I dissent from parts II, IV, X, XII, XV, and XVII.A of the lead opinion. I dissent in part from part VI of the lead opinion. I concur in the result reached in part VIII of the lead opinion but dissent from its rationale. The first three parts of this opinion address issues arising from the penalty phase of Young’s trial. The next two parts address issues arising from the guilt phase of the trial. The ensuing two parts discuss jury selection issues. The final part analyzes the constitutionality of Utah’s statutory scheme for narrowing the class of defendants eligible for the death penalty.
I. SHACKLING OF DEFENDANT DURING PENALTY PHASE
(lead opinion part XII)
Young argues that the trial court violated his rights under the Eighth and Fourteenth Amendments when it required him to remain in shackles in the presence of the jury during the penalty phase. I concur with the lead opinion that “it is within the sound discretion of the trial court to determine the safety measures necessary to insure the security of the courtroom and its occupants. These safety measures may include shackling a defendant in appropriate circumstances.” (Emphasis added.) The problem in this case is the manifest lack of the requisite necessity. Young’s unwarranted shackling amounted to an impermissible comment on the evidence and violated his due process rights by creating unacceptable prejudice.
Because of the inherently prejudicial impact of appearing shackled before the jury, courtroom shackling is permitted only “as a last resort.” See Illinois v. Allen, 397 U.S. 337, 344, 90 S.Ct. 1057, 1061, 25 L.Ed.2d 353 (1970). In all the cases I have examined, reviewing courts have required a showing of necessity before tolerating a trial court’s decision to shackle. See, e.g., Spain v. Rushen, 883 F.2d 712, 728 (9th Cir.1989), cert. denied, 495 U.S. 910, 110 S.Ct. 1937, 109 L.Ed.2d 300 (1990); Elledge v. Dugger, 823 F.2d 1439, 1452 (11th Cir.1987), cert. denied, 485 U.S. 1014, 108 S.Ct. 1487, 99 L.Ed.2d 715 (1988); Tyars v. Finner, 709 F.2d 1274, 1284-85 (9th Cir.1983); People v. Duran, 16 Cal.3d 282, 290, 127 Cal.Rptr. 618, 623, 545 P.2d 1322, 1327 (1976); Bello v. State, 547 So.2d 914, 918 (Fla.1989).1 Thus, I agree with the Ninth Circuit that “a trial judge may ... impose restraints only when ‘confronted with disruptive, contumacious, [and] stubbornly defiant defendants.’ ... Shackling ... must be limited to cases urgently demanding that action.” Tyars, 709 F.2d at 1284 (quoting Allen, 397 U.S. at 343, 90 S.Ct. at 1060-61).
Furthermore, before a court may shackle a disruptive defendant, it must first “pursue less restrictive alternatives.” Spain, 883 F.2d at 721; see also Tyars, 709 F.2d at 1284. Lesser restraints could include increasing courtroom security personnel, *369warning the defendant of the consequences of disruptive behavior, such as the possibility of contempt or removing the defendant from the courtroom. See Spain, 883 F.2d at 726. This last alternative, however, also is an abuse of discretion if it is not limited to cases urgently demanding such action. Notably, even removal is inappropriate unless the defendant first “has been warned by the judge that he will be removed if he continues his disruptive behavior” and the defendant thereafter persists in “disorderly, disruptive, and disrespectful” conduct. Allen, 397 U.S. at 343, 90 S.Ct. at 1060-61. Young received no warning, presumably because he had not engaged in any disruptive behavior.
It is true, as the lead opinion indicates, that most of the cases discussed above arose from shackling during the trial of a defendant’s guilt, where the chief concern was the potential prejudice to the defendant’s presumption of innocence. Nevertheless, the United States Supreme Court described shackling’s potential prejudice in words showing a concern with more than just its effect on the presumption of innocence:
[T]he sight of shackles ... might have a significant effect on the jury’s feelings about the defendant, [and] the use of this technique is something of an affront to the very dignity and decorum of judicial proceedings that the judge is seeking to uphold. Moreover, ... [the defendant’s] ability to communicate with counsel is greatly reduced.
Allen, 397 U.S. at 344, 90 S.Ct. at 1061 (emphasis added). Accordingly, the Eleventh Circuit concluded in Elledge v. Dugger that “the Supreme Court has not bottomed the prohibition against shackling on presumption of innocence alone.... [T]here seems to be no reason to restrict the principles to the guilt-innocence phase of the trial.” 823 F.2d at 1451. In fact, the prejudice from shackling may be greater at the sentencing phase; a capital sentencing jury is assessing the defendant’s dangerousness and risk of future violence and “might view the shackles as first hand evidence of future dangerousness and uncontrollable behavior which if unmanageable in the courtroom may also be unmanageable in prison, leaving death as a proper decision.” Id. at 1450. Shackling conveys a message about the court’s view of the defendant’s dangerousness, a message that potentially prejudices whatever issue the jury is considering, whether guilt or sentence. I therefore believe that shackling at the penalty phase, like shackling at the guilt phase, must be justified by a showing of necessity. See Bello v. State, 547 So.2d 914, 918 (Fla.1989).
This case lacks any of the “disorderly, disruptive, and disrespectful” behavior necessary for shackling or removal from the courtroom.2 The record shows that Young had at all times comported himself appropriately and did nothing during the trial to justify either shackling or removal. On the contrary, unsettling testimony caused him to cry quietly or to leave the courtroom voluntarily. On the morning of the third day, the judge asked Young whether he wanted to stay for that day’s testimony, indicating that it would include “things that may be distressing and upsetting to you.” Young responded, “I’m going to stay. If I feel uncomfortable, I’ll leave. How is that?” He had already excused himself from the courtroom for part of the first day and for the entire second day because he was uncomfortable listening to some of the testimony about his mental capacities. Yet, in the face of Young’s demonstrated ability and intention to comport himself properly and to excuse himself from the courtroom when needed, the court required the use of a waist chain and handcuffs as a condition of Young’s presence during much of the sentencing hearing.
