the worth and wisdom of capital punishment. In this time of crisis — economic shrinkage, rising crime rate, inadequate prosecutorial and defense resources, and an understaffed judiciary, we should rethink the value of using the death penalty in our criminal-justice arsenal.
*258The staggering financial cost of maintaining a capital-murder regime negates any practical benefit of our death-penalty statute. The cost of implementing the death penalty is potentially crippling to a state’s criminal justice system. See, e.g., Tabak and Lane, “The Execution of Injustice,” 23 Loyola of L.A.L.Rev. 59, 133-38 (1989) (hereafter “Tabak and Lane”); Spangenberg and Walsh, “Capital Punishment or Life Imprisonment? Some Cost Considerations,” 23 Loyola of L.A.L.Rev. 45 (1989); Garey, “Comment: The Cost of Taking A Life: Dollars and Sense of the Death Penalty,” 18 U.C. Davis L.Rev. 1221 (1985) (hereafter “Garey”); Nakell, “The Cost of the Death Penalty,” 14 Crim.L.Bull. 69 (1978). Capital cases demand that courts honor due process guarantees punctiliously. Gardner v. Florida, 430 U.S. 349, 97 S.Ct. 1197, 51 L.Ed.2d 393 (1977); State v. Williams I, supra, 93 N.J. at 61, 459 A.2d 641. Because capital cases are necessarily more complex than non-capital cases, they generate more pre-trial motions, more extended and intense voir dire, longer and more frequently repeated trials, greater collateral support from court personnel, and increased use of investigators and experts. The singular problems raised in a capital context, as well as the discouraging effect of the death sentence on plea arrangements, also add to the burden of the death penalty on the criminal justice system. The greater hours necessary for the presentation of capital trials translate into higher court, prosecution, and defense costs. The exhaustive research and investigation required for the penalty phase add further costs. Defendants who are tried for capital murder but are acquitted or receive lesser sentences also invoke the more labor-intensive, expensive capital-prosecution machinery. As a result, less money and time are available for crime prevention and the efficient functioning of the other aspects of the criminal justice system, including the prosecution of other criminal causes and prison maintenance. See LeDuk, “Tough N.J. Laws Mean Jammed Jails,” Philadelphia Inquirer, Dec. 20, 1990 at Al, col. 3 (noting effects of severe prison overflow *259and New Jersey’s increasing allocations to corrections in a shrinking budget).
Information relating to the cost of the death penalty in New Jersey has started to become available. When it enacted the capital-murder statute, the State estimated an expense of $16 million per year to administer the death penalty. In 1983, the New Jersey Public Defender was budgeting more than $100,000 per capital case for trial costs alone for its fifty-two pending capital cases. Garey, supra, at 1261. The Public Defender has acknowledged that it exhausted 6,000 man-hours and spent approximately $300,000 to defend its first capital case, State v. Ramseur. Remarks of Dale Jones, Assistant Public Defender, “Life After Murder” (FOX WNYW Television Broadcast, January 2, 1991) (hereafter “Jones”). In 1987, the Deputy Public Defender estimated an expenditure of $1 million per defendant for defense costs. Glading, “Public Defenders Will Fight Death Penalty,” Trenton Times, March 8, 1987, A6 at col. 1 (hereafter “Glading”). That is partly a result of internal policy, whereby to ensure adequate representation the Public Defender assigns two attorneys to each capital defendant. It has been estimated that it costs $42,000 per capital defendant just for expert witnesses. Tabak and Lane, supra, at 137. Currently, the defense of the prosecution of a capital case at the trial level costs the State $500,000, and the defense of the two-dozen cases most recently prosecuted by the State costs approximately $12 million per year. Jones, supra. The Administrative Office of the Courts has estimated that the death penalty will cost over $7 million per year in court expenses. Other available information notes the cost of the maintenance of death row and the increased supervision required. See Glading, supra (reporting that it costs $2,000 more annually to house a prisoner on death row than to house him in the general prison population). Finally, it has been estimated that sentencing one person to death in New Jersey would cost about $7.3 million. See Tabak and Lane, supra, at 136 (referring to information supplied by New Jersey Office of the Public Advocate (Jan. 19, 1989)).
*260That does not reflect all of the real expenses. For example, the Supreme Court itself has consumed untold hours and expended enormous effort in deciding capital cases on direct appeal. Those appeals are onerous. Over the last five years the Court has hired two extra law clerks to work virtually full-time on capital appeals; other law clerks also invest time on capital cases. Over twenty-seven cases have reached the Court and undergone final review. Many matters involving capital causes have been before the Court on an interlocutory basis. The Court’s opinions in capital cases alone cover more than 2,000 pages in the printed official reports of the Court’s decisions, almost twenty-five percent of all the Court’s decisional work-product for that period. Research entailed in developing acceptable proportionality review will likely cost more than $300,000.
