State v. Marshall

HANDLER, Justice,

dissenting.

Last year this Court affirmed the death sentence of Robert Marshall. State v. Marshall, 123 N.J. 1 , 586 A.2d 85 (1991) (Marshall I). Marshall, a Toms River insurance salesperson, had been convicted of capital murder for his hiring of a professional killer to kill his wife. Marshall’s motive, apparently, had been to collect proceeds on insurance policies that he had taken out on his wife’s life. Pursuant to the New Jersey capital-murder statute, every defendant in New Jersey who is sentenced to death may call on the Supreme Court to “determine whether the sentence is disproportionate to the penalty imposed in similar cases, considering both the crime and the defendant.” N.J.S.A. 2C:11-3e.

Marshall has requested proportionality review. Because his is the first death sentence affirmed under the 1982 capital-murder statute, he is the first defendant to do so. The Court has now decided that Marshall’s death sentence is proportionate *230 to those of other comparable capital defendants and therefore Marshall may be executed. I disagree.

In my opinion, a defendant may not under the state constitution be sentenced to death unless that death sentence is proportionate to sentences imposed on other defendants who have committed similar crimes. That constitutional stricture places on courts the heavy obligation to adopt adequate standards to ensure that death sentencing is evenhanded and consistent. Measured against that exacting constitutional requirement, I conclude that the Court’s methodology for determining whether a death sentence is proportionate is fundamentally flawed. I also believe, however, that even under the proportionality standards adopted by the Court, Marshall’s death sentence is disproportionate.

When all is said and done, the Court’s proportionality review reveals not only the inefficacy of our death-penalty statute but the inefficacy of proportionality review itself. This case confirms that proportionality, although constitutionally required, is practically unattainable. This case lends added support to the position that our capital-murder jurisprudence fails to meet our constitutional standards that prohibit cruel and unusual punishment and demand exacting procedural protections as a matter of due process and fundamental fairness. This case further confirms that capital punishment cannot be sensibly administered, soundly applied, or rationally managed. Our capital-murder law not only cannot be integrated into our system of criminal justice, it serves to weaken and distort it. The conclusion is irresistible: capital punishment is unconstitutional, unwise, and untenable. It should be abandoned.

I

Proportionality review of capital sentencing must be appreciated as a constitutionally-inspired and constitutionally-founded procedural protection. That appreciation, more than any other insight, determines the standards for defining and applying *231 proportionality review and for validating capital sentences. The understanding of the constitutional significance of proportionality review encompasses, initially, federal constitutional doctrine, but finally turns on state constitutional principles.

—A—

In Furman v. Georgia, 408 U.S. 238 , 92 S.Ct. 2726 , 33 L.Ed.2d 346 (1972), the United States Supreme Court invalidated Georgia’s death penalty statute as violating the Eighth Amendment. A central, constant theme throughout the nine opinions of Furman is that the statute lacked sufficient safeguards to prevent the arbitrary, inconsistent, and discriminatory imposition of the death penalty. Specifically, the statute did not provide a “meaningful basis for distinguishing the few cases in which it [was] imposed from the many cases in which it [was] not.” Id. at 313 , 92 S.Ct. at 2764 , 33 L.Ed.2d at 392 (White, J., concurring); see id. at 248-50 , 92 S.Ct. at 2731-32 , 33 L.Ed.2d at 354-55 (Douglas, J., concurring); id. at 293-95 , 92 S.Ct. at 2754-55 , 33 L.Ed.2d at 380-81 (Brennan, J., concurring); id. at 309-10 , 92 S.Ct. at 2762-63 , 33 L.Ed.2d at 390 (Stewart, J., concurring); id. at 364-68 , 92 S.Ct. at 2790-92 , 33 L.Ed.2d at 421-23 (Marshall, J., concurring). This Court has understood that

Furman suggested that to pass constitutional muster, a capital punishment statute must achieve two objectives: limit imposition of the penalty to what is assumed to be the small group for which it is appropriate, and ensure that the limited class selected for the penalty is chosen with rationality and consistency. Both requirements are aimed primarily at eliminating the arbitrary nature of capital proceedings in the past and their high risk of discrimination.

[State v. Ramseur, 106 N.J. 123, 183 , 524 A.2d 188 (1987) (citations omitted).]

In response to Furman , states drafted statutes that attempted to restrict the discretion available to capital juries either by guiding jury discretion through the enumeration of certain aggravating and mitigating factors or by eliminating jury discretion through mandatory death sentencing for certain crimes. *232 In Gregg v. Georgia, 428 U.S. 153 , 96 S.Ct. 2909 , 49 L.Ed.2d 859 (1976), and accompanying cases, the Court upheld guided discretion statutes. See Jurek v. Texas, 428 U.S. 262 , 96 S.Ct. 2950 , 49 L.Ed.2d 929 (1976); Proffitt v. Florida, 428 U.S. 242 , 96 S.Ct. 2960 , 49 L.Ed.2d 913 (1976).

Subsequent to Gregg , the Supreme Court made plain that capital-sentencing schemes are permitted only if they limit the range of crimes for which the death penalty may be imposed. See, e.g., Maynard v. Cartwright, 486 U.S. 356, 363-64 , 108 S.Ct. 1853,1858-59 , 100 L.Ed.2d 372, 381-82 (1988); Godfrey v. Georgia, 446 U.S. 420 , 100 S.Ct. 1759 , 64 L.Ed.2d 398 (1980). However, in Woodson v. North Carolina, 428 U.S. 280 , 96 S.Ct. 2978 , 49 L.Ed.2d 944 (1976), the Supreme Court struck down schemes that made the death penalty mandatory for particular crimes. The Court stated that the Eighth Amendment “requires consideration of the character and record of the individual offender as a constitutionally indispensable part of the process of inflicting the penalty of death.” Id. at 304-05 , 96 S.Ct. at 2991, 49 L.Ed.2d at 961. Thus, although mandatory death sentencing may promote consistency in the imposition of punishments, it is forbidden because it would allow the State to take a defendant’s life as if that defendant were a mere statistic and not a human being with unique, personal qualities. Id. at 301-03, 96 S.Ct. at 2989-91, 49 L.Ed.2d at 959-60; see also Lockett v. Ohio, 438 U.S. 586 , 98 S.Ct. 2954 , 57 L.Ed.2d 973 (1978) (ruling that sentencing juries may consider any characteristic of the defendant as evidence mitigating against imposition of the death penalty).

Thus two major strands run through United States Supreme Court jurisprudence. A death-penalty statute is constitutional only if (1) it limits or guides juror discretion and (2) it permits completely individualized consideration of a defendant’s character and culpability. Both strands share a common purpose of “identifying a group of defendants most deserving of death.” Scott E. Sundby, The Lockett Paradox: Reconciling Guided Discretion and Unguided Mitigation in Capital Sentencing, *233 38 UCLA L.Rev. 1147, 1176 (1991). Despite their broad common purpose, however, the two principles work in tension: Furman and Gregg strip juries of authority to deem certain crimes worthy of death, and thereby reduce the likelihood that individuals convicted of similar crimes will receive different sentences, while Woodson and Lockett endow juries with the power to act on any and all mitigating factors that they consider to be relevant to punishment, and thereby increase the likelihood that individuals convicted of similar crimes will receive different sentences. The inevitable result of the tension between those principles is some degree of arbitrariness. Disparate sentencing for defendants who commit comparable crimes will continue.

Proportionality review is the obvious antidote for the enduring arbitrariness of disparate sentencing. Nevertheless, the United States Supreme Court has held that comparative proportionality review is not required constitutionally of a capital-sentencing scheme if that scheme otherwise provides some rational mechanism to protect against arbitrariness. Pulley v. Harris, 465 U.S. 37, 45-46 , 104 S.Ct. 871, 876-77 , 79 L.Ed.2d 29, 37 (1984). In practice, following Pulley , no federal court ever has required a state to include comparative proportionality review in its system for imposing capital punishment.

—B—

This Court has taken a different approach with respect to the importance of proportionality in death sentencing. In State v. Ramseur, supra, 106 N.J. 123 , 524 A.2d 188 , the Court expressed preliminary views on the framework and purposes of proportionality review. The Court identified consistency as its paramount purpose. Id. at 331, 524 A.2d 188 . According to the Court, proportionality review “is an important procedural mechanism to safeguard against the arbitrary and capricious imposition of the death penalty.” Id. at 330, 524 A.2d 188 . The Court identified two reasons for requiring proportionality review. First, because the death penalty is different from all *234 other punishments in its finality, the unfairness attendant to inconsistent sentencing is especially intolerable. Id. at 326, 524 A.2d 188 . Second, proportionality review “is a means through which to monitor the imposition of death sentences and thereby to prevent any impermissible discrimination in imposing the death penalty.” Id. at 327, 524 A.2d 188 .

The legislative history to New Jersey’s death-penalty statute evinces similar concerns: proportionality review ensures that death sentences “are being meted out in a fair, even-handed way throughout the State.” Hearings on S.112 Before the N.J. Senate Judiciary Comm., 200th Leg., 2d Sess. 20-21 (1982) (statement of Director of the Division of Criminal Justice Edwin H. Stier), cited in Joseph H. Rodriguez at al., Proportionality Review in New Jersey: An Indispensable Safeguard in the Capital Sentencing Process, 15 Rutgers L.J. 399 , 421 (1984) (Rodriguez, Proportionality Review).

The policy that demands proportionality in capital sentencing also was acknowledged by the Special Master in this case. Evenhanded sentencing, he noted, assures that “death sentences are only imposed in categories of cases on which there is a clear societal consensus as to their deathworthiness.” Final Report 25 n. 23. Proportionality review serves “(a) to insure that the cases in which death sentences are carried out can be meaningfully distinguished from those cases in which lesser penalties are normally imposed and (b) to limit death sentencing to categories of death eligible cases that are most aggravated and in which death sentences are the usual, routine practice.” Ibid.

I believe that exacting proportionality review as a procedural protection is required by the state constitution, grounded in our own principles concerning cruel and unusual punishments, due process, and equal protection. See State v. Ramseur, supra, 106 N.J. at 370 , 524 A.2d 188 (Handler, J., dissenting) (observing that death penalty implicates several constitutional provisions, particularly paragraphs 1 (right of enjoying life), 5 (no *235 civil rights denied because of discrimination); and 12 (cruel and unusual punishments shall not be inflicted) of Article I); see also Rodriguez, Proportionality Review, supra, 15 Rutgers L.J. at 421 (arguing that authority to conduct proportionality review is found in cruel and unusual punishment clause; the general appellate power to examine criminal sentence for manifest excessiveness; and the Supreme Court’s power to insure that justice is truly and equally done). Those constitutional doctrines are to be interpreted in light of our strong and distinctive traditions demanding evenhandedness and consistency in all criminal sentencing, traditions that become absolute imperatives when criminal sentencing encompasses capital punishment. Any failure to assure evenhandedness and consistency in the sentencing of all capital defendants and any failure to assure comparative proportionality of individual death sentences violates those constitutional principles, and, indeed, fails to honor the constitutional values placed on individual dignity and human life.

We should not have the slightest hesitancy to look to our own constitution and jurisprudence to determine the rights that must be protected when the State seeks to impose capital punishment. The United States Supreme Court itself has recognized that the death penalty is a matter of peculiar state concern. California v. Ramos, 463 U.S. 992 , 103 S.Ct. 3446 , 77 L.Ed.2d 1171 (1983). States, without question, “are free to provide greater protections in their criminal justice system than the federal Constitution requires.” Id. at 1013 , 103 S.Ct. at 3460 , 77 L.Ed.2d at 1188-89 . The federal constitution sets only the barest minimum, leaving the heart of the matter to the individual states.

We already have acknowledged that proportionality review is an integral component of a constitutional capital-punishment scheme, providing a “means through which to monitor the imposition of death sentences.” State v. Ramseur, supra, 106 N.J. at 327 , 524 A.2d 188 . We similarly have acknowledged that “capital punishment is a matter of particular state interest *236 or local concern and does not require a uniform national policy.” Id. at 167 , 524 A.2d 188 . We have, in fact, analyzed our death-penalty statute under our state constitution many times. E.g., State v. Gerald, 113 N.J. 40, 75-76 , 549 A.2d 792 (1988). Thus, it is exceedingly perplexing and regrettable that the majority, in expounding and applying a protection as essential as proportionality review, shuns our own constitution and timidly fails to exercise the freedom of action granted by the United States Supreme Court.

Proportionality review embodies two complementary policies, policies that provide a basis for the independent application of our constitution. The first is New Jersey’s concern with evenhanded sentencing. The policy of evenhanded sentencing is derived from constitutional sources that prevent cruel and unusual punishments, unequal treatment under the law, and invidious discrimination in the application of laws, as well as the guarantee of due process and the complementary doctrine of fundamental fairness in the administration of criminal justice. The policy is expressed in legislation, judicial decisions, and administrative rules and practices.

Closely associated with the policy of evenhandedness is the recognition that appellate review plays a principal role in vindicating substantive and procedural rights in criminal causes and effectuating the goal of evenhandedness. Conducted by the highest court with statewide jurisdiction, proportionality review is an indispensable aspect of appellate review that “tests the capital murder-death penalty scheme for fairness and consistency.” State v. Ramseur, supra, 106 N.J. at 407 , 524 A.2d 188 (Handler, J., dissenting). The principles of evenhanded sentencing and appellate review express a legal philosophy and tradition that is distinctive in our state. They constitute strong justification to seek within our own constitution the standards for determining the validity of capital sentencing. See State v. Johnson, 127 N.J. 458, 473-74 , 606 A.2d 315 (1992) (declaring that matters of particular state interest and concern warrant reliance on state constitution) 606 A.2d at 322 ; State v. *237 Williams, 93 N.J. 39, 57 , 459 A.2d 641 (1983); State v. Hunt, 91 N.J. 338, 366 , 450 A.2d 952 (1982) (Handler, J., concurring).

The coalescence of those principles — substantive evenhandedness in sentencing and appellate review to secure evenhandedness — compels the conclusion that comparative proportionality review is constitutionally mandated. Because we are dealing with constitutional doctrine, we must examine those significant state principles to identify and understand the constitutional interests to be protected.

The judiciary has long committed itself to achieving greater uniformity in sentencing. Prior to the enactment of the Code of Criminal Justice (Code), when sentencing courts enjoyed unfettered discretion, this Court interposed the constraint of consistency onto the sentencing determination. Because the legislature had not corralled the virtually unrestrained sentencing discretion exercised by trial courts, our Court attempted to do so by identifying the legitimate, basic aims of criminal punishment, see State v. Ivan, 33 N.J. 197, 199-200 , 162 A.2d 851 (1960), and by cataloguing the relevant sentencing considerations and requiring a statement of reasons to protect effective appellate review of the sentences. See State v. Leggeadrini, 75 N.J. 150 , 380 A.2d 1112 (1977). Although the Court acknowledged the need for individualized sentencing, it also understood the importance of consistency and uniformity. E.g., State v. DeStasio, 49 N.J. 247, 254-55 , 229 A.2d 636 (1967) (recognizing that “[u]niformity of treatment is an ideal of law enforcement” and requiring as a matter of administrative policy that all gambling sentences be handled by a single judge in each county), cert. denied, 389 U.S. 830 , 88 S.Ct. 96 , 19 L.Ed.2d 89 (1967); see State v. Hodge, 95 N.J. 369, 379-80 , 471 A.2d 389 (1984) (Code decision explaining consistency as sentencing goal).

The legislature shared the judiciary’s concern with evenhanded treatment of criminal offenders. Prompted by reports of arbitrary and disparate sentencing, the legislature overhauled the Code’s sentencing provisions in 1978. See Robert F. Knowl *238 ton, Comments Upon the New Jersey Penal Code, 32 Rutgers L.Rev. 1, 15 (1979). One of the key sentencing purposes of the reformed Code is “[t]o safeguard offenders against excessive, disproportionate or arbitrary punishment.” N.J.S.A. 2C:1-2b(4). In State v. Roth, 95 N.J. 334, 345 , 471 A. 2d 370 (1984), the Court found that “[t]he central theme of the Code’s sentencing reforms is the replacement of the unfettered sentencing discretion of prior law with a structured discretion designed to foster less arbitrary and more equal sentences.” Taking its lead from this Court’s efforts, the legislature identified the penological purpose of punishment and prescribed presumptive terms of imprisonment related to the gravity of crimes, and listed the determinative aggravating and mitigating factors that can justify modifications to the presumptive terms. See id. at 351 , 471 A.2d 370 .

This Court repeatedly has reaffirmed the overriding principle of uniformity in interpreting the sentencing sections of the Code. Eg., State v. Pineda, 119 N.J. 621 , 575 A.2d 855 (1990) (stating that the polestar of the Code is uniformity in sentencing); see State v. Pillot, 115 N.J. 558, 577 , 560 A.2d 634 (1989); State v. Towey, 114 N.J. 69, 79 , 552 A.2d 994 (1989); State v. O’Connor, 105 N.J. 399, 405-06 , 522 A.2d 423 (1987); State v. Yarbough, 100 N.J. 627 , 498 A.2d 1239 (1985). Thus, in both interpreting and applying the Code and reviewing the sentencing decisions of lower courts, the Court has embraced the Code’s dedication to uniform, evenhanded sentencing.