The trial court’s shackling decision was not motivated by any actual or threatened misconduct, but solely by the court’s own assessment of Young’s character and a *370laudable but misplaced desire to minimize risk. Unfortunately, the very absence of courtroom misconduct that would justify shackles might have caused the jury to speculate improperly about why shackles nevertheless were suddenly in use. Additionally, shackling impermissibly conveyed to the jury the trial court’s own view of Young’s dangerousness. It amounted to an inappropriate assessment by the court of the credibility and weight to be given to testimony by an Indiana police officer on the second day of sentencing about Young’s demeanor during a past incident. The officer testified that Young once got angry when a television crew attempted to film him in court and that it took four officers to restrain him. Given that Young had been in the presence of the jury for part of the first day of sentencing without shackles, the jury could have interpreted the appearance of shackles on the third day as an expression of the court’s uncritical acceptance of the officer’s testimony.
In these circumstances, shackling was an abuse of discretion that introduced intolerable prejudice into the sentencing phase, thereby violating Young’s Fourteenth and Eighth Amendment rights to carefully channelled and reliable capital sentencing. See Mills v. Maryland, 486 U.S. 367, 383-84, 108 S.Ct. 1860, 1869-70, 100 L.Ed.2d 384 (1988); McCleskey v. Kemp, 481 U.S. 279, 303-04, 107 S.Ct. 1756, 1773-74, 95 L.Ed.2d 262 (1987); Gardner v. Florida, 430 U.S. 349, 358-59, 97 S.Ct. 1197, 1204-05, 51 L.Ed.2d 393 (1977); Woodson v. North Carolina, 428 U.S. 280, 305, 96 S.Ct. 2978, 2991-92, 49 L.Ed.2d 944 (1976). It also forced Young to compromise his Sixth Amendment right to be present at the sentencing phase, for as the trial court acknowledged, Young based his decision to leave the courtroom for the third day of sentencing at least partially, if not wholly, on the shackling order. I cannot conclude that this was constitutional harmless error, particularly given the inherent difficulty, discussed more fully in the section on Young’s statutory right to allocution (part II.A), of reweighing on appeal the subjective moral judgment made by a capital sentencing jury.3
II. DEFENDANT’S REQUEST FOR ALLOCUTION DURING THE PENALTY PHASE
(lead opinion part XV)
During his penalty hearing, defendant specifically asked the court to give him an opportunity to address the jurors before their deliberations. The lead opinion concludes that even if the trial court erred in denying this request, the error was harmless beyond a reasonable doubt. I cannot agree, and I am dismayed by the lead opinion’s cavalier dismissal of the potential impact of defendant’s own statement on a capital sentencing jury. I conclude not only that Young had both a statutory right and a constitutional right to allocution, but also that the court’s denial of these rights was harmful error under any standard. Furthermore, the trial court’s failure to allow Young to speak briefly to the jury likely compounded the prejudice created by the shackling order. The order conveyed to the jury the court’s assessment of Young’s character and dangerousness; the denial of allocution prevented Young from presenting his personal view of his circumstances and character.
A. Statutory Error
Historically, the common law right of allocution provided capital defendants— who were incompetent to testify and not entitled to counsel—an opportunity to plead for their lives, and until the defendant had a chance to speak, the court could not properly pronounce sentence. See Ball v. United States, 140 U.S. 118, 129-31, 11 S.Ct. 761, 765-66, 35 L.Ed. 377 (1891). Perhaps because of the fairness of this practice, the common law eventually extended the right to all criminal defendants. Today, despite modern innovations in criminal procedure that allow defendants to testify and to have the assistance of counsel, the ancient common law right of allocution exists as a right by statute or rule in most *371states, including Utah, as well as in the federal system. See Utah R.Crim.P. 22(a); Fed.R.Crim.P. 32(a)(1)(C).
Notably, the high courts of many states with statutes preserving the practice of allocution have insisted that the right be held inviolate. In this case, the lead opinion, although correctly asserting that “[m]ost other jurisdictions that recognize the right of allocution consider it to be statutory [rather than constitutional] in nature,” ignores the fact that in recent years these courts have resisted all requests to assess the harmlessness of a denial of allo-cution and instead have used a per se practice of requiring a new penalty hearing. For example, the Colorado Supreme Court concluded that failure to grant allocution “renders [the sentence] invalid” and explained that the defendant had a “right of allocution” prior to sentencing “which cannot be withheld from him.” People v. Emig, 177 Colo. 174, 493 P.2d 368, 369-70 (1972). The Alaska Supreme Court stated, “[T]here is no substitute for the impact on sentencing which a defendant’s own words might have if he chooses to make a statement.” Mohn v. State, 584 P.2d 40, 44 (Alaska 1978). The New Mexico Supreme Court concluded that “failure to [give the defendant an opportunity to speak before sentencing] renders the sentence invalid.” Tomlinson v. State, 98 N.M. 213, 215, 647 P.2d 415, 417 (1982). The Rhode Island Supreme Court stated that violation of the right to allocution would “require this court to remand the case for resentencing.” State v. Nicoletti, 471 A.2d 613, 618 (R.I.1984). Similarly, the United States Supreme Court in its memorandum opinion in Van Hook v. United States, 365 U.S. 609, 81 S.Ct. 823, 5 L.Ed.2d 821 (1961), apparently held that on direct appeal a defendant who was not offered an opportunity to make a statement prior to sentencing was entitled to resentencing. See also Hill v. United States, 368 U.S. 424, 429 n. 6, 82 S.Ct. 468, 472, 7 L.Ed.2d 417 (1962).
The above decisions, all invoking a per se rule of resentencing when a trial court has denied a defendant’s statutory right of allo-cution, demonstrate that allocution continues to play a unique and important role in criminal sentencing, despite innovations in criminal procedure that permit defendants to testify and to have counsel. As the United States Supreme Court has stated, “None of these modern innovations lessens the need for the defendant, personally, to have the opportunity to present to the court his plea in mitigation.” Green v. United States, 365 U.S. 301, 304, 81 S.Ct. 653, 655, 5 L.Ed.2d 670 (1961). Although most of the above cases involved noncapital sentencing, their rationale is only strengthened in a capital case. The nature of capital sentencing, particularly the constitutional requirement that a convicted capital defendant have an unrestricted opportunity to present mitigating evidence, see, e.g., Eddings v. Oklahoma, 455 U.S. 104, 112, 102 S.Ct. 869, 875, 71 L.Ed.2d 1 (1982); Lockett v. Ohio, 438 U.S. 586, 604, 98 S.Ct. 2954, 2964, 57 L.Ed.2d 973 (1978); Jurek v. Texas, 428 U.S. 262, 276, 96 S.Ct. 2950, 2958, 49 L.Ed.2d 929 (1976); State v. Wood, 648 P.2d 71, 86 (Utah), cert. denied, 459 U.S. 988, 103 S.Ct. 341, 74 L.Ed.2d 383 (1982), leads me to conclude that if a trial court has denied allocution erroneously, this error typically will create more prejudice to a capital defendant than to a non-capital defendant. As it was originally, allocution today is of the utmost importance in permitting defendants to plead for their lives.