Studies conducted elsewhere also suggest that the price of death is exorbitant. See, e.g., Moran & Ellis, “Death Penalty: A Luxury Item,” Newsday, June 14, 1989 at 60, col. 1 (death penalty would cost New York taxpayers $1 million per capital trial); New York State Defenders Association, Inc., Capital Losses: The Price of the Death Penalty in New York State, A Report From the Public Defense Backup Center to the Senate Finance Committee, the Assembly Ways and Means Committee, and the Division of the Budget at 12 (1982) (estimating a cost of $1.8 million per capital defendant from trial through initial United States Supreme Court review, three times the cost of life imprisonment); Gradess, “A Lesson in Execution Economics,” Sunday Times Union, April 16, 1989, at col. 1 (noting that the New York State Bar Association has pledged to oppose the reimposition of capital punishment in New York because of its cost); Von Drehle, “Bottom Line: Life In Prison One-Sixth As Expensive,” Miami Herald, July 10, 1988, 12A at col. 1 (concluding that it costs $3.2 million to execute someone in Florida, six times the cost of life imprisonment); Magagnini, “Closing Death Row Would Save State $90 Million a Year,” Sacramento Bee, March 28, 1988, at 1, col. 1 (death penalty *261costs California an additional $90 million per year); Indiana Legislature, A Fiscal Impact Statement Re Senate Bill 531 (1989) (reporting that capital cases cost $204,000 more per case than non-capital cases, and that replacing the death penalty with life-without-parole would save Indiana $5 million annually); Gradess, “The Death Penalty Doesn’t Pay,” Houston Chronicle, March 13, 1988, § 6 at 1, col. 1 (Texas spends $2 million per capital case); Gradess, “Execution Does Not Pay,” Washington Post, February 28, 1988 at 37 C5, col. 3 (hereafter “Gradess, ‘Execution’ ”) (noting that high costs of death penalty will increase as death row numbers grow); see also Gottlieb, “The Death Penalty in the Legislature: Some Thoughts About Money, Myth, and Morality,” 37 U.Kan.L.Rev. 443 (1989) (hereafter “Gottlieb”) (describing the Kansas Senate’s defeat in 1987 and 1989 of capital punishment bills and the role of cost arguments in the debates). Indeed, the overwhelming cost of the death penalty has nearly bankrupted one county. See Ricks, “Seminole Borrows to Pay for Alday Case,” Atlanta J. and Const., Aug. 8, 1988, at B6, col. 4 (describing how the prosecutions of three capital cases drove Seminole County, Georgia heavily into debt). At least two jurisdictions have complained about capital appeal workload clogging their dockets. Gradess, “Execution,” supra (noting that more than thirty percent of the workload of the Eleventh Circuit consists of capital cases); Cox, “Inaction In Action in California,” Nat’l L.J., July 11, 1988, at 1, col. 1 (hereafter “Cox”) (because it hears so many capital appeals, the California Supreme Court has had to markedly reduce the number of non-capital cases it can hear). In fact, observers of the California Supreme Court have noted that the scores of backlogged death penalty cases have stalled the development of California’s jurisprudence in other areas. Cox, supra. Capital cases have been “a real albatross around [the Justices’] necks,” severely limiting the number of cases the Court can hear and opinions it can publish. Ibid, (quoting Prof. Stephen R. Barnett, Boalt Hall School of Law).
*262The inordinate cost of death may yet divert resources from other areas of the criminal justice system. Due process demands that if capital punishment is to exist, safeguards requiring the substantial expenditure of public funds be in place. Because we must spend substantial amounts on relatively few cases, limited funds will be deflected from the general pool of cases in the criminal-justice system. One commentator, remarking on the Kansas legislature’s rejection of the death penalty, largely because of the tremendous drain of resources that the state would suffer, has noted: “The enormous effort and complex jurisprudence in [capital] cases have produced genuine distortions in the criminal justice systems of states that actively pursue capital punishment.” Gottlieb, supra, at 448.