The sentencing reforms did not eliminate the need for judicial oversight, and appellate review constitutes an important tool in achieving the desired uniform results. See Knowlton, supra, 32 Rutgers L.Rev. at 15. “The Court foresaw a central role to be played by appellate courts in ... enhancing uniformity in the application of statutory standards to similar facts under similar circumstances.” State v. Jarbath, 114 N.J. 394, 400-01 , 555 A.2d 559 (1989); State v. Dunbar, 108 N.J. 80, 97 , 527 A.2d 1346 (1987) (ruling that the trial court must explain basis of sentence to facilitate judicial review); State v. Roth, supra, 95 *239 N.J. at 361, 471 A.2d 370 (commenting that “central to the reform of sentencing procedures is provision for appellate review of sentences to provide a greater degree of uniformity”).

In addition to legislation and court decisions, other sources sustain the policy of evenhanded sentencing. In Pillot , we traced some of the measures taken by a variety of state bodies, noting that “[o]ur state’s concern over the evils of sentencing inconsistency and disparity has been longstanding.” 115 N.J. at 572 , 560 A.2d 634 . For example, various committees have studied sentencing uniformity and proposed measures to achieve it. Id. at 573 , 560 A.2d 634 (noting that counties undertook studies as early as 1933 and 1940); id. at 575 , 560 A.2d 634 (noting that this Court and the Administrative Office of the Courts both have established committees to address issues relating to consistency in sentencing). Judicial Rules also reflect the policy commitment to uniform sentencing. See, e.g., R. 3:25A (allowing consolidation of charges pending against one defendant in multiple jurisdictions). The resentencing panel constituted to reconcile pre-Code and post-Code sentences exemplifies one of the administrative steps taken to further uniformity in sentencing. See 104 N.J.L.J. 369 (1979).

Those sources — legislation, decisional law, Court Rules and directives, actual practices, and prevailing attitudes — manifest our state’s commitment to the premise that “there can be no justice without a predictable degree of uniformity in sentencing.” State v. Hodge, supra, 95 N.J. at 379, 471 A.2d 389 . In the capital-sentencing context, we require an elevated level of uniformity, a level that cannot possibly be achieved in the absence of strong, well-thought-out, and honestly applied proportionality review.

This state’s strong public policy of uniform sentencing and the concomitant role of appellate review to effectuate that policy clearly implicate significant state constitutional concerns. Those concerns involve the constitutional guarantees that prevent cruel and unusual punishments, unequal treatment under *240 the law, and invidious discrimination in the application of laws, and that assure due process and protect the right to life. Those concerns also involve the singular protection derived from the complementary doctrine of fundamental fairness in the administration of criminal justice.

Although the state constitution does not echo the federal constitution’s precise language on equal protection of laws, Article I, paragraph 1 provides “analogous or superior” protections. Peper v. Princeton Univ. Bd. of Trustees, 77 N.J. 55, 79 , 389 A.2d 465 (1978). This Court has always insisted on proper respect for the protection against unequal treatment of those who are similarly situated. The concern for equal protection of the laws intensifies when the law is applied to take a human life, and fully justifies recourse to the state constitution to require proportionality review.

Concededly, few defendants succeed on equal protection claims relating to sentencing in ordinary criminal cases. See, e.g., State v. Corbitt, 74 N.J. 379 , 378 A.2d 235 (1977), aff'd, 439 U.S. 212 , 99 S.Ct. 492 , 58 L.Ed.2d 466 (1978); State v. Fernandez, 209 N.J.Super. 37 , 506 A.2d 1245 (App.Div.1986). Nevertheless, as the relative severity of the punishment increases, sentencing disparities become more difficult to overlook. Of course, no punishment is more severe than the death penalty. As expressed by Justice Brennan:

Although we may tolerate ... irrationality in other sentencing contexts, the premise of Furman was that such arbitrary and capricious decisionmaking is simply invalid when applied to “a matter [as] grave as the determination of whether a human life should be taken or spared.”

[Pulley v. Harris, supra, 465 U.S. at 64 , 104 S.Ct. at 886 , 79 L.Ed.2d at 49 (Brennan, J., dissenting) (quoting Zant v. Stephens, 462 U.S. 862, 874 , 103 S.Ct 2733, 2741 , 77 L.Ed.2d 235, 248 (1983)). Accord State v. Bey, 112 N.J. 45, 105-19 , 548 A.2d 846 (1988) (Handler, J., concurring) (Bey I).]

The equal protection concern invariably arises when the legislature has created categories of offenses that result in dispa *241 rate treatment of offenders. That the death-penalty scheme is unfocused enough to generate disparate results, therefore, renders an equal protection analysis appropriate. “A capital sentencing system which results in differential treatment of similarly situated capital felons has effectively classified similar felons differently with respect to their rights to life.” Gary Goodpaster, Judicial Review of Death Sentences, 74 J. Crim. L. & Criminology 786 , 788, 802-03 (1983) (hereinafter Judicial Review).

The majority depreciates and excuses the occasional unequal sentencing treatment of similar defendants by attributing that to sentencer “mercy.” Ante at 153, 613 A.2d at 1081. But that rationalization explains aberrational life sentences, not aberrational death sentences. As one commentator explained:

[B]ecause individualized sentencing is constitutionally required, it may be argued that differential treatment of similar capital defendants is the inevitable result of the system. Given constitutional and human constraints, there is no way to perfect the system. It could further be argued that the fact that some defendants are underpenalized does not demonstrate that similar capital defendants who also deserve to die should not receive the death penalty or that the state has discriminated against the latter if they do. Sentencer mercy, as the argument runs, does not offend the Constitution.

These arguments are correct when applied to isolated cases, for the Constitution requires only that the decision to impose the death penalty be principled. There is a tipping point, however, at which these arguments lose their validity. Where capital sentences cannot be rationally distinguished from a significant number of cases where the result was a life sentence, more is present than the irremediable failure of an imperfect human system. When this occurs, the capital sentencing system has become constitutionally arbitrary. Further, where mercy becomes the norm for specific crimes and defendants, there is no justification for an isolated capital sentence. At this point, which is determinable only be comparative review, the state’s interests in executing the capitally sentenced criminal fail. As his fundamental right to life would then be taken without compelling justification, his execution would violate the equal protection clause of the Constitution.

[Goodpaster, Judicial Review, supra, 74 J.Crim.L. & Criminology at 802 (footnotes omitted).]

Comparative proportionality review is required to ensure that the capital-murder sentencing scheme is producing overall fair *242 and consistent results — a principle firmly lodged in our tradition of equal treatment under the law.

Moreover, even accepting that the state has a legitimate penological goal for adopting capital punishment, the means chosen to effect that interest must minimize the possibility that defendants will be wrongly deprived of their right to life. Ibid. “Since the basic issue is whether a state’s capital sentencing system differentially classifies similar defendants, a state must devise some means of ensuring proportional sentencing. Comparative review, since it is explicitly directed to this issue, is the least restrictive means of making this determination.” Ibid. Again, appellate review surfaces as a basic mechanism by which to achieve the desired goal of equal treatment of those who are similarly situated.

The unequal treatment of capital defendants also implicates Article I, paragraph 5 of the Constitution, which bars the denial of civil rights or discrimination against persons in the exercise of any right because of religion, race, or national origin. When inequality arises from improper considerations such as race, those concerns are most acute. The Special Master’s study preliminarily shows the race of the victim to be a factor in many decisions to impose the death penalty, Final Report, supra, at 100-06, and seems to confirm many long-held suspicions. See State v. Gerald, supra, 113 N.J. at 163-66 , 549 A.2d 792 (Handler, J., dissenting); State v. Ramseur, supra, 106 N.J. at 181, 327 , 524 A.2d 188 ; id. at 426, 524 A.2d 188 (Handler, J., dissenting). One purpose of proportionality review is to “prevent discrimination on an impermissible basis, including but not limited to, race and sex.” State v. Ramseur, supra, 106 N.J. at 330 , 524 A.2d 188 .

The doctrine of evenhandedness and the accordant requirement of comprehensive proportionality review implicates, perhaps most directly, the protections secured by the cruel and unusual punishment clause. N. J. Const. art. I, ¶12. The Court has concluded that our state constitution’s cruel and unusual *243 punishment clause “affords greater protections to capital defendants than does the eighth amendment of the federal constitution.” State v. Gerald, supra, 113 N.J. at 76 , 549 A.2d 792 . In order to comport with the principles underlying the cruel and unusual punishment clause, there must be a “meaningful basis for distinguishing the few cases in which [the death penalty] is imposed from the many cases in which it is not.” Furman v. Georgia, supra, 408 U.S. at 313 , 92 S.Ct. at 2764 , 33 L.Ed.2d at 392 (White, J., concurring). That purpose is accomplished by a proportionality review system “that assures similar results in similar cases.” State v. Ramseur, supra, 106 N.J. at 330 , 524 A.2d 188 .

Implicit in the prohibition of cruel and unusual punishment is that arbitrary sentencing is unacceptable because it fails both to comport with contemporary standards of decency and to serve a legitimate penological purpose. See id. at 169 , 524 A.2d 188 (citing State v. Des Marets, 92 N.J. 62, 82 , 455 A.2d 1074 (1983)). A disproportionate death sentence, by definition, is inconsistent with contemporary standards of decency and morality and does not match the community consensus of an appropriate punishment for the crime and the defendant. Societal retribution cannot justify the imposition of an aberrant death sentence because society has not generally demanded capital punishment to vindicate the offenses of similarly situated offenders.

Whether a particular jury verdict accurately reflects contemporary standards cannot be determined in isolation or by looking at the death sentence on its face; instead, it must be assessed comparatively. Proportionality review serves that function. “[W]hen juries generally do not impose the death sentence in a certain kind of case, the appellate review procedures assure that no defendant convicted under such circumstances will suffer a death sentence.” Gregg v. Georgia, supra, 428 U.S. at 206, 96 S.Ct. at 2940 , 49 L.Ed.2d at 893 . Moreover, “decisions to impose the death sentence must be consistent (in the sense of consistency with other decisions to *244 impose or not to impose death).” State v. Ramseur, supra, 106 N.J. at 185 , 524 A.2d 188 (interpreting Furman and Eighth Amendment).

Constitutional due process and the complementary doctrine of fundamental fairness also support exhaustive proportionality review. Due process is strengthened by fundamental fairness, particularly in the context of capital punishment. The principle of fundamental fairness “serves to protect citizens generally against unjust and arbitrary government action, and specifically against governmental procedures that tend to operate arbitrarily.” Id. at 377 , 524 A.2d 188 (Handler, J., dissenting). “Considerations of fundamental fairness are particularly heightened where the potential harm to the individual from arbitrary state action is greatest.” Id. at 378, 524 A.2d 188 . When the potential harm is the taking of a human life, the concerns about arbitrary state action are greatest and the safeguards against that arbitrariness must be at their peak. Thus, the doctrine of fundamental fairness mandates that this Court guard against the infringement of a defendant’s right to life by conducting an exhaustive, careful review of capital sentencing. State v. Matulewicz, 115 N.J. 191, 207 , 557 A.2d 1001 (1989) (Handler, J., concurring) (observing that proportionality review is mandated by precepts of fundamental fairness).

—C—

I have called attention to the infirmities of our death-penalty statute many times. E.g., State v. Jackson, 128 N.J. 136 , 607 A.2d 974, 975 (1992) (Handler, J., dissenting) (Jackson II) (noting that prosecutors have unguided discretion to decide who will be exposed to the death penalty); State v. Hunt, 115 N.J. 330, 391-95 , 558 A.2d 1259 (1989) (Handler, J., dissenting) (noting that a bifurcated trial with the guilt phase conducted before a death-qualified jury unfairly increases a defendant’s chance of being convicted); State v. Ramseur, supra, 106 N.J. at 382-404 , 524 A.2d 188 (Handler, J., dissenting) (noting that *245 the statutory definitions of murder and the aggravating factors do not sufficiently narrow the class of death-eligible murderers). In a statutory scheme fraught with such inadequate protections, the role of judicial review to determine sentencing proportionality attains heightened significance. Judicial review provides the best vantage point from which to make an overall examination of how the system is functioning. It is a last bastion against arbitrariness in imposing capital punishment. Only such a review can measure whether “ ‘we have designed procedures which are appropriate to the decision between life and death and ... [whether] we have followed those procedures.’ ” Pulley v. Harris, supra, 465 U.S. at 68-69 , 104 S.Ct. at 888-89 , 79 L.Ed.2d at 52 (Brennan, J., dissenting) (quoting Kaplan, The Problem of Capital Punishment, 1983 U.Ill. L.Rev. 555, 576). As I stated in State v. Ramseur:

The statute with its serious flaws of overbreadth, vagueness, and the blurring of decision-making, to which may be added unchecked prosecutorial discretion, is grossly defective if it cannot provide an ultimate fall-safe that could otherwise rectify individual injustice and spare the life of a defendant improvidently sentenced to death.

[ 106 N.J. at 406 , 524 A.2d 188 (Handler, J., dissenting).]

Simply stated: “If a given system produces arbitrary results, and a procedural mechanism exists that can correct those arbitrary results, the procedural mechanism is constitutionally required.” Goodpaster, Judicial Review, supra, 74 J.Crim.L. & Criminology at 799.

II

The Court has made some fundamental assumptions about the purpose and methodology of proportionality review. By understanding the Court’s philosophy and methodology, one can better assess the constitutional and statutory sufficiency of its standards for proportionality review of capital sentences.

The Court has determined that Marshall’s death sentence is proportionate. The proportionality review undertaken by the Court does not in principle or in application bring the Court to a *246 sound and correct result. No method of proportionality review can fully remedy the constitutional defects inherent in our death-penalty statute, but some methods are better than others. Had the majority adopted a method genuinely designed to address inconsistent sentencing, it would have corrected the injustice done in this case.

—A—

Until today, the Court never has stated precisely its standard for assessing proportionality. Today the Court answers the most fundamental question that arises in the design of a system of proportionality review: what does “disproportionate” mean? The Court’s answer is extremely disappointing. In several recent cases, I have observed that the Court has developed what I would now describe as a “split personality” on capital-punishment issues. At times, the Court has been laudably committed to fairness in capital punishment prosecutions, but at others it has applied the law with seeming severity and indifference. E.g., Marshall I, supra, 123 N.J. 1, 170 , 586 A.2d 85 (noting the severe and rigid applications of principles derived from prior capital-punishment decisions) (Handler, J., dissenting). The Court’s conflicted philosophy is exemplified by its holding today on the standard for assessing proportionality. The majority goes to great lengths to trumpet the virtues of evenhanded justice and to decry unfairness wherever it may lurk in capital sentencing. Yet, when the time arrives to deliver a holding that would impose real requirements of fairness and rationality on capital sentencing, the Court offers only a placebo.

As the Court points out, comparative proportionality review is supposed to be about consistency and evenhandedness; proportionality review “should seek to ensure that the death penalty is being administered in a rational, non-arbitrary, and evenhanded manner, fairly and with reasonable consistency.” Ante at 130-131, 613 A.2d at 1069. When conducting comparative *247 proportionality review, courts determine whether it would be unfair to single out a particular defendant for capital punishment when similarly situated defendants receive life sentences. In that sense, comparative proportionality review is quite unlike traditional proportionality review, in which courts examine the crime committed and determine whether executing any defendant for that crime would offend contemporary standards of decency.

Like the majority, I emphasize at the outset the ways in which the meaning of “disproportionate” differs depending on the form of proportionality review being used. Traditional proportionality review is what all courts undertake when evaluating sentences under the Eighth Amendment to the United States Constitution and what New Jersey courts undertake when evaluating sentences under Article 1, Paragraph 12 of the state constitution. See, e.g., Coker v. Georgia, 433 U.S. 584 , 97 S.Ct. 2861 , 53 L.Ed.2d 982 (1977) (declaring death sentence disproportionate for crime of rape); Enmund v. Florida, 458 U.S. 782 , 102 S.Ct. 3368 , 73 L.Ed.2d 1140 (1982) (declaring death sentence disproportionate for crime of felony murder when defendant neither killed nor intended to kill); Tison v. Arizona, 481 U.S. 137 , 107 S.Ct. 1676 , 95 L.Ed.2d 127 (1987) (declaring death penalty not disproportionate under federal constitution for crime of felony murder when defendant was recklessly indifferent to human life and was a major participant in the felony committed); State v. Gerald, supra, 113 N.J. 40 , 549 A.2d 792 (declaring death penalty disproportionate under New Jersey constitution when defendant acted without intent to kill and without knowledge that death would result). The death sentence is “disproportionate” in those cases in the sense that it never can be imposed for the commission of the offenses at issue, regardless of how consistently a particular jurisdiction tries to impose it.

The sort of review conducted in Coker, Enmund, and Gerald often helps to reduce arbitrariness in death sentencing. If a court deems a death sentence morally disproportionate, a *248 strong possibility exists that many juries would draw similar conclusions in similar cases. Thus, in Coker and Enmund , the Supreme Court amassed considerable evidence to show that most people convicted of the crimes in question did not receive the death penalty, and that the few who did had been terribly unlucky. Nevertheless, we must remember that the outcomes of those traditional proportionality cases did not depend on the inconsistencies the Court found in death-sentencing patterns. Those cases were not about evenhanded sentencing but were strictly moral decisions. See Enmund, 458 U.S. at 797 , 102 S.Ct. at 3376-77 , 73 L.Ed.2d at 1152 ; Coker, 433 U.S. at 597 , 97 S.Ct. at 2868 , 53 L.Ed.2d at 992 . The majority correctly perceives that the traditional form of analysis (which the majority describes as “offense-oriented” analysis) is an unsuitable substitute for comparative proportionality review. Ante at 129, 613 A. 2d at 1069. See also State v. Ramseur, supra, 106 N.J. at 324 n. 84, 524 A.2d 188 (noting that a claim that a death sentence is “manifestly excessive” and “inappropriate” is “distinct” from a claim that the death sentence is disproportionate when compared with the sentences of other defendants); accord Pulley v. Harris, supra, 465 U.S. at 43 , 104 S.Ct. at 875 , 79 L.Ed. 2d at 36 .