The State has conceded that Utah Code Ann. § 77-35-22 (now Utah R.Crim.P. 22(a)), which provides that “before imposing sentence the court shall afford the defendant an opportunity to make a statement in his own behalf and to present any information in mitigation of punishment,” required the court to allow Young to speak to the jury before it retired to deliberate his sentence. Having conceded the statutory error, the State argues only that it was harmless error not meriting a new sentencing trial. I disagree and conclude that it is at least reversible error, if not error per se, under our statutory harmless error standard.
The lead opinion has failed to acknowledge the State’s concession of statutory error. Understandably, once the lead opin*372ion has determined that even under a constitutional error standard, the denial of al-locution to Young was harmless beyond a reasonable doubt, those joining in it would be free to ignore whether statutory error occurred. Instead, however, they decide, sua sponte and despite the State’s concession, that in fact no statutory error occurred here because the trial court satisfied section 77-35-22 when it permitted Young to address the trial judge after sentencing. Particularly in light of this court’s recent opinion in State v. Andrews, 843 P.2d 1027, 1028 (Utah 1992), concluding that “[i]n substance and effect, the verdict of the jury is the sentence,” the lead opinion’s conclusion today is unsupportable. The right of allocution contemplates permitting a convicted defendant to speak pri- or to the fixing of sentence in order to provide the defendant a chance to present a plea in mitigation. See, e.g., Green, 365 U.S. at 304, 81 S.Ct. at 655. In capital prosecutions in Utah, as in most states, the jury actually determines the defendant’s sentence, and the judge has no discretion to modify the verdict on sentencing. See Andrews, 843 P.2d at 1028. Therefore, if a jury determines sentence and a statute has established a capital defendant’s right of allocution, the defendant must have the opportunity to make this plea to the jury. Allocution in a capital sentencing trial after the jury has returned its verdict on sentencing is a meaningless formality, “no more than an empty gesture.” Tomlinson, 647 P.2d at 417; see J. Thomas Sullivan, The Capital Defendant’s Right to Make a Personal Plea for Mercy: Common Law Allocution and Constitutional Mitigation, 15 N.M.L.Rev. 41, 56-60 (1985) [hereinafter J. Thomas Sullivan].
Furthermore, allocution in precisely the form Young requested, not the lead opinion’s “postverdict allocution,” comports most closely with Utah’s death penalty statute. After stating that at sentencing the parties may present evidence of any fact either in aggravation or mitigation of the penalty, the statute provides, “The state’s attorney and the defendant shall be permitted to present argument for or against sentence of death.” Utah Code Ann. § 76-3-207(2). If the defendant has no personal entitlement to argue to the jury, one would expect the statute to permit “the state’s attorney and the defendant’s attorney” to present argument or, in the alternative to permit “the state and the defendant” to present argument. Therefore, one obvious implication of the literal language in Utah’s death penalty statute is that it permits the defendant personally to address the sentencer as part of Utah’s extensive sentencing procedure.
In footnote 179, the lead opinion quotes a case asserting that allocution in its original common law form had nothing to do with “afford[ing] an opportunity to present mitigating evidence or to plead for leniency.” In fact, for centuries allocution has served exactly this role, and Utah now expressly preserves this role in its allocution rule, which states, “Before imposing sentence the court shall afford the defendant an opportunity to make a statement in his own behalf and to present any information in mitigation of punishment.” Utah R.Crim.P. 22(a). Because by statute allocution is a right to introduce a mitigating statement and because capital defendants are entitled to present to the sentencing jury any mitigating information, courts must permit allocution at the sentencing phase, when requested, rather than post-verdict. Accord People v. Davis, 794 P.2d 159, 191-92 (Colo.1990) (interpreting allocution statute identical to Utah’s rule as establishing right of capital defendant to make statement to sentencing jury), cert. denied, 498 U.S. 1018, 111 S.Ct. 662, 112 L.Ed.2d 656 (1991).
Thus, the only question I believe this court should address regarding the statutory error claim is whether the error is reversible or harmless. We have stated that statutory error is reversible “when a reasonable likelihood exists that absent the error, the result would have been more favorable to the defendant.” State v. Dibello, 780 P.2d 1221, 1230 (Utah 1989). When the error occurs in a penalty trial in which the jury voted unanimously for death, this can only mean that there must be a reasonable likelihood that one of the *373twelve jurors would not have voted for death had the error not occurred. We also have explained that this “reasonable likelihood” standard is met when, as a result of the error, our confidence in the outcome is undermined. Id.
The State bases its harmless error contention on two points: (1) the “overwhelming evidence of aggravation”; and (2) the fact that Young’s proffered statement “did not add anything to the evidence presented through other witnesses.” For both of these reasons, the State suggests that Young cannot meet the “reasonable likelihood” standard. The lead opinion similarly concludes, after one paragraph of discussion, that given the “many aggravating circumstances” and defendant’s presentation of “days of testimony in mitigation,” Young’s allocution statement “was not likely to influence the result” and its potential impact upon a jury was “minimal at best.” The application of this calculus in a capital sentencing proceeding is astonishing.
It is simply inconceivable, given the nature of our process for imposing capital punishment, that a defendant’s personal statement would not add anything to the jury’s understanding and view of his character, his crime, and his life. As many of the cases cited above adequately express, “there is no substitute” for a defendant’s personal statement. Mohn, 584 P.2d at 44. “The most persuasive counsel may not be able to speak for a defendant as the defendant might, with halting eloquence, speak for himself.” Green, 365 U.S. at 304, 81 S.Ct. at 655. The verbal content of the statement could be completely duplicative of evidence otherwise introduced, and yet the statement would not be cumulative, because it would convey to the jury intangible personal qualities of the defendant that the jury can experience and assess in no other way.