There may also be an intellectual drain on our criminal-justice system attributable to the enormous commitment of resources and efforts to capital-murder prosecutions. The Court has, I think, failed to anticipate or acknowledge the strength of the gravitational pull of capital-murder jurisprudence on the criminal law itself, as well as on the general administration of criminal justice. One may legitimately wonder whether the capital-murder jurisprudence that is emerging from this horrendous effort itself is not exerting an influence on our criminal law in general. For example, with respect to capital cases, despite initial strong statements concerning the protections entailed in juror qualification, see, e.g., State v. Biegenwald, 106 N.J. 13, 524 A.2d 130; State v. Williams II, supra, 113 N.J. 393, 550 A.2d 1172, we have shown ambivalence if not outright laxity in succeeding cases. See, e.g., State v. Moore, 122 N.J. 420, 585 A.2d 864 (1991); State v. Hunt, supra, 115 N.J. 330, 558 A.2d 1259; State v. Koedatich, supra, 112 N.J. 225, 548 A.2d 939. This Court has now adopted the same standard with respect to effective assistance of counsel as applies in ordinary cases. See, e.g., State v. Davis, 116 N.J. 341, 561 A.2d 1082 (1989). It does not appear, however, that its application in capital cases is in any way extraordinary. Compare State v. Oglesby, 122 N.J. 522, 585 A.2d 916 (1991); State v. Davis, *263supra, with State v. Savage, 120 N.J. 594, 577 A.2d 455 (1990). The Court purports to follow the same principles governing expert testimony in both capital and noncapital cases. Ironically, it has tended to be less scrupulous in authorizing the State’s use of expert testimony in capital cases. Compare State v. Moore, supra, 122 N.J. 420, 585 A.2d 864; State v. Zola, 112 N.J. 384, 548 A.2d 1022 (1988) with State v. Pitts, supra, 116 N.J. 580, 562 A.2d 1320. But cf. State v. Harvey, 121 N.J. 407, 581 A.2d 483 (1990) (expert testimony without proper foundation should not be admitted); State v. Johnson, 120 N.J. 263, 576 A.2d 834 (1990) (same). The Court started out by expressing a stringent standard governing prosecutorial misconduct. See State v. Ramseur, supra, 106 N.J. 123, 524 A.2d 188; State v. Biegenwald, supra, 106 N.J. 13, 524 A.2d 130; see Williams II, supra, 113 N.J. 393, 550 A.2d 1172; yet it has rarely applied the standard. See, e.g., State v. Moore, supra, 122 N.J. 420, 585 A.2d 864; State v. Koedatich, supra, 112 N.J. 225, 548 A.2d 939. Our standards governing the admissibility of controversial evidence, such as the use of prior convictions and other-crimes evidence, if anything have become less protective of capital defendants. See, e.g., State v. Moore, supra, 122 N.J. 420, 585 A.2d 864; State v. Clausell, 121 N.J. 298, 580 A.2d 221 (1990); State v. Hightower, supra, 120 N.J. 378, 577 A.2d 99; State v. Pennington, 119 N.J. 547, 575 A.2d 816; State v. Long, supra, 119 N.J. 439, 575 A.2d 435. The Court’s application of Miranda standards protects a capital defendant only in the clearest cases, see, e.g., State v. Harvey, supra, 121 N.J. 407, 581 A.2d 483; State v. Johnson, supra, 120 N.J. 263, 576 A.2d 834, but fails to do so in less clear cases that arguably would have been different had a capital offense not been involved. See, e.g., State v. Bey I, supra, 112 N.J. 45, 548 A.2d 846. One may believe that the corroboration rule applicable to confessions was diluted in a capital prosecution. See, e.g., State v. Di Frisco, supra, 118 N.J. 253, 571 A.2d 914. The Court has seemingly been less protective in invoking the doctrine of lesser-included offenses in capital causes than it would *264be in ordinary cases. Compare, e.g. State v. Rose, supra, 112 N.J. 454, 548 A.2d 1058 with e.g., State v. Grunow, 102 N.J. 133, 506 A.2d 708 (1986). The standards governing plea-bargaining have been applied with aspects of strictness in capital causes to the disadvantage of capital defendants. See State v. Kiett, supra, 121 N.J. 483, 582 A.2d 630; State v. Davis, 116 N.J. 341, 561 A.2d 1082. Further, capital-murder sentencing has resulted in a polyglot of principles and rules, many of which would be alien in an ordinary criminal sentencing proceeding, and which can operate with peculiar unfairness against a capital defendant. Thus, a defendant is given the freedom to ask for death, State v. Hightower, supra, 120 N.J. 378, 577 A.2d 99, but the defendant’s own family may not implore for mercy. State v. Moore, supra, 122 N.J. 420, 585 A.2d 864. The Court adopted a conventional standard of appellate review believing that it could secure greater protections by a more scrupulous application, see State v. Bey I, supra, 112 N.J. 45, 548 A.2d 846; State v. Rose, supra, 112 N.J. 454, 548 A.2d 1058, but, if anything, that standard is ineffective except with respect to the most egregious kind of error. See, e.g., State v. Pitts, supra, 116 N.J. 580, 562 A.2d 1320.