The majority succinctly states that a death sentence is comparatively disproportionate if “ ‘other defendants with similar characteristics generally receive sentences other than death for committing factually similar offenses.’ ” Ante at 129, 613 A.2d at 1069 (quoting Tichnell v. State, 297 Md. 432 , 468 A.2d 1 , 17 n. 18 (1983)). Had the Court genuinely embraced that holding, I still would have differed with it, for I believe that a death sentence is disproportionate unless defendants with similar characteristics generally receive death sentences for committing factually similar offenses. Nonetheless, had the Court been faithful to its announced standard, I would have agreed with it insofar as it identified the consistency of sentencing as the pivotal factor to be considered in the course of proportionality review.

*249 Given the rule the Court announces, its review should have involved the following steps and only the following steps: the Court first should have identified a universe of cases from which comparison cases could be drawn (as it did); it then should have formulated comparison groups of cases similar to the case under review (as it did); it then should have documented the actual sentencing outcomes in those groups to determine how frequently death sentences actually have been imposed in them (it did not); finally, it should have decided whether life sentences were imposed with sufficient frequency — and death with sufficient infrequency — to warrant reversal of the sentence under review (again, it did not). Had the Court followed that procedure, Marshall’s sentence would have been reversed. Unfortunately, the Court undertook only the first two steps, ignoring the final two.

Although the Court identifies a proper universe of cases from which to draw comparison cases, it dilutes its putative holding in two critical ways. First, it states that when a death sentence has been reversed, it still may be counted as a death sentence for the purpose of proportionality review, even if the defendant whose sentence has been reversed has been sentenced subsequently to life in prison. I cannot overemphasize that those sentences were reversed because we found errors at trial that infected the very deliberative process that led to the death verdicts in the first place. The reversals destroyed the reliability of those cases as valid determinations that the defendants deserved the death sentence. The Court’s artificial construct thus produces a flawed survey of jurors’ attitudes about the death penalty as well as a colossal exaggeration of the frequency with which that penalty is meted out in New Jersey. Second, by adopting a precedent-seeking method of proportionality review that is supposed to operate in some undefined “functional relationship” with the frequency analysis method of proportionality review, the Court blurs the focus of its endeavor. By grafting one method onto another entirely incompatible method, the Court creates a mutated and illogical definition of propor *250 tionality: a death sentence is disproportionate if other defendants with similar characteristics generally receive sentences other than death for committing factually similar offenses unless the Court concludes through its subjective intuitive examination of precedent that the sentence is fair. In my view, that standard for determining proportionality is indefensible. By embracing it the Court turns its back on its earlier pronouncements in Ramseur that comparative proportionality review must ensure factual “statewide uniformity” in capital sentencing, 106 N.J. at 329 , 524 A. 2d 188 , and that proportionality review is “distinct” from a review that asks if the punishment is morally justifiable. Id. at 325-26 n. 84, 524 A.2d 188 .

—B—

The majority correctly concludes that the universe of cases to be considered in assessing the proportionality of death sentences must consist of all death-eligible cases. Ante at 137, 613 A.2d at 1073. It soundly rejects both the Staté’s proposal that the universe be limited to death-sentenced cases and Justice Garibaldi’s proposal that the universe be limited to cases in which prosecutors sought the death sentence. Ante at 222, 1138, 613 A.2d at 1116, 1073.

If one looks only at death-sentenced cases, one cannot know if the death sentence for a particular case is aberrational. If one cannot know if a particular death sentence is aberrational, one cannot know if it is disproportionate. The issue was addressed graphically and persuasively in an opinion by Chief Justice Krivosha of the Nebraska Supreme Court:

The purpose of [the Nebraska proportionality review statute] was to ensure that persons were not being arbitrarily sentenced to death. To therefore suggest that we look only at those individuals who may have been discriminated against to determine whether or not they have been discriminated against is an exercise in futility. If one wants to determine whether individuals are being discriminated against in public transportation, one does not merely look at those who are required to sit in the back of the bus and conclude that since everyone in the back of the bus looks alike, there is no discrimination. One, of necessity, must look at who is riding in the front of the bus as well in order to determine *251 whether the persons in the back are being discriminated against. So, too, there is no way that we can determine whether those who are sentenced to death are being discriminated against if we do not examine those cases having the same or similar circumstances which, for whatever reason, did not result in the imposition of a death sentence.

[State v. Palmer, 224 Neb. 282 , 399 N.W.2d 706, 752 (1986) (Krivosha, C.J., concurring in part and dissenting in part).]

Further, because prosecutors are vested with initial authority to decide whether a defendant should die for his or her crime, confining the universe to cases in which prosecutors sought the death penalty would undermine proportionality review, for it would insulate arbitrary prosecutorial charging decisions from appellate review. As I noted in Jackson II, supra, 128 N.J. at 138 , 607 A.2d 974 , 975 our current decentralized system of capital prosecution is rife with potential for prosecutorial abuse of power. Prosecutors exercise their capital charging authority with almost no supervision, and the decisions they make are often highly idiosyncratic. All data collected to date reveal that prosecutors seek the death penalty for death-eligible defendants with little consistency or predictability. The report of the Special Master finds that there were 246 homicides committed in New Jersey since 1982 that could have been prosecuted as capital cases, but that prosecutors sought the death penalty in only 132 (fifty-six percent) of them. Final Report, supra, Table 13. The independent expert retained by the Attorney General reached very similar conclusions, estimating that 154 of 264 death-eligible cases (fifty-eight percent) resulted in capital prosecutions. Herbert I. Weisberg, Proportionality Review of Death Sentences in New Jersey (Nov. 26, 1991) at 19 (Weisberg Report). Even when the figures are broken down so that charging decisions are examined among sub-categories of crimes, wide charging disparities still exist. See, e.g., Final Report, supra, Table 13; Weisberg Report, supra, Table 2.

Thus, the need for judicial supervision of the capital-charging function is pressing, and the need to prevent arbitrary charging *252 from resulting in arbitrary capital sentencing even more so. Were we to look only at cases in which prosecutors actually sought the death penalty, we never would know if those prosecutorial decisions were aberrational. And if we could not know if a death sentence resulted from an aberrational charging decision, then we also could not know if it was disproportionate. I therefore conclude, with the majority, that the universe from which comparison cases are to be drawn must include all cases that could have resulted in the death penalty, including those cases in which prosecutors did not actually seek it. 1

*253 —c—

The majority’s calculation of death-sentencing frequencies is grossly distorted by its treatment of death-sentenced cases that have been reversed by this Court. The majority counts thirty-three cases as death-sentenced cases even though the death sentences handed down in them have been vacated and have not been reinstated. See Appendix A, infra at 289, 613 A.2d at 1149. In all but two cases, the Court does so without analysis, comment, or justification. See Ante at 169 n. 5, 613 A.2d at 1090 n. 5. In doing so it obscures both the extreme infrequency with which our state sentences death-eligible defendants to death and the glaring disproportionality of Marshall’s sentence. I would count reversed death-sentenced cases as life-sentenced cases when conducting proportionality review unless the case has resulted in a valid death sentence on re-trial.

Since the current death-penalty statute was enacted, there have been important interpretations and applications of the laws governing capital punishment that have changed the results in many individual cases. For example, we now require that jurors be told that defendants may be acquitted of a death sentence by a non-unanimous verdict and that trial courts not coerce juries into unanimous verdicts. State v. Clausell, 121 N.J. 298, 345-46 , 580 A.2d 221 (1990); State v. Ramseur, supra, 106 N.J. at 301 , 524 A.2d 188 . We have narrowed aggravating factor c(4)(c) so that a defendant may be sentenced *254 to death based on it only if he or she intentionally inflicted “severe physical or psychological pain or suffering to the victim” or if he or she killed merely “for the enjoyment” of killing. State v. Ramseur, supra, 106 N.J. at 211 , 524 A.2d 188 . A defendant who killed his or her victim but intended to inflict only serious-bodily-injury no longer may be sentenced to death. State v. Gerald, supra, 113 N.J. at 85 , 549 A.2d 792 . A defendant may not be sentenced to death if the prosecution relied on victim-impact evidence. Williams II, supra, 113 N.J. at 453-54, 550 A.2d 1172 . A defendant may not be sentenced to death if the jury was instructed incorrectly on how mitigating factors can be found and weighed. State v. Bey, 112 N.J. 123, 162-77 , 548 A.2d 887 (1988) (Bey II).

Our capital-punishment rulings have different implications for the counting of cases in which juries returned life sentences from those that have for the counting of cases in which juries returned death sentences. As the Special Master explains,

[b]ecause the effect of the new penalty-trial procedural rules has been to lower the likelihood of a death sentence, we can be confident that had the new procedural rules been applied in the earlier penalty trials that resulted in life sentences the life sentencing outcome would have been the same.

[Final Report, supra, at 58.]

The Special Master points out that “[t]he same cannot be said ... for the earlier death sentenced cases.” Ibid. That is because the new procedural rules have made the obtaining of death sentences more difficult. Every reversed case was reversed because the error “was clearly capable of affecting either the verdict or the sentence.” Bey I, supra, 112 N.J. at 94, 548 A.2d 846 . The unfathomable irony of the Court’s holding today is that although the Court found the reversed verdicts too untrustworthy for use in sentencing the individual defendants subject to them, it now finds them sufficiently trustworthy for use in sentencing other defendants who were not subject to them. Given the Court’s determination that the errors were sufficiently harmful to warrant reversals, the *255 Court’s current use of reversed cases as death-sentenced cases is utterly unprincipled.

The majority’s distorted calculation of death-sentencing frequency destroys the foundation of proportionality review. Proportionality review is about evenhanded sentencing in fact, not evenhanded sentencing in theory. The Court’s task is to look at how the system, as a whole, actually functions. To repeat, proportionality review exists to ensure that the procedures we have established do what they are supposed to do. State v. Ramseur, supra, 106 N.J. at 326 , 524 A. 2d 188 . It is supposed to “protect defendants from the arbitrary and capricious imposition of the death penalty.” Id. at 328, 524 A. 2d 188 (emphasis added). If the death penalty is not consistently imposed for a specific sort of crime, then the aberrant death sentences for that sort of crime should be vacated. Under our system as it actually works, the vacated deathworthiness judgments are not the final deathworthiness judgments in these cases. The final judgments — that is, the ones that govern the outcomes of cases — belong to the prosecutors who decide not to retry reversed cases capitally, and to the second juries that decide to sentence defendants to life in prison.

Reversed death sentences cannot be used reliably as “predictors” of future jury behavior. Death-sentencing rates have been decreasing in recent years. Final Report, supra, at 15-16. Juries do not return death sentences with the same frequency that they once did, and prosecutors, who base their charging decisions in part on their own predictions of what juries will do, do not charge cases capitally with the samé frequency that they once did. Id. at 20. When prosecutors choose not to seek the death penalty in reversed cases, or when juries choose not to impose it, those decisions are the best indicators of contemporary values as they interact with our state’s current laws on capital punishment. Thus, the prior sentencing decisions — made as they were under misguided legal notions that this Court has condemned in its own opinions— cannot serve as data in a legitimate proportionality review.

*256 The majority may be taking its cues from the Special Master, who believes that some reversed death-sentenced cases can be “rehabilitated” if, on retrial, they result in a new death sentence, or if they were reversed for certain guilt-phase errors rather than for penalty-trial errors. According to the Special Master, the reversed deathworthiness judgments in some cases may reflect community opinions concerning crimes of the sort that the juries in those cases thought had occurred. Final Report, supra, at 61-62. Still, the Special Master cautions that reversed death-sentenced cases should be “rehabilitated” only if the Court analyzes each one individually to ensure that the error causing reversal does not impugn its reliability. Id. at 61. The Court barely undertakes that requisite analysis, and instead uncritically assumes that all reversed cases can be salvaged. See ante at 194 n. 10, 613 A.2d at 1102 n. 10.

In the few cases in which the Court purports to “rehabilitate” a reversed death sentence, the analysis is untenable and unjustifiable. See Appendix B. To illustrate, the majority finds that Clausell’s case can be reported as a death verdict because the Court’s reversal on Gerald grounds did not “undermine the findings of deathworthiness.” Ante at 169 n. 5, 613 A.2d at 1090 n. 5. That conclusion is insupportable given the nature and scope of error that pervaded Clausell’s trial. In addition to the Gerald error, which itself impugns the jury’s sentencing conclusions, several errors related directly to the jury’s deliberations on the appropriateness of death as a punishment for Clausell’s crimes. For example, the trial court did not define the mental state in the “grave risk” aggravating factor, a problem exacerbated by its incorrect instruction on the aggravated assault charge. Clausell, supra, 121 N.J. at 344 , 580 A.2d 221 . The court’s explanation of the c(4)(d) aggravating factor was inadequate, ibid., as were its instructions on the mitigating factors. Id. at 344-45 , 580 A.2d 221 . By not explaining the function and meaning of mitigating factors, the instructions risked diluting the importance of the factors relied on by the defendant. Ibid. Moreover, the trial court improper *257 ly instructed the jury on the weighing process. Id. at 347 , 580 A.2d 221 . Further, the verdict form and instructions misled the jury on whether unanimity was required on the death verdict. Id. at 346 , 580 A.2d 221 . Finally, the jury instructions diminished the jury’s sense of responsibility for the death sentence. Id. at 346-47 , 580 A.2d 221 . Each of those errors significantly undermines the reliability of the decision to sentence Clausell to death independently of the Gerald issue. Accordingly, we can have no confidence that the death verdict accurately reflects the community’s assessment of Clausell’s deathworthiness. I do not believe that any case can be considered a death-sentenced case when, as in Clausell , the State has decided not to seek a retrial of the penalty phase and has accepted a life sentence. However, even if one were to accept the possibility that certain cases could be “rehabilitated,” Clausell’s case is not one of them. The Court’s erroneous “rehabilitation” of Clausell’s case clearly is an error, an error highlighted by its inclusion of both the original death judgment and the subsequent life judgment in several comparison groups.

I think the Court’s engaging in a theoretical construct to justify a death sentence when the reality is that the system generally is not imposing death sentences for similar crimes is highly inappropriate. Unlike the majority, I would not count a case twice when the person sentenced to death can be executed for the crime only once. And unlike the majority, I would not count cases that resulted in life sentences as death-sentenced cases. 2

*258 —D—

Compounding the Court’s mistake in counting death-sentenced cases is its decision to use a “precedent-seeking” approach, which, according to the Court, is to operate in some undefined “functional relationship” with the frequency approach. The Court’s explication of that aspect of its methodology is unclear and unconvincing, and, by the end, it appears to be virtually indistinguishable from traditional proportionality review that determines only whether the sentence is morally justified. Its irretrievable defects demonstrate that only a frequency approach can come close to measuring comparative proportionality.

The Court writes that “[i]f we are correct that death need not be normal or general to be a licit sentence, it follows that the frequency approach will not provide a sole-source, fail safe method of proportionality review.” Ante at 152-153, 613 A.2d at 1081. That conclusion hardly follows the premise. The Court is saying, in effect, that if it is correct in rejecting a high cut-off point for death-sentencing frequency, then the conclusion follows that no level of infrequency can justify a finding that a death sentence is aberrational or the product of inconsistent and arbitrary decisionmaking. The Court’s assumptions and analysis are simply wrong. Even accepting the notion that a moderate death-sentencing rate for similar cases would not impugn a sentence under review, it still may be true that a low or a very low death-sentencing rate for similar cases would impugn a sentence under review.

The Court, writes: “Rather than employ the frequency method as a cutoff, we believe that it will serve as a coefficient of consistency.” Ante at 153, 613 A.2d at 1081. What the Court *259 means by that expression is difficult to know. When the Court concludes that “[t]he lower the frequency, the more strictly the Court must scrutinize the case for the possible influence of impermissible factors,” ibid., it all but abandons two of its three stated objectives for proportionality review. Those were described initially as a protection against (1) arbitrary decisions made by juries, (2) arbitrary decisions made by prosecutors, and (3) decisions made by either juries or prosecutors based on impermissible considerations such as race. Ante at 133-135, 613 A.2d at 1070-1072. However, now only the third purpose seems to remain. In effect, the majority holds that defendants may be singled out unfairly for capital punishment unless they were the victims of discrimination based on race or sex or some other impermissible factor. See also ante at 181, 613 A. 2d at 1096 (stating that when conducting precedent seeking the Court’s “search should be for some impermissible or invidious factor or pattern”).

