State v. Josephs

COLEMAN, J.,

concurring in part and dissenting in part.

I concur with the majority that the sequential charge was so prejudicial to defendant that it had the effect of depriving him of a fair trial on the “by own conduct” death-eligibility issue. The Court, however, requires a new trial. I disagree that a new capital murder trial should be conducted because I conclude that the evidence was insufficient as a matter of law to. support the jury’s determination on the “by own conduct” issue even if the jury had been properly instructed. Hence, I would preclude a second trial on death eligibility.

I.

Our death penalty statute makes defendant death eligible only if he purposely or knowingly caused the death of one or more of the victims “by his own conduct.” N.J.S.A. 2C:ll-3c. That means defendant must have “actively and directly participated in the homicidal act, ie., in the infliction of the injuries from which the victim died. The critical elements are that defendant in fact acted, and the immediacy of his conduct to the victim’s demise.” State v. Gerald, 113 N.J. 40, 97 , 549 A.2d 792 (1988). The “by own conduct” requirement is a triggering event for determining whether a defendant is death eligible, State v. Moore, 113 N.J. 239, 300-01 , 550 A.2d 117 (1988), which the State must prove beyond a *143 reasonable doubt. Id. at 299 , 550 A.2d 117 . Consequently, an examination of the evidence to determine whether it was sufficient to establish the “by own conduct” requirement is necessary.

The “by own conduct” issue was of great concern to the trial court because the State relied exclusively on circumstantial evidence to prove death eligibility. None of the physical evidence recovered or tested forensically specifically connected defendant directly to the shootings of McLean, Mitchell or Williams. Additionally, none of the testimonial evidence connected defendant directly to the shootings. As the trial court summarized: “Except for the testimony of Emil Josephs that he heard shots from two different caliber guns and testimony from neighbors that they heard thumping and rat-a-tat sounds, the evidence as to what occurred to precipitate the shootings is very sparse and almost nonexistent.” In rejecting defendant’s motion for a judgment of acquittal based on a claim that the evidence was insufficient to sustain a verdict of murder by defendant’s own conduct, the trial court stated that the “fact that Hugh was seen with only one gun and the gunshot wounds sustained by the victims were caused by two or three guns sufficiently established the inference that a second gunman helped Hugh shoot the victims.”

Defendant contends that the testimony of Emil Josephs that Hugh fired a nine millimeter handgun- at Emil as Emil leaped from the kitchen window was the only direct evidence of who was in possession of a handgun used in the murders. He argues that the jury’s verdict is based on speculation because the only direct evidence in the case concerning his whereabouts prior to the murders came from Emil who testified that defendant was observed sleeping in his bedroom hours before the murder.

In addressing the “by own conduct” issue, the majority has framed the question presented as “whether, after viewing the evidence in the light most favorable to the prosecution, any rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt.” Post, at 81, 803 A.2d at 1095 (citing Jackson v. Virginia, 443 U.S. 307, 319 , 99 S.Ct. 2781, 2789 , *144 61 L.Ed.2d 560, 573 (1979)). I submit that in this death penalty case, the question should be whether proof of “by own conduct” should'be allowed to rest exclusively on circumstantial evidence where inferences are not based on independently proven facts. My answer is no. I believe that to allow proof of an essential precondition to death eligibility based exclusively on circumstantial evidence “allow[s] impermissibly discretionary findings and death sentences based on the slimmest of evidence.” State v. Ramseur, 106 N.J. 123, 208 , 524 A.2d 188 (1987), cert. denied, 508 U.S. 947 , 113 S.Ct. 2433 , 124 L.Ed.2d 653 (1993).

II.

A.

The death penalty as a form of punishment is categorically different. Furman v. Georgia, 408 U.S. 238, 286 , 92 S.Ct. 2726, 2751 , 33 L.Ed.2d 346, 376 (1972) (Brennan, J., concurring). “ ‘[T]he imposition of death by public authority is ... profoundly different from all other penalties.’ ” State v. Ramseur, supra, 106 N.J. at 326 , 524 A.2d 188 (quoting Lockett v. Ohio, 438 U.S. 586, 605 , 98 S.Ct. 2954, 2965 , 57 L.Ed.2d 973, 990 (1978)).

“From the point of view of the defendant, it is different in both its severity and its finality. From the point of view of society, the action of the sovereign in taking the life of one of its citizens also differs dramatically from any other legitimate state action. It is of vital importance to the defendant and to the community that any decision to impose the death sentence be, and appear to be, based on reason rather than caprice or emotion.”

[Mickens v. Taylor, 535 U.S. -, -, 122 S.Ct. 1237, 1253 , 152 L.Ed.2d 291 (2002) (Stevens, J., dissenting) (quoting Gardner v. Florida, 430 U.S. 349, 357-58 , 97 S.Ct. 1197, 1204 , 51 L.Ed.2d 393 (1977)).]