The Court believes that the legal principles governing capital causes are indistinguishable from those governing ordinary cases. State v. Davis, supra, 116 N.J. at 341, 561 A.2d 1082. Hence, there is no reason to suppose that its capital cases will not constitute precedential authority in ordinary criminal cases. See, e.g., State v. O’Donnell, 117 N.J. 210, 217-18, 564 A.2d 1202 (1989) (citing Ramseur's interpretation of N.J.S.A. 2C:11-3c(4)(c) in its interpretation of “cruel” in N.J.S.A. 2C:44-1a(1) as an aggravating factor); State v. Stevens, 115 N.J. 289, 293, 302, 558 A.2d 833 (1989) (citing Ramseur as governing admissibility of evidence under Evidence Rule 55); State v. Murphy, 110 N.J. 20, 34-36, 538 A.2d 1235 (1988) (upholding assignment judge’s decision not to dismiss indictment where prosecutor failed to notify assignment judge of possible grand juror bias, citing Ramseur); State v. Breakiron, 108 N.J. 591, 609, 614, *265532 A.2d 199 (1987) (ruling that defendant’s mental disease or defect may negate mens rea, but does not, as a matter of law, reduce murder to an unspecified degree of manslaughter, citing Ramseur)-, State v. Murray, 240 N.J.Super. 378, 392-93, 573 A.2d 488 (App.Div.1990) (finding voir dire adequate under standards of Ramseur and Biegenwald)-, State v. Slattery, 239 N.J.Super. 534, 571 A.2d 1314 (App.Div.1990) (trial judge’s comment tending to dilute jury’s sense of responsibility not reversible, citing Rose and Ramseur)-, State v. Jenkins, 234 N.J.Super. 311, 316-17, 560 A.2d 1240 (App.Div.1989) (applying standards of admissibility of evidence under Evidence Rule 55, citing Ramseur)-, State v. Gary, 229 N.J.Super. 102, 113-14, 550 A.2d 1259 (App.Div.1988) (effect of pretrial publicity governed by Biegenwald and Koedatich)-, State v. Fiorilla, 226 N.J.Super. 81, 88, 543 A.2d 958 (App.Div.1988) (citing Biegenwald to support determination of defendant’s right to waive jury trial); State v. Davidson, 225 N.J.Super. 1, 9-10, 541 A.2d 700 (App.Div.1988) (citing Ramseur as authority for requiring more extensive voir dire)-, State v. Watson, 224 N.J.Super. 354, 362, 540 A.2d 875 (App.Div.) (citing Ramseur as authority in determining effect of prosecutorial misconduct), certif. denied, 111 N.J. 620, 546 A.2d 537, cert. denied, 488 U.S. 983, 109 S.Ct. 535, 102 L.Ed.2d 566 (1988); State v. Juinta, 224 N.J.Super. 711, 722-24, 541 A.2d 284 (App.Div.) (citing Ramseur as support for effect of diminished capacity on requisite mens rea of “recklessness”), certif. denied, 113 N.J. 339, 550 A.2d 453 (1988); State v. Washington, 223 N.J.Super. 367, 373-74, 538 A.2d 1256 (App.Div.) (determination of inclusion of offenses of aggravated manslaughter and manslaughter in jury charge, citing Ramseur), certif. denied, 111 N.J. 612, 546 A.2d 531 (1988); State v. Garafola, 226 N.J.Super. 657, 665, 545 A.2d 257 (Law Div.1988) (citing Biegenwald standard with respect to effect of pretrial publicity).
There is every reason to believe, therefore, that the judicial vacillation reflected in the application of those principles in capital prosecutions and the subtle weakening of those protec*266tions even for capital defendants will be mirrored in the prosecution of ordinary criminal causes. That process has possibly already started. See, e.g., State v. Murray, supra, 240 N.J.Super. at 392-93, 573 A.2d 488 (voir dire deemed sufficient); State v. Slattery, supra, 239 N.J.Super. 534, 571 A.2d 1314 (improper instruction on jury’s responsibility not reversible); State v. Gary, supra, 229 N.J.Super. at 113-14, 550 A.2d 1259 (pretrial publicity not prejudicial); State v. Watson, supra, 224 N.J.Super. at 367, 540 A.2d 875 (prosecutorial misconduct not prejudicial); State v. Garafola, supra, 226 N.J.Super. at 665, 545 A.2d 257 (pretrial publicity not prejudicial).