The precedent-seeking approach to proportionality review is based, at least loosely, on the traditional form of legal reasoning in which courts decide cases by following the rules enuncated in similar, previously decided, cases. When courts conduct proportionality review using the precedent-seeking method, they identify prior cases similar to the case under review and decide if the case under review more closely resembles the prior cases that resulted in death sentences or the prior cases that resulted in life sentences. Moore v. State, 233 Ga. 861 , 213 S.E.2d 829, 832-33 (1975), cert. denied, 428 U.S. 910 , 96 S.Ct. 3222 , 49 L.Ed.2d 1218 (1976); cf. Coleman v. State, 378 So.2d 640, 650 (Miss.1979) (finding no comparable cases, but holding death sentence excessive because case was less aggravated than the death cases to which it was compared). The obvious failing of that approach is that when there are cases in both the life-sentenced group and the death-sentenced group that resemble the case under review, the approach fails to account for the number of cases that fall within each group. If the cases that are most similar fall within the death-sentenced group, then the *260 sentence will be affirmed, even if the vast majority of very similar cases resulted in life sentences.

One might hope that the Court would respond to such a situation by taking into account the fact that the bulk of the precedent was in the life-sentenced group. In other words, one might hope that the Court would reverse the death sentence in the case under review because the similar precedents generally result in life sentences. If the Court did so, of course, it simply would be employing a surreptitious and less rigorous form of frequency analysis. It would engage in no distinctive precedent-seeking analysis at all.

When discussing its selection of a universe of comparison cases, the Court aptly observes that the selection of a universe for proportionality review “depends on the purposes to be served by that review.” Ante at 132, 613 A.2d at 1070. The Court then soundly concludes that because proportionality review was designed to ensure consistency in sentencing decisions by juries and prosecutors, a universe is required that includes all death-sentenced cases and all cases that could have resulted in the death sentences. Somehow the Court fails to appreciate that the reasoning it uses in choosing a universe applies with greater force to the choice of the standard for assessing proportionality. The frequency-analysis method is the only method of review that addresses the core concern of comparative proportionality review: whether the death penalty is being imposed in an arbitrary or non-evenhanded way. That conclusion is tautological. Because the core concern relates not to the abstract appropriateness of the death penalty but to the consistency with which the criminal justice system hands down death sentences in particular circumstances, the Court’s final determination must depend on the actual frequency with which the death sentence is imposed in those circumstances.

The Court’s discursive methodology appears to have been spawned from a concern that the frequency approach simply would not work standing alone. The majority suggests that a *261 frequency analysis, by relying too heavily on statistics, can ignore real human experiences. Ante at 154, 613 A. 2d at 1082. That objection is well taken, but it applies not so much to the frequency approach as to the entire enterprise of comparative proportionality review. To the extent that the criticism is valid, the conclusion to be drawn from it is that even soundly conducted proportionality review cannot prevent arbitrary sentencing. I, in fact, draw that very conclusion. See infra at 159-163, 613 A.2d at 1085-1087.

I agree with the Court insofar as it disapproves of formulaic or mechanical classification of defendants. To be sure, any decent proportionality review must focus, to the extent possible, on “real people,” ante at 154, 613 A.2d at 1082, and must take into account, to the extent possible, the unique characteristics of defendants and crimes that render certain cases especially aggravated. The Court fails to grasp that those features are considered in frequency analysis during the selection of cases for inclusion in the comparison groups whose death-sentencing rates are to be calculated. When a comparison case bears only superficial similarities to a case under review, it is excluded from the comparison group, notwithstanding the similarities, for the frequency calculation must be made for truly similar cases. 3 See Final Report, supra, 34 (suggesting that cases *262 sharing common facts may be dissimilar and unsuitable as comparison cases in a frequency analysis if they reflect distinctly different levels of culpability). In the end, however, once proper comparison groups have been formulated (taking into account special individual characteristics of the cases under consideration), the only issue that remains is the consistency with which the cases in those groups have resulted in death sentences or life sentences. 4 The Court mistrusts numbers, but numerical questions require numerical answers.

A second objection of the majority to the frequency approach is that it does not work well in cases such as Marshall’s, which are similar to few other cases. When the comparison groups are small, as the majority implicity recognizes, the frequencies may be the products of chance more than anything else. Ante at 167, 613 A. 2d at 1089. Unfortunately, replacing the frequency approach with a more obscure, less rigorous method like the precedent-seeking approach does not solve the problem of having few comparison cases. Irrespective of the method of analysis the Court chooses, the Court still must make extraordinarily weighty life-or-death decisions on the basis of only a handful of *263 cases. If the decision would be unreliable using a frequency method, it cannot be made any more reliable by using some other method that does not even purport to address issues of evenhandedness.

Instead of confronting the related problems of (1) how to draw empirical conclusions based on little data and (2) what to do when no reliable empirical conclusions can be reached, the majority simply pretends that proportionality review can be conducted without reference to the actual conduct of juries and prosecutors. If the Court truly were committed to evenhanded capital sentencing and to the proposition that there must be a coherent basis for distinguishing the cases in which the death penalty is imposed from those in which it is not, then the Court would not have “supplemented” the frequency approach with another non-empirical method.

The infirmity of the precedent-seeking approach is magnified when the reviewing court uses the approach and includes the case under review in the comparison group. The majority does precisely that. When the case under review is in the comparison group, the case under review always will resemble the death-sentenced group more closely than the life-sentenced group because the case under review always will resemble itself more closely than it resembles any other case. Here the cases most similar to the Marshall case, other than the Marshall case itself, are the cases of William and Herbert Engel, both of which resulted in life sentences. The Court concedes that those cases are almost on all fours with the subject case. See ante at 179-182, 613 A.2d at 1095-1096, but affirms Marshall’s sentence anyway. Had Marshall’s case itself not been considered as precedent, any precedent-seeking analysis would have dictated the reversal Marshall’s sentence. Only by including Marshall’s sentence in the death-sentenced group is the majority able to circumvent the Engel precedents. And, only by including Marshall’s sentence in the death-sentenced group is the majority able to overcome the fact that after ten years of murder prosecutions under the capital-murder statute there has *264 been no independent precedent — not even reversed precedent— to support imposition of the death sentence in cases like Marshall’s.

I should point out, moreover, that even if one were to accept the majority’s designation of reversed death-sentenced cases as valid death-sentenced cases, the controlling precedent still would come from the life-sentenced category. Of the so-called death-sentenced cases, the majority concludes that John Martini and James Clausell possessed higher culpability than Marshall, ante at 170, 183, 186, 188, 613 A.2d at 1090, 1097, 1098, 1099, while Anthony Di Frisco was of equal culpability. Ante at 183, 613 A.2d at 1096. The life-sentenced cases were almost equally divided between those of equal culpability (William Engel, Herbert Engel, Michael Rose, David Russo) and those of less culpability (Francis Brand, Miguel Melendez, and Randy Burroughs). Ante at 182-186, 613 A.2d at 1096-1098. The majority’s conclusion from that configuration is a striking departure from the conclusions suggested by the Special Master.

In a hypothetical given in the Special Master’s report to the Court, the Special Master assumes that six similar cases exist with two death sentences and four life sentences. Final Report, supra, at 33. Under the precedent-seeking approach, the Special Master describes the examination a court should undertake:

If the review case was deemed of comparable or greater culpability than the death-sentenced cases ..., the sentence would be affirmed. However, if some features of the review case made it appear less culpable than the two death-sentenced cases or comparable to one or more life-sentenced cases ..., the death sentence would be vacated.

[Id. at 34.]

That hypothetical fairly reflects the situation that the majority says it is facing, as it has deemed Marshall to be less culpable than two death-sentenced cases (Martini and Clausell) and of equal culpability to one death-sentenced case (Di Frisco) and four life-sentenced cases (H. Engel, W. Engel, M. Rose, and Russo). Despite the majority’s conclusion that the review case *265 is not of greater culpability than the death-sentenced cases, the majority misapplies the approach to conclude that Marshall’s case is not disproportionate.

The results of the Court’s precedent-seeking analysis are hardly surprising, for the Court does not even define that process, let alone conscientiously apply it. Ante at 181, 613 A. 2d at 1095 (asking whether some invidious factor or pattern has been discerned from the precedent); ante at 187, 613 A.2d at 1099 (concluding without analysis that Marshall’s sentence is not disproportionate). The Court’s judgments that some cases are similar to Marshall’s and that others are not are moral and hopelessly subjective and value-laden assessments. On the one hand, I cannot blame the majority for its failure to use the precedent-seeking method as the Special Master, and several other courts, suggest it be used. On the other hand, however, I find unconscionable the Court’s use of some utterly unstructured method of review, its calling that method “precedent-seeking,” and its claim that that method justifies affirmance of a death sentence when the most reliable measures dictate that the sentence be reversed.

—E—

By definition, a court engaged in frequency analysis must establish at least a loose cut-off point for permissible death-sentencing frequencies among comparison cases. As noted, the majority purports to hold that a death sentence is disproportionate only if other defendants with similar characteristics generally receive life sentences for committing factually similar offenses. Ante at 131, 613 A. 2d at 1070. I believe that a death sentence is disproportionate unless other defendants with similar characteristics generally receive death sentences for committing factually similar offenses.

In State v. Jeffries, 105 Wash.2d 398 , 717 P.2d 722 (1986), a case in which the Washington Supreme Court applied a precedent-seeking approach using a universe containing only death- *266 sentenced cases, Justice Utter dissented from the holding of the Court, arguing that a frequency approach was required by the Washington proportionality review statute. Id. 717 P.2d at 743 . Elaborating on the approach he thought required, he said: “If the frequency [of death sentencing for similar cases] is less than ‘generally,’ the death sentence should be reversed. Use of the word ‘generally’ suggests that the ‘threshold frequency’ at which a death sentence becomes appropriate is significantly greater than 50 percent.” Ibid.

If the death penalty is not imposed generally in cases similar to the crime under review, the death sentence should be vacated. Contrary to the position of the majority, the constitutional requirement that there be individualized sentencing does not mean that the State may impose the death penalty only occasionally for a certain type of crime. Ante at 152, 613 A.2d at 1081. The constitutional requirement of individualized sentencing means that the death penalty cannot be imposed invariably for a certain type of crime, not that the death penalty can be imposed sporadically for a certain type of crime. See Woodson v. North Carolina, supra, 428 U.S. at 296-98, 96 S.Ct. at 2987-88 , 49 L.Ed.2d at 956-57 .

Once the death penalty is not imposed generally for a type of crime, two conclusions can be drawn that should bar the imposition of the death penalty in the cases involving that type of crime. Both of those conclusions have constitutional dimensions. First, when the death penalty is not imposed generally for a type of crime, there is no societal consensus that those who commit that type of crime deserve the death penalty. “[J]ury sentencing constitutes the link between contemporary community values and the penal system.” Bey II, supra, 112 N.J. at 162, 548 A.2d 887 . When juries routinely find that the mitigating factors outweigh the aggravating factors even when no obvious mitigating factors are present, and when prosecutors follow their lead and routinely decide not to charge cases capitally, see, e.g., Jackson II, supra, 128 N.J. at 146 , 607 A.2d 974, 979 , those judgments must reflect public ambivalence *267 about imposing the death penalty in those cases. See Goodpaster, Judicial Review, supra, 74 J.Crim.L. & Criminology, at 802-03. Execution of defendants in those cases would be unconstitutional, for doing so would offend “ ‘ “the evolving standards of decency that mark the progress of a maturing society.” ’ ” Bey II, supra, 112 N.J. at 162, 548 A.2d 887 (quoting Gregg v. Georgia, supra, 428 U.S. at 173, 96 S.Ct. at 2925 , 49 L.Ed.2d at 874 (quoting Trop v. Dulles, 356 U.S. 86, 101 , 78 S.Ct. 590, 598 , 2 L.Ed.2d 630, 642 (1958))).

Second, when juries assessing a certain type of crime and defendant routinely find that the mitigating factors outweigh the aggravating factors, the conclusion follows that many of the defendants who receive the death penalty are no more deserving of death than are the defendants' whose lives are spared. There is no rational basis on which to distinguish those who are sentenced to death from those who are not. The ones who are sentenced to death just happen to be unlucky. When that occurs, the lives of the few who have been sentenced to death must be spared. State v. Ramseur, supra, 106 N.J. at 183, 197 , 524 A.2d 188 . Moreover, when that occurs, one can infer that the system has not been designed in a way that enables jurors to make rational, informed, and reasonably consistent decisions. See id. at 382-404, 524 A.2d 188 (Handler, J., dissenting).

For both of those reasons, I still would have dissented from the majority’s putative holding, even had that holding not been diluted by methodological errors. Our constitution requires no less.

—F—

The majority would draw comfort from the Special Master’s observation that contract murderers are among the defendants most likely to receive the death penalty, behind only those who kill multiple victims, those who have killed before, those who have kidnapped their victims, and those who kill public ser *268 vants. See ante at 171-172, 613 A. 2d at 1091; Final Report, supra, Table 7. That observation ignores both the fact that the only contract-murderer currently under death sentence is Marshall and the fact that, even counting reversed death-sentenced cases as valid, in no group of cases has the death sentence regularly been imposed. The fact that cases like Marshall’s are among the most likely to result in a death sentence is inconsequential unless the cases that are most likely to result in a death sentence actually result in death sentences with regularity. Because, after ten years, no types of capital murder cases regularly result in death sentences, the fact that contract-killings are more likely to result in the death penalty than other types of killings cannot alter or mitigate the irregularity attendant to death sentencing in general.

Ill

The Court’s standards for proportionality review are profoundly deficient. Only by proceeding as it does from a misguided understanding of the essential requirements of proportionate sentencing is the Court able to uphold Marshall’s sentence. To reiterate, there is a sounder approach: a frequency analysis based on actual sentencing practices. A conscientious application of that approach leads inexorably to one conclusion: the death penalty rarely is imposed in New Jersey in cases like Marshall’s.

—A—

One obstacle to conducting proportionality review in this case — and in any case — is that many defendants do not have the benefit of thorough penalty hearings. If the universe contains cases in which some defendants were able to establish a record of mitigating evidence and others were not, the comparisons made among the cases in the universe necessarily will be prone to error. Moreover, the comparisons will work to the *269 disadvantage of defendants like Marshall whose trials did not include the presentation of mitigating evidence.

As I pointed out in my dissent in Marshall I, Marshall’s penalty trial was a perfunctory exercise that involved little more than the recitation by counsel of facts brought out during the guilt phase. 123 N.J. at 242 , 586 A. 2d 85 (Handler, J., dissenting). The defense presented no new information concerning defendant’s character, his personal history, or his relationship with his children. The only evidence the jury considered was that bearing on Marshall’s guilt, not on whether he deserved to live. Having convicted Marshall of capital murder, the jury already had rejected that guilt-phase evidence. In the absence of any other evidence or a genuinely independent penalty hearing, Marshall’s fate effectively had been decided before the jury even commenced sentencing-phase deliberations. “There is no reason to suppose that the evidence offered by defendant at the guilt trial by way of defense to the charges could be more than marginally helpful to whether he deserved only life imprisonment.” Ibid. The brevity of the hearing, coupled with the fact that it was held only hours after the guilty verdict had been announced, severely undermined defendant’s ability to establish the applicability of particular mitigating factors. Today that flaw hampers our ability to conduct a proportionality review of his sentence. However, even in the absence of a complete record of mitigation in Marshall’s case, the data still suggest that Marshall’s sentence is disproportionate when compared with those of other defendants who committed similar crimes and who, like Marshall, failed to produce substantial mitigating evidence.

—B—

Like the majority, I acknowledge that proportionality review must use several comparison groups, each reflecting a different measure of similarity. Each comparison group must serve as a *270 cross-check on the others. In this case, because the results derived from the various comparisons are consistent, the ultimate conclusion derived from them may be regarded as reliable. Had the results been inconsistent, a closer re-evaluation or a refined approach would have been necessary.

Distilled to its essence, Marshall’s crime was a contract murder of a family member for pecuniary gain. The murder was carefully planned and was not committed in the heat of passion. However, the murder was not especially brutal — for a capital murder — and did not involve torture or depravity of mind. Marshall had no criminal record prior to his conviction. I consider the case to be similar, for the purposes of proportionality review, to several types of cases. Closest to Marshall’s case, perhaps, are those cases involving contract killings. Because I consider the ends of a crime to be as important as or more important than the means, I also consider Marshall’s crime to be similar to those cases involving murders for pecuniary gain and murders of spouses, family members, and friends.

For the most part, therefore, I settle on the same criteria for creating comparison groups that the majority uses. I rely heavily on comparison groups comprised of cases involving the following types of criminals: contract-murderers (both principals and agents); spousal murderers who did not kill in the heat of passion and whose victims were defenseless; and murderers who robbed or kidnapped their victims, planned their crimes extensively, had pecuniary motives, and deceived or entrapped defenseless victims. In addition, I consider two comparison groups proposed by the Special Master that the majority did not consider: cases involving murderers who killed for pecuniary gain whose crimes did not involve burglary, robbery, or kidnapping, Final Report, supra, at 80-84; table 7, app. E at 5, app. F at 8, and cases involving murderers who robbed acquaintances, relatives, or friends. Final Report, supra, at 80-84; table 7, app. E at 3, app. F at 5-6.

*271 In Appendix D to this opinion the death-sentencing frequencies for several comparison groups are listed. Those frequencies are based on a universe of all death-eligible cases. All living defendants not currently under death sentence are counted as life-sentenced defendants. No case is counted more than once, unless it is a case in which a defendant was sentenced separately for more than one murder. I have excluded from the universe all defendants not currently under conviction for the homicides of which they have been accused. All comparison groups are listed in Appendix F.