For that reason, our federal constitution requires extra protections for defendants who stand “in deadly peril of their lives.” Powell v. Alabama, 287 U.S. 45, 71 , 53 S.Ct. 55, 65 , 77 L.Ed. 158, 171 (1932). To that end, a “heightened reliability [is] demanded by the Eighth Amendment in the determination whether the death penalty is appropriate in a particular case.” Sumner v. Shuman, 483 U.S. 66, 72 , 107 S.Ct. 2716, 2720 , 97 L.Ed.2d 56, 63 (1987). *145 That Amendment “draw[s] its meaning from the evolving standards of decency that mark the progress of a maturing society.” Atkins v. Virginia, — U.S. -, -, 122 S.Ct. 2242, 2247 , 153 L.Ed.2d 335 , - (2002) (quoting Trap v. Dulles, 356 U.S. 86, 100-01 , 78 S.Ct. 590, 598 , 2 L.Ed.2d 630, 642 (1958)).

The notion that a sentence of death is different has caused some scholars and commentators to disagree with the majority’s application of the typical non-capital “beyond a reasonable doubt” standard to capital cases, rather than the more stringent “beyond any doubt” standard. See, e.g., Sherry F. Colb, The Qualitative Dimension of Fourth Amendment “Reasonableness”, 98 Colum. L.Rev. 1642, 1674 (1998) (stating that “[t]he Supreme Court ... has rested an entire jurisprudence of capital punishment on the premise that state killing is sui generis and that noncapital precedents sometimes provide insufficient protection when applied in the capital context. The Court has declared by way of justification that ‘death is different.’ ”); Stephen P. Garvey, Death-Innocence and the Law of Habeas Carpus, 56 Alb. L. Rev. 225 , 233 (1992) (stating that “[djeath is different, and doctrinally its distinctness is expressed by this attention to moral detail. Nothing of moral relevance must be disregarded or overlooked.”); Note, The Rhetoric of Difference and the Legitimacy of Capital Punishment, 114 Harv. L.Rev. 1599, 1603 (2001) (distinguishing between “legal guilt (the state of having been tried with due process of law and found guilty) and actual guilt (the state of having actually committed some crime)” and noting that the Supreme Court “has made ... procedural adjustments to attempt to ensure that a legally guilty capital defendant is actually guilty of a capital crime.”); Laura S. Underkuffler, Agentic and Conscientic Decisions in Law: Death and Other Cases, 74 Notre Dame L.Rev. 1713, 1729 (1999) (citing Supreme Court cases that describe uniqueness of death penalty and stating that “[a]s a result of this uniqueness, it is particularly critical that the correct decision in each case be made, and correctness requires that all factors, circumstances, and aspects of the case be heard and weighed, without hindrance, by the sentencer.”).

*146 In addition, the Model Penal Code bars death where “although the evidence suffices to sustain the verdict, it does not foreclose all doubt respecting the defendant’s guilt.” Margery Malkin Koosed, Averting Mistaken Executions By Adopting The Model Penal Code’s Exclusion Of Death In The Presence Of Lingering Doubt, 21 N. Ill. U.L.Rev. 41, 50-51 (2001) (quoting Model Penal Code § 210.6(l)(f)). The commentary notes that “[t]his provision is an accommodation to the' irrevocability of the capital sanction. Where doubt of guilt remains, the opportunity to reverse a conviction on the basis of new evidence must be preserved, and a sentence of death is obviously inconsistent with that goal.” Ibid. “It may not be a ‘reasonable doubt’ but it is a real doubt nonetheless, in the sense that those who possess it ‘can be expected to resist those who would impose the irremedial penalty of death.’ ” Id. at 55 (quoting Smith v. Balkcom, 660 F.2d 573, 581 (5th Cir.1981), cert. denied, 459 U.S. 882 , 103 S.Ct. 181 , 74 L.Ed.2d 148 (1982)). Indeed, prior to “the mid-1700’s, jurors may have been allowed to acquit on the basis of any doubt,” and the reasonable doubt standard “was likely a concession to prosecutors.” Id. at 72.

Professor Koosed observes that concession has been costly:

A report prepared by James Liebman and a team of lawyers and criminologists at Columbia University, A Broken System: Error Rates in Capital Cases, 1973-1995, at 6-7 (June 12, 2000), reveals:

The “overall error rate” for the entire 1973-1995 period, i.e. the proportion of fully reviewed capital judgments overturned at one of the three stages (direct appeal, post-conviction, and/or federal habeas corpus) due to serious error, was 68%. On retrial, when the errors are cured, “an astonishing 82% (247 out of 301) of the capital judgments that were reversed were replaced on retrial with a sentence less than death, or no sentence at all. In the latter regard, 7% (22/301) of the reversals for serious error resulted in a determination on retrial that the defendant was not guilty of the capital offense.”

[Koosed, supra, 21 N. Ill. U.L.Rev. at 108 (quoting Kamisar, LaFave & Israel Modem Criminal Procedure 80-81 (9th ed, 2000 Supp.) (quoting the study)).]