Furthermore, in this case the content of the proffered statement was not duplica-tive of evidence already introduced. It included Young’s recognition that he had made a mistake. The lead opinion asserts that Young showed no remorse. In fact, however, the proffered statement did contain an implicit plea for forgiveness, and had Young personally made the statement, the jury, rather than this court, would have been left to evaluate the existence and depth of his remorse. The statement also included his assessment of his ability to work and contribute to the community. It included an expression of his desire to “start a new life.” None of these elements could have been communicated by anyone other than Young himself.
Finally, because the death penalty is inherently different from any other punishment, modern capital punishment jurisprudence forcefully establishes that the sentencing jury must consider the unique personality and circumstances of every capital defendant. This court has echoed the United States Supreme Court’s requirements that a capital sentencer must consider all factors in mitigation, including “any aspect of a defendant’s character ... that the defendant proffers,” Wood, 648 P.2d at 86 (quoting Lockett, 438 U.S. at 605, 98 S.Ct. at 2965), and that the sentencer must make an individualized determination of the appropriateness of death for the particular defendant, id.
The lead opinion argues that to communicate these factors to the sentencing jury, Young simply needed to take the witness stand and testify under oath and subject himself to cross-examination. It even suggests that this alternative would have produced greater impact on the jury than an unsworn statement. Yet the lead opinion also acknowledges that the trial court itself advised Young against testifying, because of the high risk of adverse impact from impeachment on cross-examination. A capital defendant should have the right to make a plea for mercy free from this adverse impact. “[T]he accused making a plea for mercy does not intend to advance or to dispute facts, but instead uses the plea to ask for lenience or understanding in the sentencer’s decision. If the accused does not stray from this subject matter in his statement, the plea made does not justify traditional impeachment.” J. Thomas Sullivan at 63. Therefore, if Young did *374have a statutory right to allocution, as the State concedes, then the fact that he could have introduced his statement by taking the witness stand is irrelevant in determining whether denial of his right was harmless.
The State's only remaining argument is that allocution would have been unavailing to Young because of the “overwhelming” evidence of aggravation. Given the nature of capital sentencing, very few circumstances would allow a reviewing court to properly conclude that aggravating evidence “overwhelms” mitigating factors. In capital trials, we ask the sentencing jury to. make a subjective moral evaluation of the “worth” of the defendant and of the defendant’s character. As this court stated in Wood, “The ultimate purpose in the penalty phase is not one of factfinding, but ... is a matter of judgment.” 648 P.2d at 84. As a result, an “inherent lack of predictability” characterizes the jury’s assessment of the appropriate sentence. McCleskey v. Kemp, 481 U.S. 279, 311, 107 S.Ct. 1756, 1777, 95 L.Ed.2d 262 (1987). When an extended penalty trial has been held, presenting substantial issues in both aggravation and mitigation, a reviewing court should undertake to reweigh the jury’s subjective assessment only with extreme caution.
Many others have recognized that the role played by a capital sentencing jury is unlike the role of any other jury. For instance, the United States Supreme Court explained the “fundamental difference” as follows:
In returning a conviction, the jury must satisfy itself that the necessary elements of the particular crime have been proved beyond a reasonable doubt. In fixing a penalty, however, there is no similar “central issue” from which the jury’s attention may be diverted_ [T]he jury ... is free to consider a myriad of factors to determine whether death is the appropriate punishment.
California v. Ramos, 463 U.S. 992, 1008, 103 S.Ct. 3446, 3457, 77 L.Ed.2d 1171 (1983). Justice Powell also articulated this distinction:
Underlying the question of guilt or innocence is an objective truth: the defendant, in fact, did or did not commit the acts constituting the crime charged. From the time an accused is first suspected to the time the decision on guilt or innocence is made, our criminal justice system is designed to enable the trier of fact to discover that truth according to the law. But triers of fact can err, and an innocent person can be pronounced guilty. In contrast, the law provides only limited standards for assessing the validity of a sentencing decision. The sentencer’s function is not to discover a fact, but to mete out just deserts as he sees them. Absent a mandatory sentence, there is no objective measure by which the sentencer’s decision can be deemed correct or erroneous if it is duly made within the authority conferred by the legislature.
Bullington v. Missouri, 451 U.S. 430, 450, 101 S.Ct. 1852, 1864, 68 L.Ed.2d 270 (1981) (Powell, J., dissenting). And Justice Marshall wrote:
Unlike the determination of guilt or innocence, which turns largely on an evaluation of objective facts, the question whether death is the appropriate sentence requires a profoundly moral evaluation of the defendant’s character and crime_ Even in the face of overwhelming aggravating evidence, the sen-tencer has discretion to act with leniency and refuse to impose the death sentence .... [Predicting the reaction of a sentencer to a proceeding untainted by constitutional error on the basis of a cold record is a dangerously speculative enterprise.
Satterwhite v. Texas, 486 U.S. 249, 261-62, 108 S.Ct. 1792, 1800, 100 L.Ed.2d 284 (1988) (Marshall, J. dissenting).
In Young’s case, the denial of allocution had no impact on any factual determinations; instead, it directly affected only the jury’s exercise of moral judgment. This judgment inherently requires a subjective rather than an objective weighing; therefore, it is a judgment that, like a credibility determination, belongs uniquely to the trier *375of fact. As we explained in State v. Holland, 111 P.2d 1019 (Utah 1989), capital sentencing factors “are not, in truth, weighable.... These factors have largely subjective value and therefore vary in their ‘weight’ or persuasiveness for or against the death penalty with each judge or juror according to his or her own background and prior experiences.” Id. at 1028. I do not understand how we as a reviewing court can conclude that there is no reasonable likelihood that defendant’s personal plea for mercy would have affected this personal moral judgment. Almost every case presents a reasonable likelihood that “[a] previously held opinion regarding the appropriate sentence may well be modified after listening to ... defendant’s statements in mitigation of punishment.” Kent v. State, 287 Md. 389, 412 A.2d 1236, 1239 (1980), quoted in Harris v. State, 306 Md. 344, 509 A.2d 120, 127 (1986). Furthermore, in State v. Wood, 648 P.2d 71 (Utah 1981) (per curiam), we held that to impose the death penalty, a capital sentencing jury must both “be persuaded beyond a reasonable doubt that total aggravation outweighs total mitigation, and ... must further be persuaded, beyond a reasonable doubt, that the imposition of the death penalty is justified and appropriate in the circumstances.” Id. Even if this court could properly accept the State’s claim that aggravation entirely overwhelms mitigation, we nevertheless could not reliably conclude under the second part of the Wood analysis that a statement of allocution would create no reasonable likelihood of a different outcome.