A frequency calculation for each comparison group reveals that the death sentence in Marshall’s case is disproportionate. For example, of the nine contract murderers convicted of murder in the last decade, only one, Marshall, was sentenced to death. Of the ten murderers who killed for pecuniary gain and whose killings did not involve robbery, burglary, or kidnapping, only one, Marshall, was sentenced to death. Of the eleven murderers who killed while robbing friends, acquaintances, or relatives, none was sentenced to death. In most of the comparison groups, the death-sentencing frequency hovers around one in ten. In only one comparison group — that consisting of spousal murderers — does the death-sentencing frequency reach as high as one in three; in that group there were only three cases, and Marshall’s is the one that resulted in the death penalty.

Marshall’s case has achieved considerable notoriety. That notoriety, coupled with the highly unusual nature of the facts of the murder itself, might lead one to believe that Marshall’s case must be more aggravated than most other capital-murder cases, and that his sentence must be proportionate. In fact, although Marshall’s was a horrible crime, it was no worse (by many different standards) than several murders committed in New Jersey since 1982, when the current capital-murder statute *272 was enacted. Yet since 1982 only four death sentences 5 have been returned that still may be understood as valid death sentences, and the vast majority of murderers who committed crimes comparable to and more heinous than the killing of Maria Marshall have received life sentences. Those facts operate to make Marshall’s death sentence disproportionate.

—C—

Marshall’s sentence is disproportionate when compared both to defendants who committed other crimes and to defendants who participated in Marshall’s crime with him. His case exemplifies the intra-case disparity that is rampant in our death-penalty system. In State v. Di Frisco, 118 N.J. 253 , 571 A.2d 914 (1990), I expressed concern that the system allowed the hired killer to be tried for capital murder while the principal was not pursued or indicted. Id. at 302-05 , 571 A.2d 914 (Handler, J., concurring in part and dissenting in part). The exercise of prosecutorial discretion appeared to have been unguided and unsupervised, and appeared to have resulted in arbitrary and disproportionate outcomes. The concern over disproportionality within a case also was present in the Clausell case, see State v. Clausell, supra, 121 N.J. at 309, 312 , 580 A.2d 221 (principal not prosecuted), and in the Brand case. See Final Report, supra, at 54, 62.

Here Marshall was prosecuted capitally as a contract-murder principal, but the death penalty was waived against the confessed contract-killer agent, Billy Wayne McKinnon, in order to secure his testimony against Marshall. See Final Report, supra, at 11 n. 8. McKinnon pleaded guilty to conspiracy to commit murder and received a sentence of five years. The disparate results are even more magnified as the alleged shoot *273 er, Larry Thompson, was acquitted of capital murder and conspiracy — his defense was that McKinnon had been the shooter — and the other participant, Robert Cumber, was sentenced to thirty-three years for his role as an accomplice in the conspiracy. Intra-case disparity alone might not justify the vacation of Marshall’s death sentence, but it does bolster the conclusion that his sentence is disproportionate.

IV

Although I believe proportionality review is required constitutionally, I do not believe that its use can save an otherwise unconstitutional death-penalty statute from constitutional infirmity. Proportionality review has its limits. It cannot keep its own promise — it cannot assure evenhandedness, uniformity, or consistency in the imposition of the death sentence. It does not and cannot overcome the arbitrariness that plagues the imposition of the death penalty. Our experience with proportionality review itself now confirms that chronic and profound constitutional deficiencies endure in the New Jersey system for administering capital punishment. Finally, proportionality review is not simply a final straw placed on the back of the public. It is a massive undertaking that adds substantially to the enormous governmental, institutional, and social burdens that are imposed and sacrifices that are exacted as a result of the decision to maintain capital punishment as part of our criminal justice system.

—A—

The fallacy of proportionality review inheres in its attempt to identify precisely what makes some cases similar and others dissimilar. It inheres in the notion that we, as judges, can identify the qualities that make some murderers deathworthy and others not. The Court’s efforts to do that — by assessing moral culpability — amply demonstrates that proportionality review often serves to perpetuate rather than eliminate arbitrary *274 sentencing. Moral culpability determinations may appear to be a more sophisticated and refined exercise than the statutory process by which jurors determine death eligibility and death-worthiness. As I have often said, the latter process does not sufficiently restrict the discretion of jurors or direct and guide them to sound and reliable decisions concerning capital murder and life-and-death judgments. E.g., State v. Ramseur, supra, 106 N.J. at 382-404 , 524 A.2d 188 (Handler, J., dissenting). The objective of proportionality review is to enable the reviewing court to compensate for such subjectivity and correct its unfair results, as when similar defendants committing similar crimes do not in fact receive the same sentence. The irony of proportionality review is that the reviewing court itself resorts to extra-statutory factors that are also inherently subjective. Because the Court at this end-point of capital prosecution validates or invalidates the defendant’s death sentence based on its own judgment of culpability, the Court, make no mistake about it, becomes the ultimate sentencer.

As the majority itself recognizes, a “value judgment is built into practically every measurement” of proportionality. Ante at 119, 613 A.2d at 1064. That inescapable entanglement with subjective moralistic judgment renders proportionality part of the problems rather than the remedy to the problems inherent in the statutory scheme. The Court’s decision only perpetuates the myth that the death penalty can be imposed in an objective, principled, non-random manner, and that any given death sentence can be validated as a fair, reliable expression of just punishment or community values. The Court thus makes comparison after comparison with absolutely no legal or logical foundation for support. Is it worse to kill for money or for hatred? Is it worse to kill over a woman or over a dog? Is it worse to kill to support a gambling habit or to support a drug habit? Is it worse to kill a relative or a stranger? To pose those questions is to pose insoluble moral conundrums. Because life or death is determined by the answer, no process will give us an answer that can assure us that it is right. We are *275 then not reasonably better off then when we first heard the brilliant insight of Justice Stewart, who fingered the critical deficiency of capital punishment: its freakish and wanton selection of persons who are to be sentenced to death. The selection of defendants for death always will be freakish and wanton, proportionality review notwithstanding.

—B—

In Ramseur , I expressed the view that the death penalty is unconstitutional because it offends contemporary standards of decency. I there observed with respect to paragraphs 1, 5, and 12 of Article I of the New Jersey Constitution:

Both the language and structure of these provisions affirm that the life of the individual — including a quality of life embodied in liberty, safety, security and happiness — is accorded the highest value and greatest protection possible under the State Constitution. Government cannot take life, or detract from the essence of life, or restrict the rights that are supportive of life, based on arbitrary and unreasoned actions or invidious factors; even when sanctions must be imposed, the State may not impose any punishments that are cruel and unusual. It follows logically from these provisions and their obvious purpose that when life itself is at stake government must be its most scrupulous to avoid any wrongful harm.

[ 106 N.J. at 370-71 , 524 A.2d 188 (Handler, J., dissenting).]

I felt that our own experience with the application of capital punishment and the history of its use and evolution revealed valid measures of our common moral values as they related to the death penalty. Id. at 371, 524 A.2d 188 . The clearest feature of that historical experience was “the decreasing use of the capital murder sanction.” Ibid. The societal disaffection from capital punishment was reflected in the fact that for almost three-quarters of a century relatively few murderers faced capital trials. Ibid. The courts judicially barred practices that could provide “a ready and facile road to the gallows.” Id. at 372, 524 A.2d 188 (quoting State v. Genz, 57 N.J.L. 459, 462 , 31 A. 1037 (Sup.Ct.1895)). The Legislature enacted added protections to the imposition of capital punishment. Our actual experience demonstrates that the imposition of the death penal *276 ty was becoming progressively and dramatically less tolerated and less frequent. Id. 106 N.J. at 372-73 , 524 A.2d 188 . History and experience teach us that, despite a commonly held belief that capital punishment can be a useful part of a criminal justice system, capital punishment founders on stronger and deeper feelings that will not abide the immorality in the arbitrary or capricious taking of life.

I continue to believe that our death penalty regime cannot be reconciled with contemporary standards of decency. Thus our current experience indeed validates our past experience. Our hesitancy and ambivalence, our angst and travail over capital punishment represent more than a quixotic quest for an ideally fair system. Rather, they reflect a very pervasive and deep-seated antipathy against the taking of a life when the fairness in taking that life cannot be assured. Plainly, proportionality review is no assurance, although we may pretend that it is. Capital punishment continues to offend our common decency.

Our recent experience ratifies the lesson to be taken from our failure to devise a capital punishment system that is fair. Death-sentencing rates since the death penalty statute was enacted have declined dramatically. Not one jury in New Jersey has returned a death sentence since 1990. See Final Report, supra, at 15-16. Prosecutors, who base their charging decisions in part on their own predictions of what juries will do, charge cases capitally far less often than they once did. Id. at 20. Prosecutors sometimes strain to avoid capital prosecutions. E.g., Jackson II, supra, 128 N.J. at 142 , 607 A. 2d 974, 977 (Handler, J., dissenting).

I said in Ramseur :

Our history thus not only confirms that society places extraordinary value on individual life, but also teaches us society's moral judgment that the death penalty should not be applied unless it is right to do so, unless, in other words, substantive and procedural protections have been maximized so that the risk of arbitrariness is reduced and the possibility of the ultimate injustice is eliminated.

*277 [State v. Ramseur, supra, 106 N.J. at 374 , 524 A. 2d 188 .]

Our experience over the last ten years with capital punishment, capped by the lessons wrought from our gargantuan effort to attain proportionate sentencing, confirms the constitutional impossibility of capital punishment.

—C—

Our efforts to achieve satisfactory proportionality review has been a gigantic undertaking. It has taken about three years, untold hours, substantial outlays of moneys, and the absorption of significant resources to attempt to develop an adequate methodology. That fresh experience makes highly relevant the consideration anew of the enormous governmental, institutional, and social sacrifices entailed in capital punishment and the obvious, disturbing policy implications flowing from those sacrifices.

Although the overall costs for estimating a capital trial vary depending on the jurisdiction and the costs included, the estimates uniformly confirm that the death penalty is a costly mechanism. In California, the fourteen cases recently affirmed by the California Supreme Court cost an average of $1.7 million each, though another case cost $5 million and most cases range from $750,000 to $1 million for the trial and direct appeal. Stephen Maganini, Closing Death Row would Save State $90 Million a Year, The Sacramento Bee, Mar. 28, 1988, at A14 (estimating the average death-penalty case at $592,000 compared to $93,000 for the average murder case). A study prepared for New York estimated the costs to bring a capital case up to the point of a death sentence at $1,498,0Ú0. New York State Defenders Association, Inc., Capital Losses: The Price of the Death Penalty for New York State, Apr. 1, 1982, at 19 (Capital Losses); see also Dave Von Drehle, Bottom Line: Life in Prison One-Sixth as Expensive, The Miami Herald, July 10, 1988, at 12A (citing studies from Maryland and Kansas suggesting that the total additional cost for a bifurcat *278 ed capital trial over a non-capital murder trial is estimated at between $36,000 and $116,700) (Bottom Line); Bar Panel Reports Death Penalty is Too Expensive, Newsday, Feb. 10, 1989, City Section 34 (reporting that New York State Bar Association had estimated that the cost of a capital case is $2 million in Texas and more than $3 million in Florida).

In New Jersey, sentencing one person to death has been estimated to cost about $7.3 million. Tabak and Lane, The Execution of Injustice, 23 Loyola of L.A.L.Rev. 59, 136 (1989). At every stage of the judicial proceedings the costs are higher for the death-penalty case than for the non-capital case. The first and most enduring cost to the state is the assignment of counsel. Ninety percent of those on death row nationwide cannot afford counsel and receive court appointed counsel. Howard Goodman, Helping Those Sentenced to Die, Philadelphia Inquirer, May 11, 1991, at B1; Leigh B. Bienen, et al., The Reimposition of Capital Punishment in New Jersey: The Role of Prosecutorial Discretion, 41 Rutgers L.Rev. 27, 163 n. 671 (1988) (noting that 81.5% of 703 homicide cases in New Jersey between 1982 and 1986 were handled by the State Public Defender’s office or court appointed counsel). The state therefore pays for those defenses, straining the resources of the Public Defender’s Office. For example, an audit established that New Jersey spent nearly $3 million extra in 1984 to defend indigent defendants charged with capital crimes. Jo Astrid Glading, Bills to Defend Indigents Climbing, Trenton Times, July 4, 1985. Because of the complexity of capital cases, the State Public Defender assigns two defense attorneys to each case, compared to the single assignment for most murder cases. Ames Alexander, Penalty Comes at High Price, Asbury Park Press, July 22, 1991, at 1. Normally two prosecutors are needed in New Jersey death penalty cases, further increasing the costs throughout the proceedings. Michael Booth, Mercer Leads State in Pending Death-Penalty Cases, Trenton Times, Oct. 17, 1988, at Al. For example, the prosecution estimated that the Ramseur case cost $60,000 a year to litigate. Pat R. *279 Gilbert, Bills Add up in a Hurry for Death Row Inmates, Trenton Times, Feb. 14, 1988, at Al. Moreover, competent counsel on both sides are essential not only to ensure the fairness of the proceedings but because inadequate legal services will lead to greater costs as cases progress. “The dangers of additional extraordinary and successful post-conviction applications increases, with the likelihood for new trials, if counsel at the trial level is not thoroughly effective.” Criminal Justice Section, New York State Bar Assn., Reinstitution of Death as Punishment, Feb. 13, 1989, at 4 (Reinstitution of Death).

Although nine out of ten non-capital cases resolve themselves in plea negotiations, few capital defendants enter guilty pleas. See Margot Garey, The Cost of Taking a Life: Dollars and Sense of the Death Penalty, 18 U.C. Davis L.Rev. 1221, 1247 n. 114 (1989) (stating that capital cases go to trial ten times as often as non-capital cases). In effect, the result of the decision to seek the death sentence is trial. Bottom Line, supra, at 12A (noting that capital cases almost always require a trial because, due to the extensive pretrial publicity and notoriety of the cases, prosecutors are reluctant to plea bargain).

The pretrial process is longer and more complicated in capital cases than in non-capital criminal cases. Maganini, supra, at A1 (stating that non-capital trials average one day of pretrial motions, whereas capital trials average twelve days); see also Bienen, supra, 41 Rutgers L.Rev. at 242-45 (describing the impact that the allegation of an aggravating factor has throughout proceedings). The usual number of pretrial motions in a non-capital case is five to seven, whereas the typical capital case requires the filing of between ten and twenty-five motions. Capital Losses, supra, at 12. While pretrial practice plays an important role in most criminal cases, it takes on an added significance in the capital case. The capital case requires not only motions on the underlying offense but motions attacking the death penalty itself, for example, its constitutionality or the sufficiency of evidence to support alleged aggravating factors. Capital cases also generate a need for motions to *280 appoint experts and employ investigators, and the notoriety of capital cases requires change of venue and sequestration motions. Garey, supra, 18 U.C. Davis L.Rev. at 1247-51. The standard motions concerning discovery and the suppression of evidence also become critical, taking more time to prepare and stimulating heavy litigation. Capital Losses, supra, at 12-13.

Preparation for the pretrial motions, as well as the guilt and penalty trials, can take months of expensive investigation. Defense counsel often call psychiatrists, family members, former teachers, even accomplices in past crimes. Those witnesses must be located, interviewed, and persuaded to cooperate, all of which requires competent investigators, time, and money. Bottom Line, supra, at 12A. School, work, military, and medical records must be located and retrieved. In California, a typical death-penalty defense costs an additional $25,000 to $50,000 in special investigation costs and $15,000 for psychiatrists or other expert witnesses. Maganini, supra, at A14 (quoting a San Francisco Public Defender); Capital Losses, supra, at 15 (listing fees of various experts: crime scene reconstructionists and blood stain analysts range from $700 to $1,000 per day; psychiatrists’ average daily fees are $700, or $125 to $150 per hour of courtroom testimony; polygraph expert average bill amounts to $500; eyewitness identification expert courtroom fee is $500 per day). Specifically, of the $296,449.41 the Public Defender’s Office spent to defend Thomas Ramseur at trial, the largest portion, $191,881, was spent on psychiatrists. Gilbert, supra, at Al. The Public Defender spent more than $75,000 successfully attacking the makeup of grand and petit juries in Atlantic County. Joseph F. Sullivan, New Jersey Defenders Battle Death Law, New York Times, Feb. 26, 1991, at B1. Steep investigation expenses are not limited to the defense side. The cost of the prosecutor’s investigation of the Robert Marshall case alone was more than $200,000. Alexander, supra, at 1 (quoting detective at Ocean County prosecutor’s office).

*281 Jury selection in a death case is on the average 5.3 times longer, and hence more costly, than in a non-capital case. Garey, supra, 18 U.C. Davis L.Rev. at 1257 (stating that courtroom time alone may increase cost to the system by as much as $87,440). Jury selection, a process that ordinarily takes three days, can consume two to six weeks of court time. Maganini, supra, at A14 (reporting that jury selection routinely takes six weeks or more); Capital Losses, supra, at 13 (reporting that individual examination of jurors can take two to four weeks). Potential jurors must be questioned individually to explore their personal biases and feelings on capital punishment, an inquiry that does not arise in non-capital cases. Jury voir dire also may increase the cost of experts, as both sides use jury experts to assist in the selection of fair, impartial jurors. Id. at 15 (reporting that Witherspoon consultant fee is $100 per hour); Sullivan, supra, at B1 (reporting that defense counsel at capital trials in New Jersey routinely hire juristic psychologists to help pick a jury).