That study was updated in February 2002 and recommended that capital-cases should require a higher standard of proof — beyond any doubt — rather than proof beyond a reasonable doubt. James S. Liebman, et al., A Broken System, Part II: Why There Is So *147 Much Error In Capital Cases, And What Can Be Done About It, at 397-99 (Feb. 11, 2002), available at http://justice.poliey.net/ ejreform/dpstudy/ (last visited June 6, 2002).

More than fifty years ago, this Court adopted the “beyond any doubt” standard for criminal eases in which the State relied substantially on circumstantial evidence to prove a defendant’s guilt. In State v. Donohue, 2 N.J. 381, 390-91, 67 A.2d 152 (1949), this Court held that, to justify a murder conviction based largely on circumstantial evidence,

all of the circumstances not only must concur to indicate a defendant’s guilt but they must also be inconsistent with any other rational conclusion. It is not enough that they coincide to render probable the hypothesis advanced by the prosecution; they must also exclude beyond a reasonable doubt every other hypothesis except that of guilt. Jackson v. Delaware, L. & W.R. Co., 111 N.J.L. 487 , 170 A. 22 (E. & A.1933). Where the essential facts are proved, and where they cannot be rationally explained on any theory other than that the defendant is guilty of the crime charged, such circumstantial evidence will be considered as convincing as evidence of a direct and positive character.

mm

That formulation was later rejected in favor of what has become known as the State v. Reyes “beyond a reasonable doubt” standard. State v. Mayberry, 52 N.J. 413, 436-37 , 245 A.2d 481 (1968) (stating that the Donohue standard has been repudiated because it would have defeated too many legitimate circumstantial prosecutions based on speculative hypothetical consistent with the defendant’s innocence). I believe the Court should adopt the Donohue “beyond any doubt” standard for “by own conduct” death-eligibility determinations in cases in which the State relies exclusively, or nearly exclusively, on circumstantial evidence. By restricting that standard to such death-eligibility determinations, the problem associated with defendant’s advancing speculative hypotheticals consistent with innocence will not be presented because the purpose of such a standard is to separate principals from accomplices rather than the guilty from the innocent.

B.

The majority has used the principles articulated in State v. Reyes, 50 N.J. 454 , 236 A.2d 385 (1967), that establish the stan *148 dard in criminal cases for determining whether to submit a case to the jury for its ultimate determination. Although this Court in State v. Gerald concluded that the relevant inquiry concerning the “by own conduct” issue is whether or not the defendant actively and directly participated in the homicidal act, ie., in the infliction of the injuries from which the victim died, the Court in Gerald did not indicate whether Reyes was the appropriate standard to be used in making that determination. The Court has, however, addressed the sufficiency of the evidence to prove an aggravating factor. Those cases are helpful in deciding whether to submit a “by own conduct” issue to the jury. For instance, to submit to the jury the aggravating factor of whether the defendant committed the murder to escape detention, N.J.S.A 2C:11-Bc(4)(f), “the State [must] produce[ ] sufficient evidence on which a jury can reasonably conclude that at least one of the motives of the defendant in killing his or her victim was to eliminate a witness or avoid subsequent apprehension and prosecution for criminal acts.” State v. Martini, 131 N.J. 176, 282-83 , 619 A.2d 1208 (1993), cert. denied, 516 U.S. 875 , 116 S.Ct. 203 , 133 L.Ed.2d 137 (1995).

In State v. Hunt, 115 N.J. 330, 388 , 558 A.2d 1259 (1989), the defendant contended that the State failed to present sufficient evidence that the murder of the victim involved either “torture” or an aggravated battery, which is required to establish an aggravating factor under c(4)(c). The evidence in Hunt established that the victim was stabbed twenty-four times, was shocked by the attack and bled for twenty minutes before dying.- Id. at 389 , 558 A.2d 1259 . The State relied solely on the twenty-four stab wounds to establish aggravating factor c(4)(c). Notably, the Court expressed concern that “if the c(4)(c) factor could be sustained on this evidence alone, there would be ‘no principled way to distinguish this case, in which the death penalty was imposed, from the many cases in which it was not.’” Ibid, (quoting Godfrey v. Georgia, 446 U.S. 420, 433 , 100 S.Ct. 1759, 1767 , 64 L.Ed.2d 398, 409 (1980)). Because the State was barred on other grounds from seeking the death penalty, however, the Court found that it did *149 not need to determine whether proof of the aggravating factor was insufficient to support submission of that factor to the jury. Ibid.

Also helpful to the analysis is a rule of law established nearly half-a-eentury ago in State v. Lucas, 30 N.J. 37, 51 , 152 A.2d 50 (1959), where the Court held that a defendant could not be convicted of felony murder (a capital offense at that time) based solely on defendant’s uncorroborated confession. The Court required “the extrinsic corroborative proofs ... be of such a nature as to give the confession an aura of authenticity.” Id. at 52, 56 , 152 A.2d 50 . That evidentiary ruling was established to minimize the chances that innocent people would be executed. Id. at 56-57 , 152 A.2d 50 . A similar rational standard of reliability should be required for a “by own conduct” determination. Indeed, this Court in State v. DiFrisco, 118 N.J. 253, 280 , 571 A.2d 914 (1990), cert. denied, 516 U.S. 1129 , 116 S.Ct. 949 , 133 L.Ed.2d 873 (1996), applied the Lucas principle and vacated defendant’s death sentence following a guilty plea because there was no corroboration of defendant’s confession. DiFrisco observed that Lucas “is a rule of evidence” that applies to both the guilt and sentencing phases of the State’s case. Id. at 270, 571 A.2d 914 .