Despite the concerns that relate to the serious defects in the capital-murder statute as enacted, interpreted, and applied, despite the convincing empirical evidence and strong intuition that tell us that the death penalty cannot be rationally invoked or fairly imposed, despite compelling countervailing considerations of public policy, the Court validates this law. It resists the conclusion that there is something fundamentally awry with our capital-murder jurisprudence, it ignores the emergence of the disquieting truth that capital punishment cannot really be made to work in a civilized society. It chooses to live with a case, this case, in which a criminal defendant is arbitrarily sentenced to die. The Court’s decision, then, more so than any that has preceded it, brings home the unmanageable conflicts that inhere in the administration of capital-murder prosecutions and the deepening incoherence of our capital-murder jurisprudence.
I believe strongly that society is entitled to express through its institutions the outrage felt when the lives of innocent persons are taken callously, senselessly or cruelly. See State v. Ramseur, supra, 106 N.J. at 468, 524 A.2d 188 (Handler, J., dissenting). None of us can escape our common humanity; we are all violated and reduced by the taking of an innocent victim’s life. Yet we can draw no satisfaction in punishment that does not conform to law. Our constitutional values, we *267must hope, retain integrity and do not allow us, no matter how outraged, to exact unprincipled retribution. The Court’s decision today dashes that hope. No one, not even the most ardent advocate of the death penalty, should rejoice in the unprincipled taking of a life. Our constitution is for all of us: if it fails the most reprehensible, it fails the rest. The death sentence in this case sounds with deafening finality for the defendant, but its discordant reverberations resound for everyone. John Donne, in Devotions XVII, gives voice to this truth: “[A]ny man’s death diminishes me, because I am involved in mankind; and therefore never send to know for whom the bell tolls; it tolls for thee.”
For affirmance as to Guilt Phase — Chief Justice WILENTZ and Justices CLIFFORD, POLLOCK, O’HERN, GARIBALDI and STEIN — 6
For reversal as to Guilt Phase — Justice HANDLER — 1
For affirmance as to Penalty Phase — Chief Justice WILENTZ and Justices CLIFFORD, POLLOCK, GARIBALDI and STEIN — 5
For reversal as to Penalty Phase — Justices HANDLER and O’HERN — 2
Defense counsel stated:
On the questionnaire, appearing on the last page are three questions pertaining to a prospective juror’s views on capital punishment. In the past at least, even if the prospective juror answered in a fashion which would be consistent with being qualified to serve on this jury, your Honor has still explored to some extent their views in this area. My request is that at least with regard to further panelists, if the answers to the questions are such that they would be qualified by virtue of their answers, I ask that your Honor not inquire any further into their views on capital punishment inasmuch as I feel it only serves to highlight further the area that I had objected to in the very beginning that deals with death qualification.
The court’s closing admonition to the jury was:
Ladies and gentlemen, it has been said, and correctly, that many of the words that we use are derived from the Latin, and one of the words in that category is the word verdict. The word verdict is derived from the Latin veredictum, and it means a true declaration. So let your verdict declare the truth.
[Ante at 135, 586 A.2d at 154.]
The lower court’s determination that the State did not have a purpose or motive to withhold information is conjectural, viz:
The conclusion that the non-disclosure was not willful is supported by examination of the question of motive. Presence or absence of motive bears upon whether an intentional wrong would be committed. There would have existed the motive to conceal the documents in order to prevent their being utilized to affect the credibility of the witness. However, the Prosecutor would have been aware that he could demonstrate that a substantial portion of Kraushaar's story was told on September 7, 1984, twenty days prior to any discussion of immunity. He was aware of what the testimony of Billy Wayne McKinnon would be, together with the fact that it would not have been plausible for McKinnon to have been involved in the murder either as an accomplice (as alleged by the State) or *230as the shooter (as alleged at trial by co-defendant Thompson) except at the instigation of Marshall. He was aware that he could demonstrate false and inconsistent statements made by Marshall following the murder. He was aware of the compelling circumstantial evidence of guilt. Being aware of all these things, on balance any motive for nondisclosure and its accompanying risks — the documents also being in Altman's file — was substantially outweighed by the motive to disclose.