Court testimony at average non-capital trials covers three court days at an average cost of $6,200 a day. By comparison, capital-murder trials take thirty days of court time at an estimated $7,500 a day. Maganini, supra, at A14 (citing a California report).

The penalty trial in death cases has no analogy in non-capital cases. The resources expended in pretrial investigation bear fruit here, as family members, friends, neighbors, teachers, and employers are called to provide the jury with a complete retrospective of the defendant’s life from prenatal care to the date of sentence. A majority of New Jersey defendants rely on mitigating factors that touch on mental conditions. Bienen, supra, 41 Rutgers L.Rev. at 259-61 & table 51 (reporting that in sixty-nine New Jersey cases proceeding to penalty trials, mitigating factor c(5)(a), mental or emotional disturbance, was alleged in 59.4% and mitigating factor c(5)(d), impairment of defendant’s capacity to appreciate the wrongfulness of his or her conduct, in 68.1% of those trials). In order to persuade the *282 jury that the alleged mental and emotional characteristics exist, the defense must rely on expert testimony. The prosecution, in seeking to rebut that evidence, also relies heavily on psychiatric testimony. See Bottom Line, supra, at 12A; Capital Losses, supra, at 17.

Moreover, the costs stemming from the high stakes involved in a capital prosecution are incurred even when the jury acquits, convicts of a lesser offense, or returns a life sentence. In 1988, for example, none of the twenty New Jersey capital juries returned a death sentence. Kathy Barrett Carter, Jurors Avoid the Responsibility for Killing a Stranger in 20 Trials, Star Ledger, Dec. 18, 1988, at 32. Of the ninety-four homicides that went to trial in New Jersey between 1982 and 1986, sixty-nine resulted in capital murder convictions, but only twenty-five defendants were sentenced to death. Bienen, supra, 41 Rutgers L.Rev. at 159-60. In other words, although the heavy machinery of a capital trial is invoked in nearly twenty percent of all homicides, only á quarter of those trials resulted in a death sentence. Id.; Final Report, supra, Tables 2 & 3 (reporting thirty percent rate between 1983 and 1991); John Rofe, Death Penalty Evades Prosecutors, The Dispatch, June 10, 1987, at 1 (reporting statewide rate around one sentence in three penalty trials). Further, the deterrence of capital punishment may be questioned in light of the fact that the number of persons executed likely will be no more than a minute percentage of those accused, tried, and convicted of capital murder. We must recognize a parallel concern “with the disproportion between the enormous costs — financially and in terms of scarce, overburdened judicial, prosecution, and defense services — and the putative ‘benefit’ of the actual infrequent imposition of death, given its infrequency.” Reinstitution of Death, supra, at 6.

Every death-sentenced defendant is entitled to a direct appeal to this Court. Those mandatory reviews for defense alone cost an average of $37,740. Bottom Line, supra, at 12A (quoting consultant for American Bar Association); see also Maganini, *283 supra, at A14 (estimating that typical death penalty appeal takes about 1,000 attorney hours, or about $62,000). Studies show that the prosecution matches those defense attorney costs dollar-for-dollar. Bottom Line, supra, at 12A; Capital Losses, supra, at 10 n. 28 (using defense-prosecution ratios of one-to-one at appellate stage). The length of the pre-trial and trial proceedings also means voluminous transcripts must be prepared for appeal. Ramseur’s trial transcript cost $33,000. Gilbert, supra, at A1.

The complexity of our capital jurisprudence and the shear volume of the records in capital trials compel appellate courts to expend inordinate energy on capital cases. See, e.g., Maganini, supra, at A14 (reporting that the California Supreme Court spends one half its time reviewing death cases); Andrew H. Malcolm, Capital Punishment is Popular, But So Are Its Alternatives, New York Times, Sept. 10, 1989, at A4 (reporting that Florida’s Supreme Court devotes one-third of its time to capital cases); Jonathan E. Gradess, Execution Does Not Pay, The Washington Post, at 5 (noting that thirty percent of the docket of the United States Court of Appeals for the Eleventh Circuit consists of death penalty cases).

Half of all death sentences are overturned on appeal. David J. Gottlieb, The Death Penalty in the Legislature: Some Thoughts About Money, Myth, and Morality, 37 Kan.L.Rev. 443, 449 (1989); Dave Von Drehle, Capital Punishment in Paralysis, The Miami Herald, July 10, 1988, at 1A ("Punishment in Paralysis")', Gradess, supra, at 5; David A. Kaplan, Death Mill, USA, National Law Journal, May 2, 1989, at 38 (observing that Louisiana Supreme Court has reversed 36.8% of the death sentences to come before it; Texas 32 .9%; Florida, 46.3%). Even if the case is upheld by the state courts, federal courts, reviewing the sentences in habeas corpus proceedings, find reversible error in 40.7 percent of the death sentences. Kaplan, supra, at 38.

*284 When prosecutors pursue retrials, many of the costs discussed above are repeated. See, e.g., Renee Winkler, Prosecutors Pleased but Wait for Chance to Study Opinion, May 6, 1987, at 27 (quoting a Camden County prosecutor as concluding that a retrial of sentencing phase requires “a virtual retrial of the whole case”); Matt Lait, Second Death Penalty Seen as Costly Effort, Los Angeles Times, Apr. 5, 1992, at B1 (quoting attorneys and court officials in California as estimating that retrial of one defendant would eventually cost county taxpayers between $500,000 to $1 million, even though the defendant changed his plea to guilty early in the trial); Ian Gray, A Shameful — and Expensive — Logjam at Death’s Door, Los Angeles Times, Dec. 17, 1989, at M3 (reporting that retrial carried a $4.6 million price tag after defendant had received ineffective assistance of counsel during the original trial). Although the high costs continue, the death-sentencing rate on retrial is low. See Nick Fox, Condemned Inmates may Win Review of Sentences, Bergen Record, Mar. 2, 1988, at A1 (citing NAACP Legal Defense Fund figures showing that of the 696 resentencing trials nationwide, only 197 resulted in a second death sentence).

Additional appellate costs are incurred when the death-sentenced defendant seeks post-conviction relief in state and federal courts. In most states, volunteer lawyers must be found for collateral review in which they can incur out-of-pocket expenses of $10,000 and spend an average of 665 hours. Linda Greenhouse, Right to Death-Row Lawyer Curbed, New York Times, June 24, 1988, at A8; see Stephen Wermiel, Delays on Death Row Fuel Hot Debate, Wall Street Journal, Mar. 6, 1989, at 136 (noting that federally funded death-penalty centers, which together have budgets totalling $5.9 million budget, recruit and train lawyers, and that the 1989 federal budget allocated $10 million for appointing attorneys to represent death-row inmates); Punishment in Paralysis, supra, at 1A (estimating that capital cases reviewed in federal courts will cost the nation’s taxpayers $30 million a year for appointed counsel). *285 New Jersey, however, is one of the few states that authorize the Public Defender to represent a capital defendant through all stages of judicial review. Thus, the costs of post-conviction review also will be incurred by the state. Alexander, supra, at 1.

The strain on the court system is also felt in non-monetary terms. The burden on the judiciary increases when courts must address last-minute appeals during midnight vigils. See Dave Von Drehle, Fairness was Fatal Blow to Fast Executions, The Miami Herald, July 11, 1988, at 1A (describing midnight appeal handled by Justice Powell); Punishment in Paralysis, supra, at 1A (reporting that the Florida Supreme Court hears thirty to forty last minute appeals a year). In those frenzied appeals, reams of documents and legal arguments are deposited on the courts’ doorstep. Incidental costs also increase during these last-chance proceedings, as the process shifts into overdrive, air travel must be arranged and overnight couriers used. Dave Von Drehle, Political Pressure Thwarts Clemency, The Miami Herald, July 12, 1988, at 1A.

Clemency proceedings before the Governor are another stage in the process, which increases the cost of executing a criminal defendant. Amnesty International USA, The Truth About the Death Penalty in the United States is Real News, June 1990.

The prisons, entrusted with housing and protecting the defendants awaiting execution, account for additional costs. When defendants face a death sentence, they receive individual cells and increased security. Housing a death sentenced inmate in the New Jersey State Prison’s Capital Sentence Unit at Trenton costs an additional $3,000 a year or 20 percent more than housing a non-death-sentenced inmate in the general prison population. Gilbert, supra, at A1 (noting that the close supervision of death row inmates requires two officers to escort an inmate anywhere and notations in the file every ten minutes on the inmate’s activities); Pamela Geurds, The “Fantasy World” of Life on Death Row, Trenton Times, Feb. 15, 1988, at A10 *286 (reporting that death row inmate must be strip searched before and after greeting a visitor); Jack Knarr, Spaces on Death Row Available at $22G a Year, Trenton Times, Sept. 21, 1987, at A3 (reporting that the cost per inmate in Trenton’s capital sentence unit is $22,000 a year, or $2,000 above that of the general population because of additional supervision). The total cost to operate the special unit amounted to $465,355 in 1986. Ibid.; see also Maganini, supra, at A14 (noting that California spends an extra $2.8 million for special housing of death row inmates); Capital Losses, supra, at 23 (estimating that Florida spends an additional $15,000 per year for each inmate to maintain individual cells and to separate the death row inmates from the general prison population). In Florida, once a death warrant is signed, the guards keep a twenty-four-hour watch to ensure that the inmate does not commit suicide. The cost in overtime for those guards is $13,800. Bottom Line, supra, at 12A.

The actual execution can generate additional costs. Georgia spent more than $250,000 for the anticipated but aborted execution of one inmate in 1980. Capital Losses, supra, at 27. Incidental costs such as special telephone lines between the prison and the United States Supreme Court and Governor’s office are necessary as are costs for extra police personnel for crowd control, security, and the shutdown of federal airspace over the prison. Ibid. In New Jersey, volunteer executioners — two are used so that neither knows who was responsible for the actual lethal injection — are paid $500 for the procedure. Pat R. Gilbert, N.J.’s Lethal Injections are Ready, Trenton Times, Feb. 15, 1988, at A1.

Once those figures are totaled, the sums are staggering. Between 1973 and 1988, Florida spent an estimated $57 million on death-penalty costs. Punishment in Paralysis, supra, at 1A. The re-enactment of the death penalty cost Texas $183.2 million. Leigh Bienen, No Savings in Lives or Money with Death Penalty, New York Times, Aug. 7, 1988, at 24. Because the capital machinery is summoned for many cases in which the *287 jury returns a life verdict or the death sentence is reversed on appeal, much of those costs are “unsuccessful” expenditures in the death-penalty system. For example, the $57 million expended in Florida resulted in eighteen executions, meaning that each execution cost $3.2 million. Bottom Line, supra, at 12A; Amnesty International USA, supra (reporting Florida’s cost at $3,178,000 per person executed). In California, where one out of every ten cases filed as a capital case results in a death verdict, the estimated six executions per year will each cost taxpayers about $15 million. Maganini, supra, at A14.

By contrast, the cost to house an inmate for life, based on an average life of forty years, is $930,240. Id. at Al; Capital Losses, supra, at 23 (estimating cost in 1978 at $602,000 for forty years of imprisonment); see also Andrew H. Malcolm, Society’s Conflict on Death Penalty Stalls Procession of the Condemned, New York Times, June 19, 1989, at B10 (Society’s Conflict) (stating that “Life imprisonment would require massive prison construction, costing about $55,000 per cell”). Thus the public cost of each battle over a death sentence, including courts, judges, prosecutors, defense attorneys, and guards, may be three times the cost of imprisoning a typical thirty-year-old defendant for life. Society’s Conflict, supra, at B10 (quoting consultant for ABA).

Ultimately the question is one of policy, whether the money expended could be dedicated to more productive uses. See, e.g., Dave Von Drehle, Cries for Change, The Miami Herald, July 13, 1988, at 1A (quoting Florida Supreme Court Justice McDonald as saying, “I think society needs to ask itself if the results justify the cost,” and former Florida Supreme Court Chief Justice England as asking, “Is the value derived really worth all the trouble?”); Maganini, supra, at Al (reporting that small California county was forced to cut police services to finance capital prosecution). The millions of dollars spent on the death penalty could be spent to prevent crime, to compensate victims, or other criminal justice costs, see Reinstitution of Death, supra, at 6. In these times of scarce governmental *288 resources, million dollar expenditures on capital punishment can hardly be justified.

—D—

No matter how sincere and conscientious, a genuinely civilized society cannot, consistently with its own legal and moral code, punish by death its most reprehensible criminals. Perhaps law, which limits its logic and vision to fair punishments in the administration of criminal justice, could accomplish capital punishment. But law cannot ignore the dictates of morality or reject the influences of decency that a civilized society absorbs into its criminal justice system.

We have in our society a constitution that places a moral value on human life and a jurisprudence that accommodates decency as a component of justice. To oppose capital punishment one need not stigmatize a sentence of death as atavistic or barbaric. Governments struggle to protect their citizens, and they attempt to mold criminal laws that will effectively punish and deter criminals. Capital punishment understandably may be thought to be responsive to a public need for retribution and deterrence. That need, however, is not repudiated or denigrated if it must yield to, because it cannot be reconciled with, a much stronger need on the part of a civilized society for a criminal justice system that recognizes both a core morality that values life and a basic common decency that insists on fair and humane punishments.

GARIBALDI, J., concurring in result.

For affirmance — Chief Justice WILENTZ and Justices CLIFFORD, POLLOCK, O’HERN, GARIBALDI and STEIN— 6.

For reversal — Justice HANDLER — 1.

*289 APPENDIX A

Cases in Which Death Sentences Have Been Vacated and Not Re-imposed

The majority counts the following cases as death-sentenced cases even though the death sentences returned in them have been reversed and have not been re-instated. See ante at 194 n. 10, 613 A.2d at 1102 n. 10.

The majority regards cases preceded by asterisks as similar to Marshall’s and counts them as death-sentenced cases for the purpose of determining whether his sentence is disproportionate when compared to sentences handed down in similar cases.

1. Bey, Marko (Alston murder)

*2. Biegenwald, Richard (Olesiewicz murder) (first trial and second trial)

*3. Clausell, James

4. Coyle, Brian

5. Davis, Steven

*6. Di Frisco, Anthony

7. Dixon, Phillip

*8. Erazo, Samuel

9. Gerald, Walter

*10. Harvey, Nathaniel

*11. Hightower, Jacinto

*12. Hunt, James

*13. Jackson, Kevin

*14. Johnson, Walter (Sharp, A. murder)

15. Kise, Raymond

16. Koedatich, James (Hoffman murder)

17. Lodato, Benjamin

*18. Long, Ronald

*19. McDougald, Anthony (Bass, W. murder)

*20. McDougald, Anthony (Bass. M. murder)

21. Moore, Marie

*22. Moore, Samuel (Moore, M. murder)

*23. Moore, Samuel (Moore, K. murder)

*24. Oglesby, Walter

25. Pennington, Frank

26. Perry, Arthur

*290 27. Pitts, Darryl

28. Purnell, Braynard ■

29. Ramseur, Thomas

30. Rose, Teddy

*31. Savage, Roy (first trial)

*32. Williams, James

*33. Zola, James (first trial)

The majority also counts Marko Bey’s conviction for the Pensiton murder twice, both times as a death-sentenced case. I also count it as a death-sentenced case, but I count it only one time.

APPENDIX B

Analysis of Illustrative Reversed Death-Sentenced Cases Used as Part of the Majority’s Precedent-Seeking Analysis

(1) State v. Clausell

[See discussion supra at 142-143, 613 A.2d at 1075-1076]

(2) State v. Oglesby

The majority includes Oglesby in several of its comparison groups, including those with the same number of aggravating and mitigating factors and the index of outcomes test. Ante at 170-172, 613 A.2d at 1090-1091. The majority reports Oglesby as a death verdict without analyzing or commenting on whether this Court’s reversal of that death sentence undermines the confidence in the jury’s death verdict. That oversight is unconscionable, as the Court reversed because the jury instructions on the sole aggravating factor in the case, c(4)(c), were constitutionally deficient. State v. Oglesby, 122 N.J. 522, 532 , 585 A.2d 916 (1991). The instruction did not comply with this Court’s narrowing construction of that factor in Ramseur . Although the conviction was overturned due to an erroneous jury instruction on the burden of proof of the diminished-capacity defense, the Court found that error in the aggravating factor instruction provided an independent basis for the vacation of the sentence. *291 Clearly, an overbroad definition of the sole aggravating factor in the case raises substantial questions about the reliability of the jury’s decision to sentence Oglesby to death.

(3) State v. Zola

Zola’s case provides yet another example of the Court’s questionable treatment of reversed cases. Zola’s death sentence was reversed on grounds that affected the weight and treatment of the aggravating and mitigating factors, rendering its value as precedent for a death sentence highly suspect. The trial court failed to inform the jury that the aggravating factors must outweigh the mitigating factors beyond a reasonable doubt. State v. Zola, 112 N.J. 384, 390-91 , 548 A.2d 1022 . That error alone forces the conclusion that the deathworthiness judgment is unreliable. Yet the deathworthiness decision is also suspect because the trial court told the jury that the mitigating factors must be found unanimously. Id. at 433 , 548 A.2d 1022 . The combined effect of the two errors firmly establishes that the death verdict is untrustworthy. Zola pleaded guilty to murder and was sentenced to life after the reversal, further supporting the counting of his case as a life-sentenced case.