In DiFrisco , the issue was whether there was sufficient evidence to corroborate the defendant’s confession that Franeiotti had paid the defendant to kill Potcher, the owner of Jack’s Pizzeria. The Court stated that the

pointed question is whether there was, within the circumstances of the case, corroboration of Franeiotti’s role as one who paid the hired gun to kill in order to avoid detection. Surely these surrounding circumstances bespeak the crime. There is no known connection between DiFrisco and the pizzeria owner. Why else would he have done it? Is the fact that there is no other explanation enough corroboration?

[Ibid.]

Despite those compelling circumstances, the Court nonetheless concluded that the absence of corroborating evidence of the confession meant that the confession was not reliable enough to allow a sentence of death to stand, notwithstanding the fact that the defendant’s confession had been found to be admissible as evi *150 dence. Id. at 273 , 571 A.2d 914 . Here, as in DiFrisco , reliability is the missing link in the case absent some direct or independently reliable evidence of “by own conduct.”

Defendant also was convicted of conspiracy to commit murder, a non-capital offense. A Lucas-type reliability standard has been applied in conspiracy cases where the Court has held that there must be independent proof of the conspiracy before a co-conspirator’s statements can be used against the co-conspirator. State v. Phelps, 96 N.J. 500, 519 , 476 A.2d 1199 (1984). It is indeed ironic that in the conspiracy charge against defendant in this case the judge was obligated to determine whether there was independent evidence “substantial enough to engender a belief in the conspiracy’s existence and the defendant’s participation,” ibid., yet in the capital murder charges against defendant the judge was not required to consider the reliability of the circumstantial evidence presented to establish death eligibility. Id. at 509-10 , 476 A.2d 1199 .

Of the more than fifty cases on death row in New Jersey, this is the first case in which the “by own conduct” issue seriously has been called into question. In all of our capital cases in which multiple persons have been accused of causing the death of a victim, there always was some direct evidence to satisfy the “by own conduct” and/or direct participation in the events that led to the victim’s death. Some cases in which we have reversed a judgment of death demonstrate this Court’s requirement that a sentence of death be grounded in highly reliable evidence.

The need for reliability begins at the earliest stages of a death penalty prosecution. In State v. McCrary, 97 N.J. 132, 143 , 478 A.2d 339 (1984), the Court stated that giving notice of aggravating factors, which is the first step in a capital case, is like seeking an indictment. That is, it “is the notion that the allegations derive from some verifiable source. The need to ensure that such a source exists compels some preliminary review to satisfy the interest of the public and the defendant that such charges not proceed to trial without a factual mooring.” Ibid.

*151 Numerous cases in which multiple defendants have been charged with inflicting fatal injuries have required direct evidence of “by own conduct.” For example, State v. Simon, 161 N.J. 416, 439-41 , 737 A.2d 1 (1999), involved a case in which two individuals were charged as the trigger person in the murder of a police officer “by own conduct.” In that case, the Court noted that although there were two defendants charged with the capital murder of a single victim, there was a threshold showing made before the Grand Jury demonstrating how the defendant, Simon, directly participated in causing the death of the victim. Ibid. The Court pointed out that the Grand Jury considered evidence that Simon was the passenger and that the shot causing death was fired from the passenger side of the vehicle. Id. at 440 , 737 A.2d 1 . In addition, Simon exited the passenger side of the vehicle with the murder weapon in his hand within minutes after the murder. Ibid. In view of that evidence, the Court concluded there was a sufficient prima fade showing that Simon was the one who murdered the victim by his own conduct, although he was one of two defendants who had been charged with a murder that involved only one shooter. Ibid.

Similarly, in State v. Clausell, 121 N.J. 298, 309 , 580 A.2d 221 (1990), two defendants were indicted for capital murder. The petit jury found both Clausell and his co-defendant guilty of purposeful or knowing murder and found that Clausell, but not his co-defendant, had committed the homicidal act by his own conduct. Id. at 312 , 580 A.2d 221 . Because there was direct evidence of Clausell's direct participation in the homicidal act, this Court neither questioned nor criticized the fact that two defendants in the single-shooter murder were charged with capital murder for killing the same individual. This case is different from Simon and Clausell ; in fact, it is the first time that there is no direct evidence of “by own conduct” when multiple defendants have been charged with inflicting the fatal injuries.