One aspect of the speculation inherent in the lead opinion’s approach particularly troubles me. This case includes considerable evidence about defendant’s cognitive disabilities and long-standing brain damage. It is entirely possible, although we have no way of knowing, that something in his expression, language, or patterns of speech might have caused some jurors to reassess that evidence. At the very least, the jurors would have had the opportunity to be confronted by him as a person, with a personal voice and a personal story, before making the decision to permit life or impose death. “The right of a defendant to address the sentencing court in person may well be of inestimable impact_ A defendant’s own statement may provide a most valuable insight into his character at a time of great value to the sentencing court.” In re Stevens, 144 Vt. 250, 478 A.2d 212, 217 (1984).
The only other state court that has considered the precise issue facing us today concluded that a capital defendant “who timely asserts his right to allocute and provides an acceptable proffer must be afforded a fair opportunity to exercise this right. If the right so asserted is denied by the court, as here, the sentence must be vacated and a new sentencing proceeding conducted.” Harris, 509 A.2d at 127 (emphasis added). Similarly, in the previously cited noncapital sentencing decisions of other states and the United States Supreme Court, the courts have not attempted to determine the harmlessness of the error, but have created a per se requirement of anew penalty hearing when allocution has been denied. See Van Hook, 365 U.S. at 609, 81 S.Ct. at 823; Mohn, 584 P.2d at 44; Emig, 493 P.2d at 369-70; Tomlinson, 647 P.2d at 417; Nicoletti, 471 A.2d at 618; see also Ashe v. State, 586 F.2d 334, 336 (4th Cir.1978), cert. denied, 441 U.S. 966, 99 S.Ct. 2416, 60 L.Ed.2d 1072 (1979).
This court should adopt a per se rule requiring that when a capital defendant has requested and been denied allocution, the verdict must be vacated and a new penalty hearing conducted. We should not second-guess the unique balancing performed by capital sentencing jurors. Furthermore, I believe that this is the first time this court has applied a statutory harmless error analysis to the wrongful exclusion (as opposed to the wrongful inclusion) of information at a capital sentencing trial. The trial court’s refusal to permit Young to make his brief statement kept from the jury an entire category of mitigating evidence. Any inquiry into the harmleSsness of this type of error is an exercise in “unguided speculation.” Holloway v. Arkansas, 435 U.S. 475, 491, 98 S.Ct. 1173, 1182, 55 L.Ed.2d 426 (1978).
*376Even without a per se rule, I would hold in this case that had Young been permitted to speak to the jury, a different outcome was reasonably likely. My confidence in the result has been seriously undermined.
B. Constitutional Error
Contrary to the lead opinion’s conclusion, failure to afford defendant an opportunity for allocution rises to the level of a violation of both the federal and state constitutions.
1. Federal Constitutional Requirements
The United States Supreme Court has held that a trial court’s failure to ask whether the defendant would like to make a statement prior to sentencing was not constitutional error and hence not reversible in a collateral appeal. Hill v. United States, 368 U.S. 424, 82 S.Ct. 468, 7 L.Ed.2d 417 (1962). But the Court has never addressed whether denying allocution to a defendant who has affirmatively requested it is constitutional error. See McGautha v. California, 402 U.S. 183, 218 n. 22, 91 S.Ct. 1454, 1473 n. 22, 28 L.Ed.2d 711 (1971); Hill, 368 U.S. at 429, 82 S.Ct. at 472. Although the lead opinion states that “the federal courts of appeals have generally determined that the right to allocution is not a constitutionally protected right,” none of the lead opinion’s cited cases addresses the issue that both McGautha and Hill explicitly left unresolved. The one federal appellate court that has squarely considered whether there is a constitutional right to allocution when the defendant requests it concluded that “it is a denial of due process not to grant the defendant’s request.” Ashe v. State, 586 F.2d 334, 336 (4th Cir.1978), cert. denied, 441 U.S. 966, 99 S.Ct. 2416, 60 L.Ed.2d 1072 (1979).
Because “death is different,” the state imposes capital punishment only after following the strictest legal and constitutional safeguards. The United States Supreme Court has protected Eighth Amendment rights to be free from cruel and unusual punishment by requiring that capital punishment be imposed only by means of carefully channelled discretion and individualized determination regarding every defendant. Further, defendants are entitled to present any mitigating factor to the sentencing body. See, e.g., Eddings v. Oklahoma, 455 U.S. 104, 112, 102 S.Ct. 869, 875-76, 71 L.Ed.2d 1 (1982); Lockett v. Ohio, 438 U.S. 586, 604, 98 S.Ct. 2954, 2964-65, 57 L.Ed.2d 973 (1978); Jurek v. Texas, 428 U.S. 262, 276, 96 S.Ct. 2950, 2958, 49 L.Ed.2d 929 (1976); Woodson v. North Carolina, 428 U.S. 280, 303-05, 96 S.Ct. 2978, 2990-92, 49 L.Ed.2d 944 (1976); State v. Wood, 648 P.2d 71, 86 (Utah), cert. denied, 459 U.S. 988, 103 S.Ct. 341, 74 L.Ed.2d 383 (1982). The Court has also protected Fourteenth Amendment due process rights in capital cases by requiring an “especially vigilant concern for procedural fairness.” Strickland v. Washington, 466 U.S. 668, 704, 104 S.Ct. 2052, 2073, 80 L.Ed.2d 674 (1984) (Brennan, J., concurring in part and dissenting in part).