(4) State v. Erazo

Although Erazo’s conviction was overturned on errors in the guilt phase (a jury instruction shifted the burden of proof to defendant to prove passion/provocation, and the jury instruction did not comply with Gerald), error in the penalty phase makes Erazo’s death sentence especially untrustworthy. The Court found that the record was “devoid of any evidence” to support the submission of the c(4)(c) aggravating factor to the jury because there was no evidence that Erazo had intended to inflict pain beyond above the act of killing on his victim. State v. Erazo, 126 N.J. 112, 137-38 , 594 A.2d 232 (1991). The State *292 was prohibited from alleging that factor on retrial. The unwarranted inclusion of the c(4)(c) aggravating factor almost certainly affected the jury’s deliberations on the deathworthiness of defendant.

APPENDIX C

Factual Analysis of Cases Similar to Marshall’s in Which Prosecutors Did not Seek the Death Penalty or Juries Did not Impose It

(1) and (2) State v. William Engel and State v. Herbert Engel

The cases of William Engel and his brother, Herbert Engel, are strikingly similar to Marshall’s. The murder they committed, like the murder Marshall committed, was a contract-killing of a defenseless spouse. The majority recognizes that there is little to distinguish the culpability of the Engels and Marshall. Ante at 179, 613 A. 2d at 1095. After a penalty trial, a jury sentenced the Engels to life terms in prison.

(3) State v. Collins

Darrell Collins killed his wife and eighteen-month-old son to obtain the proceeds from of insurance policies on their lives, which had been taken out nine weeks before the murders. The jury convicted Collins of the murder of his wife and of the capital murder of his son, but it did not find the alleged c(4)(d) (murder for pecuniary value) aggravating factor to be present. Because the evidence was overwhelming that he killed for pecuniary gain, I regard the jury’s determination that there were no aggravating factors as a deathworthiness judgment rather than a death-eligibility judgment. The case is similar to Marshall’s in that there was a pecuniary motive (the trial court cited that motive in sentencing Collins) and it involved the murder of defenseless family members. Moreover, Collins’s was a multiple murder, and therefore easily could be seen as having enhanced culpability. See Final Report 81 & table 7 *293 (multiple victims is first on list of factors that increase a defendant’s culpability). The murder of the wife was violent and brutal as he slashed her neck and stabbed her neck and breasts; and the evidence showed that the victim had fought back during the attack, suggesting that she, unlike Maria Marshall, was aware of her impending death. Collins suffocated his infant son — a defenseless victim — by pressing his head into the mattress of the crib, causing hemorrhages on the child’s head.

(4) State v. Dreher

Although John Dreher’s murder conviction has been reversed, I discuss it because the majority does and, equally importantly, because the prosecution’s theory of the case was similar to the prosecution’s theory in Marshall’s case.

Dreher’s case was not prosecuted capitally. The majority assumes that that was because the prosecutor did not believe the c(4)(c) aggravating factor, a wantonly vile murder, was present in that there was insufficient evidence to show that Dreher had intended that his wife suffer pain beyond that incident to her death. Ante at 177, 613 A.2d at 1094. Yet Dreher’s case (as described by the prosecutor at trial) and Marshall’s are similar in terms of both the crime and the defendant. Both Marshall and Dreher were alleged to have killed their wives for pecuniary gain so that they could live in luxury with their mistresses. Both defendants were financially secure, mature adults without prior criminal records. Neither cooperated with authorities, and neither, consistent with his claims of innocence, expressed remorse for his crime. Both had indicated to their lovers that they wished their wives were out of their lives at a time some months before the actual murder, indicating a long period of premeditation and a motive to eliminate the spouse in order to be with another woman. Both could be understood as having involved aborted attempts to kill the victims, as Dreher sought a gun several months before the *294 murder and Marshall arranged an opportunity to have his wife killed at a restaurant parking lot a few nights before the murder. In neither case were extenuating circumstances present that would explain or justify the defendant’s actions.

One element enhancing Dreher’s culpability is that he enlisted the help of his mistress, Nance Seifrit, in the killing. The c(4)(d) and c(4)(e) aggravating factors reflect the notion that killing for profit is reprehensible and deserving of the death penalty. In part, those factors reflect the concept that hiring a killer is offensive because it involves another person in a criminal enterprise beyond the person who intends the victim to die. That notion is grounded in a concern that the plan is more likely to succeed if others are involved as partners in crime, as the partners will prod one another into action. That element was present in Marshall’s case, inasmuch as he had hired another person to kill his wife for him. The payment of money triggered the c(4)(d) aggravating factor. Similarly, Dreher was said to have involved another person in his plan to kill his wife. According to the prosecution, he asked Seifrit to bring him something sharp while he was in the process of strangling his wife; only after she gave him a pair of scissors, said the prosecution, did Dreher stab his wife in the throat.

In comparing Dreher’s case with Marshall’s, one point that stands out is that Dreher’s victim was terrified when she died, whereas Marshall’s victim seems not to have known what was about to happen to her. Before her death, Gail Dreher was said to have begged for her life as her husband dragged her down the stairs to the basement and beat her. If, as the State claims, she overheard her husband ask for “something sharp,” then she almost certainly was aware of her impending death. Marshall’s victim appears to have suffered much less by comparison. The evidence suggests that when Maria Marshall died she was dozing on the front seat of the car and probably was unaware of her assailant. Unlike Gail Dreher, who was in pain for a considerable period, Maria Marshall died immediately *295 from two bullet wounds. 123 N.J. at 32-33 , 586 A.2d 85 (noting that Marshall told police that his wife was asleep but that he had asked her to pop the trunk lever); id. at 34 , 586 A.2d 85 (noting that pathologist had testified that victim had been lying down at time of shooting).

(5) Walter Williams

The majority distinguishes Walter Williams’s case from Marshall’s because the jury did not find a pecuniary motive and thus rejected the c(4)(d) aggravating factor. Ante at 181, 613 A.2d at 1096. Although the majority treats that occurrence as a distinguishing fact, I would note, first, that the jury did not find that aggravating factor in Marshall’s case either, so that jury’s failure to find it in the Williams case hardly distinguishes it from Marshall’s. More importantly, however, I would point out that our underlying concern that motivates our entire undertaking is that disparate treatment of similar cases — by prosecutors or by juries — can lead to disproportionate sentencing. The fact that a jury treated a similar case differently does not prove that the case is dissimilar.

Williams is included in the comparison group of defendants selected as similar because the; case, like Marshall’s, involved a spousal murder of a defenseless victim and high levels of blameworthiness. State v. Robert Marshall: A Report to the New Jersey Supreme Court 27 (Marshall Report). Williams advanced to a penalty trial and the jury returned a life verdict. The Special Master concluded that Williams’s culpability was equal to Marshall’s. Marshall Report 28.

Williams’s crime is particularly relevant. Williams, a police officer with no prior record, poisoned his wife with cyanide to prevent his detection as a bigamist and to inherit her estate (he forged her will). The jury rejected the pecuniary-motive aggravating factor, but found that Williams had killed in order to escape detection for another crime. Like Marshall, Williams had a long-settled plan to murder the victim, obtaining the *296 cyanide at least five months before his wife’s death. Williams used his official position as a police officer to obtain the cyanide poison. Both cases involved a man killing his wife in order to be with another woman. Like Marshall, Williams had no prior record, was a mature, educated adult involved in his community, and had no extenuating circumstances to justify his crime. Unlike Marshall’s victim, Williams’s wife died a painful death after hours of suffering, reflecting a significant level of victimization.

APPENDIX D

Sentencing Frequencies for Various Types of Capital-Murder Cases

1. Factually Comparable Cases (Cases Sharing Salient Features)

Contract-Murderers (Principals)

The first comparison group of similar cases consists of contract-murder principals. The salient feature, hiring another person to kill a victim, is also a statutory aggravating factor, c(4)(e); thus this group is equivalent to a group of cases that match have the same statutory aggravating factor that Marshall’s had. This comparison group is small, consisting of only four cases, but it includes cases very similar to Marshall’s. The death sentencing frequencies for this comparison group are as follows:

Marshall Excluded: .00 (0/3)

Marshall Included: .25 (1/4)

Contract-Murderers (Agents)

Unlike Marshall, who was alleged to have participated in a contract-murder both out of greed and out of dissatisfaction with his marriage, these killers killed solely for pecuniary gain. *297 The death sentencing frequency for this comparison group is as follows:

Marshall Excluded: .00 (0/5)

All Contract-Murderers (Both Principals and Agents)

We never have distinguished between contract-murder principals and contract-murder agents in terms of death-worthiness. See State v. Di Frisco, 118 N.J. 253, 269 , 571 A.2d 914 (1990). Accordingly, I have created a comparison group that includes all death-eligible defendants who participated in contract murders. The death sentencing frequencies for this comparison group are as follows:

Marshall Excluded: .00 (0/8)

Marshall Included: .11 (1/9)

Spousal Murderers Who Did Not Kill in the Heat of Passion and Whose Victims Were Defenseless

This group consists of three cases. The death-sentencing frequencies for this comparison group are as follows:

Marshall Excluded: .00 (0/2)

Marshall Included: .33 (1/8)

Murderers Who Robbed or Kidnapped Their Victims, Planned Their Crimes Extensively, Had Pecuniary Motives, and Deceived or Entrapped Defenseless Victims

The Master created a group of five cases sharing several common facts: extensive premeditation to kill, a pecuniary motive, a significant measure of deception or entrapment, and a defenseless victim. These cases did not involve contract killings, unlike Marshall’s, but they did involve either robbery or *298 kidnapping. The death-sentencing frequency for this comparison group is as follows:

Marshall Excluded: .20 (1/5)

Murderers Who Killed For Pecuniary Gain Whose Crimes Did Not Involve Burglary, Robbery, or Kidnapping

Although not charged with aggravating factor c(4)(d), Marshall seems to fall within the group of murderers who had a pecuniary motive and who did not commit robbery or burglary. The death-sentencing frequencies for this comparison group are as follows:

Marshall Excluded: .00 (0/9)

Marshall Included: .10 (1/10)

Murderers Who Robbed Acquaintances, Relatives, or Friends

Like Marshall’s case, all of these murders were for pecuniary gain and involved the killing of a person who trusted the defendant. The death sentencing frequency for this group is as follows:

Marshall Excluded: .00 (0/11)

2. Numerical Preponderance of Aggravating and Mitigating Factors

The majority includes two variations on this general approach: cases with one aggravating factor and two mitigating *299 factors and cases with one aggravating factor. I would tend to agree with the majority that such a comparison group has extremely limited applicability given our repeated admonition that jurors are supposed to make qualitative rather than quantitative judgments concerning the aggravating and mitigating factors. Ante at 171, 613 A.2d at 1091. Nevertheless, I do use a comparison group consisting of cases involving one aggravating factor and two mitigating factors, because it tends to confirm the results suggested by the other comparisons. Moreover, although the jury found only one aggravating factor, specifically, the contract murder factor, I also create a comparison group of cases involving two aggravating factors and two mitigating factors because the killing of Maria Marshall appears to have been both a contract-murder and a murder for pecuniary gain. Because a comparison group that fails to consider mitigating factors is virtually useless in conducting proportionality review, I do not undertake a comparison involving all cases with one aggravating factor.

Cases With One Aggravating Factor and Two Mitigating Factors

Marshall’s jury found one aggravating factor, c(4)(e), and two mitigating factors, c(5)(f) and c(5)(h). There are forty-three other cases that had one aggravating factor and two mitigating factors. The death sentencing frequencies for this group are as follows:

Marshall Excluded: .00 (0/43)

Marshall Included: .02 (1/44)

Cases With Two Aggravating Factors and Two Mitigating Factors

Although many of the cases in this group could be regarded fairly as more aggravated than Marshall’s, I consider the group *300 because, as noted, the evidence suggests that Marshall killed for pecuniary gain, implicating aggravating factor c(4)(d), and that he hired another to kill, implicating aggravating factor c(4)(e). Marshall doubtless would object that the cases in this group are too aggravated to be regarded as similar to his, but even this comparison group yields a result suggesting disproportionality in Marshall’s sentence. The death-sentencing frequency for this group is as follows:

Marshall Excluded: .06 (2/34)

3. Cases with a Similar Likelihood of Receiving a Death Sentence (Index of Outcomes Measure)

The majority also applies the Master’s index-of-outcomes method of constituting comparison groups. The index method recognizes that there may not be enough cases on all fours with the case under review to conduct a reliable proportionality review. To compensate, the method gathers a group of cases that are similar in terms of culpability but not necessarily similar in terms of their factual contexts. The Master’s index is based on empirical evidence of the types of factual characteristics that best explain sentencing outcomes. Using a statistical technique called multiple regression analysis, each factor is weighted according to its tendency to induce juries to sentence defendants to death. Each case is given a “score” based on the factors present. Cases then are arranged in order of their scores, and cases with similar scores (“near neighbors”) are chosen to form a comparison group. The index method is designed to generate a group of cases that may be quite disparate factually but quite similar in their overall levels of culpability. According to the Master, with the index method “it is possible to determine the relative culpability of different defendants based on the case characteristics that, on a statisti *301 cal basis, best explain which defendants actually received death sentences.” Final Report, supra, at 87.

Because I have reservations concerning the index method for creating comparison groups, I would tend to emphasize the more straightforward methods for constructing comparison groups that involve the identification of salient factors. As noted, the Master’s index is based on a multiple regression that identifies what factors tend to persuade juries to impose the death penalty and what factors do not. For that sort of multiple regression to work with any accuracy, a substantial number of juries must have returned death verdicts. If there are only three or four death verdicts, conducting the multiple regression analysis may be pointless. The Master addressed that problem by mislabeling over thirty life-sentenced cases as death-sentenced cases. See Final Report, supra, Appendix B. In doing so, he created enough data to make an analysis feasible, but he also created an analysis that may not identify accurately the factors that sway juries in the system that uses the rules the Court has laid down. Stated simply, too few death verdicts issued by properly instructed juries may exist for anyone to know what factors tend to induce such juries to return death verdicts.

Because the majority seems to rely in part on the index method of forming comparison groups, I review the results those groups produce notwithstanding my concerns regarding the methodology used to create them. I include two sets of results, one using only statutorily enumerated aggravating and mitigating factors, the other using factors that have little obvious relationship to the factors enumerated in the statute.

Index Method Based Only on Statutorily Derived Factors

Using only statutorily enumerated factors to evaluate the likelihood of a death sentence, Marshall falls into culpability level “2” (with level “1” being the least aggravated and level “5” being the most aggravated cases). Twenty-one other cases *302 fell into level “2”. Marshall Report 37. The death-sentencing frequencies for this comparison group are as follows:

Marshall Excluded: .05 (1/20)

Marshall Included: .10 (2/21)

Index Method Based on Statutorily Derived Factors and Other Factors

Using all death-eligible cases in the proposed universe and using the index based on both statutorily enumerated factors and other factors, Marshall falls into culpability level “1” (to repeat, the lowest of the Master’s five levels), a ranking he shares with 177 cases. The list is too large to include in this opinion, but in the Appendix I do include a list of Marshall’s “near neighbors.” The death sentencing frequencies for this comparison group are as follows:

Marshall Excluded: .10 (1/10)

Marshall Included: .18 (2/11)

4. Weisberg’s Model

As an alternative to the models created the Master, the State offers a method fashioned by its own expert, Herbert Weisberg. Even comparison groups formulated under Weisberg’s criteria reveal that the death penalty rarely is imposed in cases like Marshall’s.

Weisberg’s approach is similar to approaches based on the counting of statutory aggravating factors. Using what amounts to a form of “educated intuition,” Weisberg divided all death-eligible cases into four “culpability categories,” with Category 1 including the defendants he thought most culpable. *303 According to Weisberg, Marshall fell into Category 1. Category 1 includes cases with the following configurations of statutory aggravating factors: c(4)(a); c(4)(a) and (c); c(4)(a), (c), and (g); c(4)(a) and (g); c(4)(b) and (d); c(4)(b), (f), and (h); c(4) (c), (f), and (g); c(4)(d); c(4)(d) and (f); c(4)(e); and c(4)(f) and (h). Weisberg Report, supra, at 15-16. Weisberg does not list death-eligible cases that did not result in a penalty trial, but even using the universe he uses, thereby creating a bias in favor of a finding of proportionate sentencing, Marshall’s sentence still appears to be disproportionate. The death sentencing frequencies for this comparison group are as follows (counting only cases that resulted in penalty trials):

Marshall Included: .12 (3/26)

Marshall Excluded: .08 (2/25)

APPENDIX E

The following two graphs indicate the death-sentencing frequencies for each comparison group. Death-sentenced cases are shaded in black. Life-sentenced cases that once were death sentenced cases are shaded with diagonal lines. Life-sentenced cases that never were death-sentenced cases are shaded with dots. The graphs make plain the fact that, even if cases in which death sentences were returned and subsequently vacated are counted as death-sentenced cases, Marshall’s sentence still is disproportionate.

The first graph expresses death-sentencing rates in percentages, the second in number of cases. One can see by looking at the graphs that mere percentages occasionally can mislead. For example, in one comparison group the death-sentencing rate reaches thirty-three percent if Marshall’s case is included in the group; one must look at the graph that expresses death-sentencing rates in numbers of cases to understand that Marshall’s case is aberrational even in that comparison group, for *304 in that group the total number of cases is three, Marshall’s being the one that resulted in a death sentence. (Hence, when Marshall’s case is excluded from the group, one hundred percent of the cases are reported as life-sentenced cases.)