After this Court’s decision in State v. Gerald, supra, 113 N.J. at 89, 549 A.2d 792 , which held that the New Jersey Constitution *152 prohibited the death penalty for serious bodily injury murders, and until the Constitution was amended effective December 3, 1992, this Court set aside several death penalties based on insufficient evidence. In State v. Gerald, supra, 113 N.J. at 89-92, 549 A.2d 792 , the Court concluded that despite the nature and multiplicity of injuries inflicted upon the victim, the jury could not infer intent to cause death rather than serious bodily injury. In State v. Moore, supra, 113 N.J. at 302, 550 A.2d 117 , the Court concluded that the facts in the case did not support the jury’s finding that the defendant committed the homicidal act by her own conduct. In that case, several persons were accused of committing the homicidal act and the record did not satisfy the requirement that defendant had delivered one of the fatal blows. Ibid. Similarly, in State v. Matulewicz, 115 N.J. 191, 194, 201 , 557 A.2d 1001 (1989), the Court held that.the c(4)(c) aggravating factor evidence was insufficient to show that defendant intended to cause physical pain in addition to death. The Court was unwilling to infer that factor from the circumstances of the crime. Ibid. So too, in State v. Perry, 124 N.J. 128, 173-76 , 590 A.2d 624 (1991), the Court reversed a death sentence because the evidence was insufficient to establish the c(4)(c) aggravating factor. Likewise, in State v. Rose, 112 N.J. 454, 531-33 , 548 A.2d 1058 (1988), the Court found insufficient evidence to submit the depravity aggravating factor to the jury. The Court observed that a jury cannot infer depravity, which focuses on the defendant’s state of mind, solely from the means used to. cause the death. Ibid.

Clearly, our death penalty jurisprudence has stressed “the importance of providing a jury with every opportunity to spare a defendant’s life.” State v. Cooper, 151 N.J. 326, 362 , 700 A.2d 306 (1997), cert. denied, 528 U.S. 1084 , 120 S.Ct. 809 , 145 L.Ed.2d 681 (2000). One such illustration is permitting a jury to return a non-unanimous verdict on “by own conduct” death eligibility. State v. Brown, 138 N.J. 481, 509-14 , 651 A.2d 19 (1994). Consistent with that view, our Court has stated “death should not be imposed as a result of what may be an extremely close determination.” State v. *153 Matulewicz, supra, 115 N.J. at 200, 557 A.2d 1001 (quoting State v. Ramseur, supra, 106 N.J. at 206 , 524 A.2d 188 ).

In State v. Hightower, 146 N.J. 239, 282-83 , 680 A.2d 649 (1996) (Handler, J., concurring in part and dissenting in part), Justice Handler noted that the record revealed little credible evidence supporting the argument that, in killing the victim, the defendant was motivated by a fear of identification and thus committed the murder to avoid apprehension, an aggravating factor under c(4)(f). Justice Handler contended that “[a]t a minimum, in order to demonstrate sufficient evidence, the State must proffer factual evidence that is independently significant to each factor.” Id. at 281 , 680 A.2d 649 . Additionally, he found the State’s reliance on indirect circumstantial evidence troubling since such evidence could be “shaped and directed ... to support different motives for murder.” Ibid.

C.

Other states that have addressed the issue of the standard required to prove “by own conduct” in death penalty cases also have made the distinction between circumstantial and direct evidence. In circumstantial evidence cases, some courts require the state to prove the defendant’s guilt not only beyond a reasonable doubt, but to the exclusion of every reasonable hypothesis consistent with innocence. See Davis v. State, 314 Ark. 257 , 863 S.W.2d 259, 264 (1993), cert. denied, 511 U.S. 1026 , 114 S.Ct. 1417 , 128 L.Ed.2d 88 (1994); Darling v. State, 808 So.2d 145, 155 (Fla.2002); Bullard v. State, 263 Ga. 682 , 436 S.E.2d 647, 650-51 (1993); People v. Jones, 105 Ill.2d 342, 86 Ill.Dec. 453 , 475 N.E.2d 832, 835 (1985), cert. denied, 489 U.S. 1040 , 109 S.Ct. 1174 , 103 L.Ed.2d 236 (1989); State v. Neal, 796 So.2d 649, 657 (La.2001), cert. denied, — U.S. -, 122 S.Ct. 1323 , 152 L.Ed.2d 231 (2002); Brewer v. State, 725 So.2d 106, 133 (Miss.1998), cert. denied, 526 U.S. 1027 , 119 S.Ct. 1270 , 143 L.Ed.2d 365 (1999).