As I have argued above, allocution can provide significant mitigating information. As I also have discussed above, the United States Supreme Court has never decided whether denial of allocution when specifically requested is a due process violation. Nor has the Court decided in the post-Furman era whether denial of allocution to a capital defendant violates the Court’s rigorous Eighth Amendment capital sentencing safeguards. In light of the Court’s silence on these issues and given the Court’s concern in other contexts for reliability, procedural fairness, and preserving the opportunity to present any mitigating information, I conclude that the denial of Young’s requested allocution violated his federal rights. Foreclosing the jury’s opportunity to hear Young’s personal statement prevented the jury from making the individualized sentencing determination required by the Eighth Amendment. Given that Utah mandates, by statute, an opportunity for allocution, its refusal in a capital trial violates the Fourteenth Amendment’s requirement of procedural fairness. Our reference in State v. Lorrah, 761 P.2d 1388, 1390 (Utah 1988), to a “defendant’s due process right of allocution” may have been without analysis, as the lead opinion points out, but it was correct.
*3772. State Constitutional Requirements
The lead opinion relies on several older cases, such as State v. Kelbach, 23 Utah 2d 231, 461 P.2d 297 (Utah 1969), vacated in part, 408 U.S. 935, 92 S.Ct. 2858, 33 L.Ed.2d 751 (1972), for the proposition that we have examined and presumably settled the issue of the constitutional status of the right to allocution. I submit that the law of capital punishment has changed so dramatically in the past twenty years that previous analysis is unhelpful and not controlling. Furthermore, this court has never, to my knowledge, undertaken a state constitutional analysis of the issue.
Article I, section 12 of the Utah Constitution states, “In criminal prosecutions the accused shall have the right to appear and defend in person and by counsel [and] to testify in his own behalf_” This provision, in my view, plainly entitles a criminal defendant to stand before judge and jury and examine witnesses, present motions, make opening and closing arguments, and do anything else that licensed legal counsel is allowed to do. The conjunctive language entitles a defendant to do any and all of this whether or not the defendant is also assisted by counsel. The fact that the defendant’s right to testify is separately enumerated makes clear that the right to “appear and defend” is an additional entitlement to address the judge and jury. I also believe that it would be a failure of justice to suggest that these rights should, in a capital trial, apply only to the guilt phase and not to the penalty phase. Further, the provision of Utah Code Ann. § 76-3-207, discussed above, stating that at the penalty phase, the “state’s attorney and the defendant” may present argument feinforces this reading.
In this case, the State has already argued that defendant’s proffered statement introduced no new evidence. If so, the State would have no grounds for objecting to it as closing argument. Had Young chosen self-representation, his closing argument would have provided the forum for his statement as a matter of course. The fact that Young had counsel at trial should not have prevented him from making the closing argument or sharing the closing argument with counsel. I believe article I, section 12 guaranteed Young the right to make his statement. Additionally, although a capital defendant’s right to make closing argument at the penalty phase may not be equivalent to the right to allocution, I would conclude that because of the similarity between the rights, the protections of article I, section 12, as well as the due process clause of article I, section 7, encompass the right of a capital defendant to speak to the sentencing jury.
3. Constitutional Harmless Error
Because I believe a constitutional error has occurred, the burden shifts to the State to show that the error was harmless beyond a reasonable doubt. See Chapman v. California, 386 U.S. 18, 24, 87 S.Ct. 824, 828-29, 17 L.Ed.2d 705 (1967); State v. Tuttle, 780 P.2d 1203, 1213 (Utah 1989), cert. denied, 494 U.S. 1018, 110 S.Ct. 1323, 108 L.Ed.2d 498 (1990). Precluding the jury from considering a defendant’s character and proffered mitigation “creates the risk that the death penalty will be imposed in spite of factors which may call for a less severe penalty. When the choice is between life and death, that risk is unacceptable and incompatible with the commands of the Eighth and Fourteenth Amendments.” Lockett v. Ohio, 438 U.S. 586, 605, 98 S.Ct. 2954, 2965, 57 L.Ed.2d 973 (1978). For reasons apparent in the preceding discussion (part II.A) of Young’s statutory right to allocution, the State has not proved beyond a reasonable doubt that the statement was cumulative or irrelevant. Furthermore, the State has not shown that absent the error, it is “unlikely beyond a reasonable doubt that the result would have been different.” Tuttle, 780 P.2d at 1213. The error kept significant information from the sentencer, namely, what the defendant had to say for himself and the manner in which he would have said it. Therefore, Young should have been allowed to present his brief statement to the jury. The sentence should be vacated, and a new sentencing trial held.
*378III. INSTRUCTION ON MERCY OR SYMPATHY
(lead opinion part XVII.A)
Young claims that mercy or sympathy based on the mitigating evidence presented at the penalty phase may play a legitimate role in the deliberations of a capital sentencing jury. From this premise, he articulates two constitutional violations: First, he claims that by instructing the jurors to disregard their sympathetic or merciful responses to the mitigation, the trial court violated his rights under the federal and state constitutions. Second, Young claims that the trial court’s refusal to give his requested sympathy instruction also violated his constitutional rights. These claims go to the heart of contemporary capital punishment jurisprudence and require a careful analysis of the precedents of this court and of the United States Supreme Court.
Historically, the death penalty was mandatory for defendants convicted of a capital offense. Prior to Furman v. Georgia, 408 U.S. 238, 92 S.Ct. 2726, 33 L.Ed.2d 346 (1972), however, most states gave their capital sentencing juries unguided discretion to decide whether to impose the death penalty. See McGautha v. California, 402 U.S. 183, 199-202, 91 S.Ct. 1454, 1463-64, 28 L.Ed.2d 711 (1971). As McGautha explained, the purpose of providing juries with this discretion was to permit them to be merciful. Id. at 200-01, 91 S.Ct. at 1463-65. The McGautha Court refused to accept the claim that because such a system lacked any principled basis for selecting those defendants who would receive death rather than mercy, it was unconstitutional. The Court said, “To identify before the fact those characteristics of criminal homicides and their perpetrators which call for the death penalty ... [is] beyond present human ability.” Id. at 204, 91 S.Ct. at 1465-66.