*305 Sentencing Outcomes for Similar Cases Expressed In Number of Cases

[[Image here]]

Codes

Composition of Comparison Groups

Code Type of Cases

A Contract-murderers (principals)

B Contract-murderers (agents)

C All contract-murderers

D Spousal murderers who did not kill in the heat of passion and whose victims were defenseless

E Murderers who robbed or kidnapped their victims, planned their crimes extensively, had pecuniary motives, and deceived or entrapped defenseless victims

F Murderers who killed for pecuniary gain whose crimes did not involve burglary, robbery or kidnapping

G Murderers who robbed acquaintances, relatives, or friends

1 Marshall excluded

2 Marshall included

*306 Sentencing Outcomes for Similar Cases Expressed in Percentages

[[Image here]]

Codes

Composition of Comparison Groups

Code Type of Cases

A Contract-murderers (principals)

B Contract-murderers (agents)

C All contract-murderers

D Spousal murderers who did not kill in the heat of passion and whose victims were defenseless

E Murderers who robbed or kidnapped their victims, planned their crimes extensively, had pecuniary motives, and deceived or entrapped defenseless victims

F Murderers who killed for pecuniary gain whose crimes did not Involve burglary, robbery or kidnapping

G Murderers who robbed acquaintances, relatives, or friends

i-< Cases with one aggravating factor and two mitigating factors

I Cases with two aggravating factors and two mitigating factors

J Index method based only on statutorily derived factors

K Index method based on statutorily derived and other factors

L Weisberg’s Model

1 Marshall excluded

2 Marshall included

*307 APPENDIX F

Lists of Comparison Cases

In this appendix I list the names of the defendants whose cases are placed within each comparison group.

I have coded the cases with the following abbreviations:

PT: case went to a penalty trial;

noPT: case decided without penalty trial;

L: life-sentence defendant;

D: death-sentenced defendant; and

[ ']: a life-sentenced defendant treated by the majority as a death-sentenced defendant.

1. Factually Comparable Cases (Cases Sharing Salient Features)

Contract-Murderers (Principals)

Life Sentence

1. Engel, William (PT/L)

2. Engel, Herbert (PT/L)

3. Brand, Francis (noPT/L)

Death Sentence

4. Marshall, Robert (PT/D)

Contract-Murderers (Agents)

Life Sentence _ Death Sentence

1. Melendez, Miquel (PT/L)

2. Rose, Michael (PT/L)

3. Burroughs, Randy (noPT/L)

4. [DiFrisco, Anthony (PT/D/Rev’d)]

5. [Clausell, James (PT/D/rev’d/PT/L)]

All Contract-Murderers (Both Principals and Agents)

Life Sentence

1. Engel, William (PT/L)

2. Engel, Herbert (PT/L)

3. Brand, Francis (noPT/L)

4. Melendez, Miquel (PT/L)

5. Burroughs, Randy (noPT/L)

6. [Rose, Teddy (PT/D/rev’d/PT/L) ]

7. [DiFrisco, Anthony (PT/D/Rev’d)]

8. [Clausell, James (PT/D/rev’d/PT/L)]

Death Sentence

9. Marshall, Robert (PT/D)

*308 Spousal Murderers Who Did Not Kill in the Heat of Passion and Whose Victims Were Defenseless'

Life Sentence

1. Collins, Darrell (PT/L)

2. Williams, Walter (PT/L)

Death Sentence

3. Marshall, Robert (PT/D)

Murderers Who Robbed or Kidnapped Their Victims, Planned Their Crimes Extensively, Had Pecuniary Motives, and Deceived or Entrapped Defenseless Victims

Life Sentence

1. Scales, Terrence (PT/L)

2. Russo, David (PT/L)

3. Mclver, Vernon (noPT/L)

4. Thompson, Howard (noPT/L)

Death Sentence

5. Martini, John (PT/D)

Murderers Who Killed For Pecuniary Gain Whose Crimes Did Not Involve Burglary, or Kidnapping

Life Sentence

1. Rose, Michael (PT/L)

2. Burroughs, Randy (noPT/L)

3. Melendez, Miguel (PT/L)

4. Brand, Francis (noPT/L)

5. Engel, Herbert (PT/L)

6. Engel, William (PT/L)

7. Williams, Walter (PT/L)

8. [Clausell, James (PT/D/rev’d/PT/L) ]

9. [DiFrisco, Anthony (PT/D/rev’d) ]

Sentence

10. Marshall, Robert (PT/D)

Murderers Who Robbed Acquaintances, Relatives, or Friends

Life Sentence Death Sentence

1. Ferrari, Salvatore (noPT/L)

2. Jefferson, Richard (noPT/L)

3. Lazorisak, George (PT/L)

4. Mclver, Vernon (noPT/L)

5. James, Marvin (noPT/L)

6. Grant, Michael (noPT/L)

7. Sullivan, Roy (noPT/L)

8. Clark, Reginald (noPT/L)

9. Graf, Clifford (noPT/L)

10. Johnson, Nathaniel (PT/L)

11. Neapolitano, Anthony (PT/L)

2. Numerical Preponderance of Aggravating and Mitigating Factors

Cases With One Aggravating Factor and Two Mitigating Factors

*309 Life Sentence

1. Jones, Tracy (noPT/L)

2. Sullivan, Roy (noPT/L)

3. Dinkins, Robert (noPT/L)

4. Worthington, Earl (noPT/L)

5. Messam, Gladstone (noPT/L)

6. Stevens, Larry (noPT/L)

7. Floyd, Lamont (noPT/L)

8. LaPointe, Pierre (noPT/L)

9. Delvalle, Efrain (noPT/L)

10. Williams, Gerald (noPT/L)

11. Calloway, Derrick (noPT/L)

12. Williams, Walter (PT/L)

13. Anderson, Antoine (noPT/L)

14. Preston, Johnie (noPT/L)

15. Graf, Clifford (noPT/L)

16. Perry, Harold (PT/L)

17. Jones, Jimmie (PT/L)

18. Stamps, Aaron (PT/L)

19. Dollard, Thomas (noPT/L)

20. Norman, Anthony (noPT/L)

21. McNeil, Keith (noPT/L)

22. Turner, John (noPT/L)

23. Clark, Hashona (noPT/L)

24. Montalvo, Orlando (noPT/L)

25. Armstrong, Joseph (noPT/L)

26. Ruano, Heriberto (noPT/L)

27. Keresty, Walter (noPT/L)

28. Scales, Terrence (PT/L)

29. Hudson, Franklin (noPT/L)

30. Slaughter, Rafael (PT/L)

31. Kershaw, Albert (noPT/L)

32. Jacoby-Irwin, Barbara (noPT/L)

33. Deeves, William (PT/L)

34. Muscio, Nicholas (PT/L)

35. Edwards, Eugene (noPT/L)

36. Melendez, Miguel (PT/L)

37. Mendez, Oscar (noPT/L)

38. Biegenwald, Richard (Ward murder) (PT/L)

39. Guagenti, Joseph (PT/L)

40. Dreher, John (noPT/L)

41. Reyes, Jose (PT/L)

42. [Oglesby, Walter (PT/D/rev’d)]

43. [Long, Ronald (PT/D/rev’d)]

Death Sentence

44. Marshall, Robert (PT/D)

*310 Cases with Two Aggravating Factors and Two Mitigating Factors

Life Sentence

1. Brown, Vincent (noPT/L)

2. Holmes, Gregory (noPT/L)

3. Culley, Carl (noPT/L) 4. Melendez, Angel (noPT/L)

5. Taylor, Wiley (noPT/L)

6. Huff, Aaron (PT/L)

7. Cancio, Gustavio (PT/N)

8. Morton, Adrian (noPT/L)

9. Thomas, Christopher (noPT/L)

10. Tucker, Stanley (noPT/L)

11. Merola, Thomas (noPT/L)

12. McCollum, William (noPT/L)

13. Wheeler, Ronald (noPT/L)

14. Jones, Larry (PT/L)

15. Mujahid, Rasheed (noPT/L)

16. Washington, Corey (noPT/L)

17. Rivera, Rafael (PT/L)

18. Watkins, Ricky (noPT/L)

19. Mendez, Incenzio (PT/L)

20. Jalil, Nelson (noPT/L)

21. Barone, Jamie (PT/L)

22. Nieves, Alberto (PT/L)

23. Spraggins, Jerry (PT/L)

24. Reese, John (PT/L)

25. [Zola, James (PT/D/Rev’d/noPT/L) ]

26. [Purnell, Braynard (PT/D/rev’d)]

27. [Dixon, Phillip (PT/D/rev’d)]

28. [Biegenwald, Richard (Olesiewicz murder) (PT/D/rev’d/PT/D/rev’d) ].

29. [Lodato, Benjamin (PT/D/rev’d)]

30. [Rose, Teddy (PT/D/rev’d)]

31. [Jackson, Kevin (PT/D/rev’d)]

32. [Ramseur, Thomas (PT/D/rev’d)]

Death Sentence

33. Martini, John (PT/D)

34. Bey, Marko (Peniston murder) (PT/D/ rev’d/PT/D)

3. Cases with a Similar Likelihood of Receiving a Death Sentence (Index of Outcomes Measure)

Index Method Based Only on Statutorily Derived Factors

Life Sentence

1. Weston, Elisha (PT/L)

Death Sentence

20. Marshall, Robert

*311 Francis 2. Brand, (noPT/L)

3. Carrol, John (PT/L)

4. Melendez, Miguel (PT/L)

5. Mincey, Samuel (noPT/L)

6. Nieves, Alberto (PT/L)

7. Reyes, Jose (PT/L)

8. Spraggins, Jerry (PT/L)

9. Taylor, Leroy (noPT/L)

10. Johnson, Walter (Sharp, B. murder) (PT/L)

11. [Johnson, Walter (Sharp, A. murder) (PT/D/rev’d)]

12. [Hightower, Jacinto (PT/D/rev’d)]

13. [McDougald, Anthony (Bass, W. murder) (PT/D/rev’d)

14. [McDougald, Anthony (Bass. M. murder) (PT/D/rev’d)

15. 'Savage, Roy (PT/D/Rev’d/PT/L) ]

16. [Moore, Samuel (Moore, M. murder) (PT/D/rev’d)]

17. 'Moore, Samuel (Moore, K. murder) (PT/D/rev’d)]

18. [Biegenwald, Richárd (Olesiewicz murder) (PT/D/rev’d/PT/D/Rev’d) ]

19. [Zola, James (PT/D/rev’d/noPT/L)]

21. Bey, Marko (Peniston murder) (PT/D/ rev’d/PT/D)

Index Method Based on Statutorily Derived Factors and Other Factors

Life Sentence

1. Brand, Francis (noPT/L)

Engel, William (PT/L) 2.

Muhammed, Jihad (noPT/L) 3.

Neapolitano, Anthony (PT/L) 4.

Mendez, Oscar (noPT/L) 5.

Biegenwald, Richard (Ward murder) (PT/L) 6.

Reese, John (PT/L) 7.

Guagenti, Joseph (PT/L) 8.

[Zola, James (PT/D/rev’d/noPT/L)] 9.

Death Sentence

10. Martini, John (PT/D)

11. Marshall, Robert (PT/D)

4. Weisberg’s Model

Life Sentence

1. Miguel Melendez, (PT/L)

2. Biegenwald, Richard (Ward murder) (PT/L)

3. Engel, Herbert (PT/L)

4. Engel, William (PT/L)

Death Sentence

24. Schiavo, Dominick (PT/D)

25. Marshall, Robert (PT/D)

*312 5. Manfredonia, Michael (PT/L)

6. Weston, Elisha (PT/L)

7. Ploppert, Charles (PT/L)

8. Booker, George (1st Victim) (PT/L)

9. Johnson, Walter (Sharp, B. murder) (PT/L)

10. [Johnson, Walter (Sharp, A. murder) (PT/D/rev’d)]

11. [Pennington, Frank (PT/D/rev’d)]

12. [DiFrisco, Anthony (PT/D/rev’d)]

13. [Coyle, Bryan (PT/D/rev’d)]

14. [Biegenwald, Richard (Oslewicz murder) (PT/D/rev’d/PT/D/rev’d) ]

15. [Rose, Teddy (PT/D/rev’d/PT/L)]

16. Thomas

17. Ulausell, James (PT/D/rev’d/PT/L)]

18. Erazo, Samuel (PT/D/rev’d)]

19. Hightower, Jacinto (PT/D/rev’d)]

20. McDougald, Anthony (Bass, W. murder) (PT/D/ rev’d) ]

21. McDougald, Anthony (Bass, M. murder) (PT/D/rev’d) ]

22. "Kise, Raymond (PT/D/rev’d/PT/L)]

23. TKoedatich, James (PT/D/rev’d/PT/L)]

26. Bey, Marko (Peniston murder) (PT/D/rev’d/PT/D)

As noted by the majority, the legislature amended the death-penalty act shortly after oral argument in this case to provide that proportionality review "be limited to a comparison of similar cases in which a sentence of death has been imposed." L.1992 c. 5 (eff. May 12, 1992). The State urges that the amendment be applied retroactively to Marshall’s case. I believe that the amendment should not be applied to any death-sentenced case, let alone a case, like Marshall’s, involving a crime committed prior to enactment of the amendment.

On its face, the amendment is unconstitutional. As noted, I believe that our constitution requires proportionate death sentencing and, to enforce that guarantee, it requires that this Court conduct comparative proportionality review. Were we to employ a death-only universe, we would completely undermine that constitutional guarantee, for proportionality review so conducted is an empty exercise. Under Ramseur , this Court cannot abide by the legislative amendment.

In any event, the amendment plainly cannot apply to Marshall, or to any other defendant who committed his or her crime prior to the date of the amendment’s enactment. Article I of the United States Constitution provides that neither Congress nor any State shall pass any “ex post facto law.” Article IV, para. 7 of the New Jersey Constitution imposes a similar prohibition. The Supreme Court has interpreted that provision as barring criminal laws that do all of the following: (1) they apply to events occurring before their enactment; (2) they disadvantage the offender affected by them; and (3) they change substantial personal rights and do not merely implicate the procedures used to establish a conviction or to fix a punishment. Miller v. Florida, 482 U.S. 423, 430 , 107 S.Ct. 2446, 2451 , 96 L.Ed.2d 351, 360 (1987). Those elements plainly would be met if the Court applied the amendment to a capital conviction for a crime committed before its enactment. Any defendant sentenced under the revised statute would be disadvantaged, for he or she would lose the chance to have the sentence set aside if found disproportionate. To be sure, the fact that a sentence is reviewed under the old statute does not mean that the sentence automatically will be vacated. Marshall’s case certainly demonstrates that *253 point. Nevertheless, the fact that a defendant cannot show "definitively” that he or she would have reaped the benefits of the old statute is not. dispositive. Id. at 432 , 107 S.Ct. at 2452, 96 L.Ed.2d at 361 . Defendants cannot be deprived of the "opportunity” to have their sentences reversed. See Weaver v. Graham, 450 U.S. 24, 34 , 101 S.Ct. 960, 967 , 67 L.Ed.2d 17, 26 (1981). Even if our state constitution would permit the doing away with proportionality review, either by subverting it as the amendment would or by abandoning it entirely, the ex post facto law clauses of the federal and state constitutions would guarantee the right of every defendant who committed capital-murder prior to the enactment of the amendment to request a full scale review of the proportionality of his or her sentence. Cf. In re Coruzzi, 95 N.J. 557, 577-79 , 472 A.2d 546 (1984).

I should point out that the Special Master raises an important point that could justify a single exception to that rule. With respect to sentences vacated as disproportionate pursuant to proportionality review, the Special Master argues that they should be included in the universe as death-sentenced cases. If they are not, he observes, then it will be impossible for changed community values to result in changed sentencing practices. Final Report, supra, at 64. That issue could arise, for example, if the first five cases of a certain type result in life sentences, but the next twenty result in death. The initial death-sentenced cases might be reversed as disproportionate. The twentieth death- *258 sentenced case also would be reversed as disproportionate — despite the fact that eighty percent of juries gave verdicts of death in that type of case — unless the reversed cases were treated as death-sentenced cases notwithstanding the fact that they actually resulted in life sentences by virtue of proportionality review.

Whether a court employs a frequency analysis or a precedent-seeking analysis, it must incorporate some notion of what makes cases similar or dissimilar to one another. I do not believe I differ very much from the majority on that issue. Like the majority, I reject the notion that a fair proportionality review should be or can be constructed by looking only at the aggravating and mitigating factors enumerated in the capital-murder statute. Under that statute, the jury in a sentencing trial is supposed to identify aggravating and mitigating factors and then, after those factors have been identified, it is supposed to weigh them. I believe that there are many facts that are not themselves enumerated as aggravating or mitigating factors that endow the enumerated factors with their respective weights. For example, in cases involving the c(4)(b) aggravating factor — that is, cases in which defendant put persons other than the victim at grave risk — that factor would have more weight if the defendant killed someone by setting fire to a crowded arena than if the defendant killed someone by shooting a single bullet into a small *262 group of people gathered on a street corner. Thus, when deciding whether cases are similar to the case under review, I would consider the statutory factors as the central determinant of similarity and other identifiable circumstances that can be thought of as clustered to or constituent parts of the enumerated statutory factors themselves.

The Court may be concerned about situations in which the bulk of the cases in the comparison gr

This text is long and has been trimmed here. Open the source document for the complete record.