Of special significance is Cummings v. State, 715 So.2d 944, 949 (Fla.1998), where the court concluded that the evidence was *154 insufficient to support defendant’s conviction for premeditated murder and reversed the judgment and death sentence. The evidence showed that a fight ensued between Karlon Johnson and one of the co-defendants, Andre Fisher. Id. at 946 . A witness testified that defendant Cummings was upset about this and asked for a ride to get his gun. Ibid. Another witness testified that, some time after the fight, he saw the defendant in a car with an “Uzhtype gun” on his lap and the defendant asked about the fight and inquired as to the whereabouts of Johnson. Ibid. Later that evening, a white Honda Accord drove by the house where Johnson’s sister and her children lived and Johnson frequently stayed. Ibid. Johnson’s car was parked in the carport of his sister’s house. Ibid. Shelton Lucas, Sr., Johnson’s brother-in-law, who is the same height and weight as Johnson and wore clothes similar to those worn by Johnson earlier that evening, was standing in the carport when he noticed the car driving down the street. Ibid. Lucas, Sr. testified that he had taken about ten steps through the kitchen and into the living room before the passenger in the car fired at least thirty-five shots at the house from three different nine millimeter guns. Id. at 946, 949 . Several bullets struck Johnson’s car, penetrated the kitchen door, and one of the bullets struck and killed five-year-old Shelton Lucas, Jr. Id. at 946 . A witness testified that the white Honda Accord passed him shortly before the shooting and there were four people in the car, including Fisher, and that the owner of the car was Marion King. Ibid. Lastly, the witness testified that the following day the defendant told him that, if the police ask him, he should say that they were together the previous evening at his cousin’s house. Ibid.

The bullet that struck the child was consistent with a bullet fired from a gun similar to one of three nine millimeter guns fired from the ear into the house, but the State’s expert was unable to link it conclusively with the defendant. Id. at 946-47 . In reversing the conviction, the Florida court concluded that it could not rule out the possibility that the defendant and his cohorts merely intended to frighten Johnson or to damage his car and there was *155 insufficient proof of premeditation to convict on first-degree murder and impose the death sentence. Id. at 949 .

The reasoning in that Florida case is applicable here. Although defendant may have had a motive to kill Emil for ordering defendant out of the apartment and fighting with Junior, or even McLean for disarming defendant, there is no direct evidence that defendant murdered the victims by his own conduct. Although premeditation is the essential element which distinguishes first-degree murder from second-degree murder in Florida, see Hoefert v. State, 617 So.2d 1046, 1048 (Fla.1993), in New Jersey, the “by own conduct” standard is one of three triggering devices to find a defendant guilty under the death penalty statute, N.J.S.A. 2C:11-3c. See State v. Gerald, supra, 113 N.J. at 99, 549 A.2d 792 . Here, the State argues that three different victims were killed with three different guns. Although the ballistics expert testified the bullets were fired from three different guns, he also stated that some of the bullets were fired from a forty-five caliber gun, some from a nine millimeter gun, and the sole thirty-eight caliber bullet recovered could have been fired from a nine millimeter gun. He further testified that the discharged shells fell into only two groups, indicating that they were fired by two guns. Thus, it is reasonable to infer that one nine millimeter weapon and a forty-five caliber gun were used as the murder weapons. Therefore, there is no direct evidence that defendant used or handled the murder weapons or that Junior did not use two different guns to murder the victims. Emil’s testimony that he heard shots coming from two different guns also is consistent with that theory. Therefore, the possibility that there were only two guns, both used by Junior to murder the victims, cannot be ruled out. Additionally, the State’s argument that multiple shooters were involved due to the distinct locations of the victims is unpersuasive since two of the victims’ bodies were found in close proximity to the kitchen where Junior attempted to shoot Emil.

In Hall v. State, 403 So.2d 1319, 1320 (Fla.1981), the evidence established that the two defendants approached the victim, who *156 was carrying a shotgun. The victim was later found dead, and the defendants were observed fleeing the scene and subsequently exchanged gunfire with a sheriff. Ibid. The weapon used to murder the victim was found under the victim’s body, while the victim’s shotgun was found inside the car in which the defendants fled. Ibid. The court determined that those facts

demonstrate[d] beyond a reasonable doubt that the two [defendants] engaged in a common criminal scheme. As such, each was a principal to the death, and the fact that the state did not prove which [defendant fired on the victim did] not necessitate either’s acquittal. By actively operating together each was guilty of the acts of the other.

[Ibid.]

Although it was clear that one of the two defendants shot the victim, the court concluded that the evidence was insufficient to prove premeditation beyond a reasonable doubt because the proof of intent was subject to conflicting interpretations. Id. at 1320-21 . The circumstantial evidence was “inconsistent with any reasonable hypothesis of innocence as to the [murder], but it [was] not inconsistent with any reasonable exculpatory hypothesis as to the existence of premeditation.” Id. at 1321 . The conviction for first-degree murder, therefore, was reversed. Ibid.

Similarly, the circumstantial evidence in this ease, at best, establishes that defendant was an accomplice in the murders. Since accomplices are not subject to the death penalty under the “by your own conduct” requirement, State v. Simon, supra, 161 N.J. at 441 , 737 A.2d 1 , it follows that defendant should not be subjected to the death penalty even if the circumstantial evidence establishes that he was engaged in a common criminal scheme. The evidence is insufficient to prove that defendant committed a murder by his own conduct, since the evidence implicating defendant in the murder of the victims is subject to conflicting interpretation as previously stated. Although the circumstantial evidence may be inconsistent with any reasonable hypothesis of innocence as to the murder of the decedents, it is not inconsistent with any reasonable exculpatory hypothesis as to the “by own conduct” requirement.