I acknowledge that the Court’s decision one year later in Furman overruled McGautha. Furman found that state statutes that gave juries unguided discretion to impose the death penalty were unconstitutional. 408 U.S. at 239, 92 S.Ct. at 2727. Significantly, however, Furman did not suggest that state statutes giving juries unguided discretion to refuse to impose the death penalty were unconstitutional, nor did it challenge McGautha’s reaffirmation of earlier Supreme Court cases establishing that mercy could be extended freely. See McGautha, 402 U.S. at 200-01, 91 S.Ct. at 1463-64. Thereafter, states sought to do exactly what McGautha had considered “beyond human ability”: to establish criteria upon which the jury properly could choose to impose the death penalty.
When these revised death penalty schemes reached the United States Supreme Court four years after Furman, it upheld those that in its view established rational standards for selecting defendants who would receive capital punishment. See, e.g., Gregg v. Georgia, 428 U.S. 153, 96 S.Ct. 2909, 49 L.Ed.2d 859 (1976) (statute permitted death penalty only upon finding at least one of ten statutory aggravating factors); Proffitt v. Florida, 428 U.S. 242, 96 S.Ct. 2960, 49 L.Ed.2d 913 (1976) (statute permitted death penalty only upon finding sufficient aggravation from statutory list of eight aggravating factors); Jurek v. Texas, 428 U.S. 262, 96 S.Ct. 2950, 49 L.Ed.2d 929 (1976) (statute permitted death penalty only for five specified types of intentional killings). In contrast, the Court struck down a mandatory death sentence statute that stripped the jury of its ability to extend mercy, explaining that it did not permit consideration of the “compassionate or mitigating factors stemming from the diverse frailties of humankind.” Woodson v. North Carolina, 428 U.S. 280, 304, 96 S.Ct. 2978, 2991, 49 L.Ed.2d 944 (1976) (plurality). In Gregg, Justice White recognized the role of mercy in capital sentencing: “The Georgia Legislature has plainly made an effort to guide the jury in the exercise of its discretion, while at the same time permitting the jury to dispense mercy on the basis of factors too intangible to write into a statute.” 428 U.S. at 222, 96 S.Ct. at 2947-48. (White, J., concurring). Two years later, the Court held that the sentencing jury must “not be precluded from *379considering as a mitigating factor, any aspect of a defendant’s character or record.” Lockett v. Ohio, 438 U.S. 586, 604, 98 S.Ct. 2954, 2964, 57 L.Ed.2d 973 (1978) (emphasis in original).
I am compelled to conclude that as of 1978, the majority of justices on the United States Supreme Court believed that mercy and compassion were central to capital sentencing. The unmistakable meaning of the line of cases beginning with Furman was that while a jury could decide to impose death only by following carefully chan-nelled procedures designed to minimize the arbitrariness of those cases in which death actually was imposed, the jury could decline to impose a death sentence for any reason—merciful, compassionate, sympathetic, or otherwise. See Gregg, 428 U.S. at 199, 203. The Court reinforced this position in Eddings v. Oklahoma, 455 U.S. 104, 110, 102 S.Ct. 869, 875, 71 L.Ed.2d 1 (1982), and again in Caldwell v. Mississippi, 472 U.S. 320, 330-31, 105 S.Ct. 2633, 2640-41, 86 L.Ed.2d 231 (1985). In California v. Brown, 479 U.S. 538, 107 S.Ct. 837, 93 L.Ed.2d 934 (1987), although the Court upheld an instruction prohibiting the jury from being influenced by “mere sympathy,” it plainly preserved a place for sympathetic responses to evidence presented at the sentencing phase, stating that a reasonable juror would interpret the challenged instruction “as a directive to ignore only the sort of sympathy that would be totally divorced from the evidence adduced during the penalty phase.” Id. at 542, 107 S.Ct. at 840.
Particularly after Brown, it is disingenuous to suggest that the watershed cases in this country’s modern capital punishment jurisprudence did not intend at least to preserve a sacred place for sympathetic and merciful responses to a defendant’s mitigating evidence.4 Nevertheless, in rejecting Young’s claim that a jury instruction that interferes with a juror’s ability to respond sympathetically or mercifully to the mitigating evidence violated his rights, today’s lead opinion ignores this history and instead focuses solely on the opinion in the habeas corpus case of Saffle v. Parks, 494 U.S. 484, 110 S.Ct. 1257, 108 L.Ed.2d 415 (1990). Those joining the lead opinion appear erroneously to believe that Parks has disposed of the federal constitutional issue. Admittedly, dicta in Parks suggest that the Court is contemplating a retreat from its prior cases, including Brown. But contrary to the lead opinion’s assertion, the Parks opinion did not reject on its merits the constitutional claim now before us. Like Young, Parks had sought a rule that an antisympathy instruction that precluded the jury from fully considering his mitigating evidence violated the Constitution. The Court held that because the principle urged by Parks was not “dictated by our prior cases” and therefore would require the creation of a “new rule of constitutional law,” then following the mandate of Penry v. Lynaugh, 492 U.S. 302, 109 S.Ct. 2934, 106 L.Ed.2d 256 (1989), Parks could not benefit from such a rule in a collateral attack. Parks, 494 U.S. at 486, 110 S.Ct. at 1258-59. The Court therefore did not decide whether, as a question of first impression, the Constitution required such a rule. The holding in Parks is only that no such rule clearly existed under the Court’s previous interpretations of the Eighth and Fourteenth Amendments. Id. at 489, 110 S.Ct. at 1260-61.