*157 The Mississippi Supreme Court has recognized that “ ‘[f]acts not directly proved but which jurors legitimately may infer from other facts which have been proved, are limited to such as naturally follow as being logically connected or related, and which reasonably derive from such established facts.’ ” Hester v. State, 463 So.2d 1087, 1093 (Miss.1985) (quoting Matula v. State, 220 So.2d 833, 836 (Miss.1969)). Thus, “a guilty verdict based upon circumstantial evidence must be supported by a much higher degree of proof.” Ibid. The court stated:

“It is fundamental that convictions of crime cannot be sustained on proof which amounts to no more than a possibility or even when it amounts to a probability, but it must rise to the height which will exclude every reasonable doubt; that when in any essential respect the state relies on circumstantial evidence, it must be such as to exclude every other reasonable hypothesis than that the contention of the state is true.”

[Ibid, (quoting Westbrook v. State, 202 Miss. 426 , 32 So.2d 251, 252 (1947)).]

In Hester, supra, the court addressed the prosecution’s hypothesis that after the defendant and the victim left a bar, they drove to a trailer park where they struggled and the defendant ultimately robbed and stabbed the victim to death. Id. at 1088 . The prosecution’s hypothesis relied on a witness’s observation of the defendant and the victim leaving the bar together; two hours later the defendant returned disheveled and alone; a folding knife seized from the defendant’s trailer with small traces of blood and no fingerprints; and a pathologist’s statement that the victim’s wounds were consistent with the type of knife. Id. at 1089-90 .

The court found that the circumstances in the case supported the prosecution’s contention that the defendant killed the victim with the knife, but the evidence did not exclude the reasonable hypothesis asserted by the defendant that, outside the bar, the defendant and the victim parted ways, the victim was killed by a third party, and the defendant’s condition upon returning to the bar was due to an unrelated incident. Id. at 1093 . The court concluded, therefore, that the “web of circumstances” established by the State did not exclude the reasonable hypothesis that a third party, rather than the defendant, murdered the victim. Id. at 1094 . Furthermore, the evidence connecting the defendant to the *158 murder weapon was weak, since the prosecution failed to connect the blood on the knife to the victim’s blood or show that the knife belonged to or was used by the defendant. Id. at 1092 . The court reversed the defendant’s conviction because the evidence only presented a probability that the defendant murdered the victim. Id. at 1094 . Likewise, in this case, the “web of circumstances” established by the State fails to exclude the reasonable hypothesis that defendant was involved in the murder of the victims, presenting only at most a probability that defendant caused the injuries that resulted in the victims’ deaths.

In White v. State, 532 So.2d 1207, 1221 (Miss.1988), the court found that although the evidence was more than sufficient to establish beyond a reasonable doubt that the defendant and the co-defendants robbed the victim, that the victim was killed during the course of the robbery, and that the defendant was present when the killing occurred, the evidence concerning the events before and after the robbery offered no indication as to which defendant killed the victim. While the felony-murder rule only required the defendant to have been an active participant in the robbery when the victim was killed, the court stated that the Mississippi legislature had directed “that greater culpability be required before the death penalty is imposed,” presumably more than accomplice liability. Id. at 1221 n. 2. Further, there was no evidence that the defendant contemplated lethal force and no direct evidence to place the defendant at the scene with the co-defendants at the time the victim was killed. Id. at 1221 . Although the defendant was seen leaving the scene with a gun in his hand, there was no proof linking the gun to the crime or the defendant prior to the shooting, and the murder weapon was not introduced as evidence at the trial. Ibid. The court also found the evidence was “wholly consistent with numerous other scenarios by which [the defendant] did not kill or anticipate a killing,” but in fact, other reasonable hypotheses suggested that either of the co-defendants could have been the murderer. Ibid. Therefore, the court found that there was a “pure absence of any proof on the elements contained in the sentencing statute” because *159 nothing in the record offered the jury a rational basis for selecting one hypothesis over the other. Ibid.

In this case, merely circumstantial evidence places defendant in the apartment with Emil at the time of the victims’ deaths. Although Delores claims that she saw either defendant or Junior place something into his pocket, there is no other evidence that that occurred or if it did, there is no evidence that the object was a gun. There also is no direct evidence that defendant handled or fired the murder weapons, because no murder weapon was received as evidence at the trial. Moreover, the evidence is consistent with a number of other scenarios in which the defendant did not kill or participate in the killing. As previously stated, at least one other reasonable hypothesis suggests that Junior alone was the murderer. Nothing in this record offered the jury a rational basis for selecting one hypothesis over the other and, therefore, the “by own conduct” requirement under the death penalty statute has not been met.

Conversely, in State v. Anthony, 776 So.2d 376, 386 (La.), cert. denied, 531 U.S. 934 , 121 S.Ct. 320 , 148 L.Ed.2d 258 (2000), the court found that the state was not required to show that the defendant actually pulled the trigger. Instead, to successfully carry its burden, the state was required to prove only that the “defendant acted in concert -with his co-perpetrators, that defendant had the specific intent to kill, and that one of the [aggravating factors] was present.” Ibid. Although the defendant contended that the evidence was insufficient to prove that he was the triggerman, the court found that the circumstantial evidence was sufficient. Ibid. Specifically, the defendant stood in the doorway brandishing a gun with a potato on the barrel immediately before the victims were shot; the defendant ordered the victims to their knees and was the last perpetrator seen by the sole survivor, just before the latter was shot; the defendant’s shoes were the most heavily encrusted with potato particles of all the co-defendants; and the defendant was repeatedly present at the residence where the weapon later was found on the day of the murder. Ibid.

*160 That case is distinguishable from the facts in this case because the State failed to prove that defendant acted in concert with Junior. In fact, the direct evidence establishes only that Junior brandished a gun and attempted to kill Emil. Moreover, the medical evidence clearly revealed that all three victims were shot by a nine millimeter gun that was last seen in Emil’s possession. Finally, and most importantly, accomplice liability is insufficient to impose a sentence of death in New Jersey.

III.

I believe that a heightened standard of reliability should be used to decide the “by own conduct” issue in death penalty eases and that the Reyes standard that the Court uses today — the same used to determine whether an individual should be found guilty or innocent of possession of a marijuana cigarette — should not be used in determining whether an accused can be executed. Despite the fact that this Court has reversed thirty-seven of fifty-two death penalties imposed since 1982 because of trial error, some of which related to the “by own conduct” death-eligibility determination, the Court today has nonetheless declined to adopt a standard that will improve the reliability of death-eligibility determinations. The “by own conduct” death-eligibility determination in this case is based exclusively on weak circumstantial evidence that consists of inferences not based on independently proven facts. Thus, the quality of the evidence undermines confidence that the death-eligibility determination is reliable, thereby leading me to the conclusion that the sentence of death also is unreliable. I am unwilling to tolerate “discretionary findings and death sentences based on the slimmest of evidence.” State v. Ramseur, supra, 106 N.J. at 208 , 524 A.2d 188 . “Circumstantial evidence cases are always better candidates for penalty leniency than direct evidence convictions.” King v. Strickland, 748 F.2d 1462, 1464 (11th Cir. 1984), cert. denied, 471 U.S. 1016 , 105 S.Ct. 2020 , 85 L.Ed.2d 301 (1985).

*161 That lack of confidence in the machinery of death is what has prompted governors across this nation to declare a moratorium on death penalty prosecutions and executions, including most recently the decisions announced by the governors of Illinois and Maryland. See also Sharone Levy, Righting Illinois’ Wrongs: Suggestions For Reform And A Call For Abolition, 34 J. Marshall L.Rev. 469 (2001) (discussing the cases of thirteen exonerated men in Illinois where “although the prosecution had mostly circumstantial evidence and unreliable witnesses, it still was able to convince a judge and jury to convict those men of murder and sentence them to death”). Apart from that growing lack of confidence is the reality that recently the 101st person condemned to die was found to be innocent and was released from death row in Pennsylvania. Pennsylvania Man is 101st Exonerated Death Row Inmate, (May 6, 2002), available at http://www.deathpenaltyinfor.org/101Exonerated.html.

The growing national concerns about the reliability of the machinery of death has not escaped New Jersey. A May 2002 survey of New Jersey residents conducted by the Eagleton Institute of Politics of Rutgers University regarding residents’ views on the death penalty indicates that sixty-six percent of the residents support a temporary halt to executions. Eagleton Institute of Polities, Center for Public Interest Polling, New Jerseyans’ Opinions on A Death Penalty Moratorium, at 4 (May 2002). One of the reasons for such a halt is the potential for convicting and executing innocent people. In fact, the survey indicated that ninety-six percent of New Jersey residents believe that innocent people are sometimes convicted of murder. Id. at 5. ‘While we do not regard public opinion polls as decisive [on] issues of constitutional law, we cannot ignore their relevance to the largely empirical determination of’ whether the public has confidence in our death penalty system. State v. Ramseur, supra, 106 N.J. at 174 n. 10, 524 A.2d 188 . In the interest of reducing the chances that defendants improperly would be found to be death eligible, I would adopt the proof “beyond any doubt” standard articulated in State v. Donohue for determining “by own conduct” death eligibili *162 ty. Because that higher standard was not satisfied in this case, the State should be precluded from retrying defendant for capital murder. I believe the “ ‘law triumphs when the natural impulses aroused by a shocking crime yield to the safeguards which our civilization has evolved for an administration of criminal justice at once rational and effective.’ ” State v. Purnell, 126 N.J. 518, 546 , 601 A.2d 175 (1992) (quoting Watts v. Indiana, 338 U.S. 49, 55 , 69 S.Ct. 1347, 1350 , 93 L.Ed. 1801, 1807 (1949)).

Chief Justice PORITZ and Justice LONG join in this opinion.