The United States Supreme Court’s refusal to interpret its prior cases as compelling the rule urged by Parks, and its view of contrary state court rulings as “well-considered precedents,” see id. at 486, 110 S.Ct. at 1258-59, undoubtedly casts doubt on the strength of Young’s federal constitutional argument. Yet until the Supreme Court actually considers the merits of the claim, I deem it this court’s obligation to determine for itself what the federal consti*380tution requires regarding this issue of first impression. The philosophy underlying this country’s approach to capital punishment in the past two decades includes a recognition that sympathetic responses by the sentencer to sentencing evidence are entirely appropriate. As the Supreme Court has stated, “The right to have the sentencer consider and weigh relevant mitigating evidence would be meaningless unless the sentencer was also permitted to give effects to its consideration.” Penry, 492 U.S. at 321, 109 S.Ct. at 2948 (emphasis added). The Court has also said, “The sen-tencer ... may determine the weight to be given relevant mitigating evidence.” Eddings, 455 U.S. at 114-15, 102 S.Ct. at 877. I am at a loss to understand how these principles can be squared with an instruction that tells sentencing juries to accord no weight to any sympathetic or compassionate feelings that evidence may produce. The Tenth Circuit arrived at precisely this conclusion when it reached the merits of the claim in the Parks case, Parks v. Brown, 860 F.2d 1545, 1554-56 (10th Cir.1988), rev’d sub nom. Saffle v. Parks, 494 U.S. 484, 110 S.Ct. 1257, 108 L.Ed.2d 415 (1990), as has the California Supreme Court, see, e.g., People v. Wade, 43 Cal.3d 366, 233 Cal.Rptr. 48, 59, 729 P.2d 239, 250 (1987), vacated by 44 Cal.3d 975, 244 Cal.Rptr. 905, 750 P.2d 794 (1988); People v. Lanphear, 36 Cal.3d 163, 203 Cal.Rptr. 122, 124, 680 P.2d 1081, 1083 (1984).
In an apparent attempt to reconcile these principles, the lead opinion quotes dictum from Parks for the proposition that individualized, nonarbitrary capital sentencing requires a "moral inquiry,” not an “emotional response.”5 It is specious, however, to suggest that sympathy, mercy, pity, or compassion does not constitute part of a “moral” response to evidence in aggravation and mitigation. A moral weighing is not an objective weighing; while it should not be based on raw emotion, it does require each juror to respond in human terms to the sentencing evidence in assessing the unique humanity of the defendant. It is not possible for jurors properly to conduct this exercise without reference to their feelings. Yet this is precisely what an antisympathy instruction asks jurors to attempt.
Ironically, the lead opinion today affirms the antisympathy instruction while relying on the United States Supreme Court’s recent decision in Payne v. Tennessee, holding that “justice due to the accuser” supports the use of victim impact evidence at the penalty phase. 501 U.S. -, 111 S.Ct. 2597, 115 L.Ed.2d 720 (1991). The inevitable (and, I am sure, intended) effect of victim impact evidence is to arouse a sen-tencer’s natural human feelings, and the Payne standard now permits juries to consider the “emotional impact” of the crime on the victim’s family. See id. 501 U.S. at -, -, 111 S.Ct. at 2604, 2609, 115 L.Ed.2d 720. Payne observed that prohibiting the use of victim impact evidence “deprives the State of the full moral force of its evidence.” Id. 501 U.S. at -, 111 S.Ct. at 2608, 115 L.Ed.2d 720. Of course, this “full moral force” arises in part precisely because of the emotional response the victim evidence produces, both in terms of sympathy for the victims and disgust, outrage, and a desire for vengeance directed toward the defendant. The lead opinion today would permit the state to benefit from the full moral force of its evidence while denying the same benefit to defendant. This result is doubly ironic given the heightened obligation our criminal justice system has to protect the rights of defendants. Accordingly, I conclude that federal constitutional safeguards should prohibit antisympathy instructions that deny the propriety of a sentencer’s compassionate responses to a defendant’s mitigating evidence.
I also believe that Utah law compels us to prohibit antisympathy instructions that mislead the jury. I reach this conclusion in part because of the high regard that this *381state places “on the value of all human life and the humanity of every human being.” State v. Holland, 777 P.2d 1019, 1028 (Utah 1989). I also base this conclusion on prior cases in which we have implicitly approved the persuasive federal constitutional analysis in the line of cases from Gregg through Eddings. See, e.g., Holland, 777 P.2d at 1028; State v. Lafferty, 749 P.2d 1239, 1259 (Utah 1988). Because the principles contained in this line of cases reflect the protections this court ought also to afford to capital defendants, I would explicitly incorporate those principles today, regardless of the apparent retrenchment from them by the current United States Supreme Court, as guidance for interpreting the cruel and unusual punishment clause of the Utah Constitution.
Furthermore, I reach this conclusion because of our familiar pronouncements in State v. Wood, 648 P.2d 71 (Utah), cert. denied, 459 U.S. 988, 103 S.Ct. 341, 74 L.Ed.2d 383 (1982), regarding the requirements of the Utah death penalty statute. We wrote:
[T]he sentencing authority must decide ... how compelling or persuasive the totality of the mitigating factors are when compared against the totality of the aggravating factors.... [To impose death,] upon consideration of all the circumstances relating to this defendant and this crime the sentencing authority must be convinced beyond a reasonable doubt that the death penalty should be imposed.
Id. at 83-84. We expounded upon the Wood standard in Holland, explaining that in Utah the factors a capital sentencer weighs “are not, in truth, weighable.... These factors have largely subjective value and therefore vary in their ‘weight’ or persuasiveness for or against the death penalty with each judge or juror according to his or her own background and prior experiences.” 777 P.2d at 1028. Consequently, I conclude that trial courts usurp the sentencing jury’s role when they instruct the jury to avoid sympathy in the sentencing phase of a capital trial. This state’s profound respect for each person’s humanity and the obligation of the sentencer to bring her or his own background and human experience to bear in assessing the persuasiveness of every aspect of the defendant’s mitigation render it error to prohibit the influence of sympathy in weighing penalty options.
Having concluded that this court should not countenance jury instructions that may preclude the sentencer from considering whatever sympathy the mitigating evidence produces, I examine the propriety of the particular instructions given in this case. This inquiry requires determining “what a reásonable juror could have understood the charge as meaning.” California v. Brown, 479 U.S. at 541, 107 S.Ct. at 839. When reviewing potential error in capital sentencing instructions, appellate courts must remand for resentencing unless they “can rule out the substantial possibility that the jury may have rested its verdict on the ‘improper’ ground.” Mills v. Maryland, 486 U.S. 367, 377, 108 S.Ct. 1860, 1866-67, 100 L.Ed.2d 384 (1988).
At the guilt stage, the trial court instructed the jurors that they must deliberate “uninfluenced by pity for the defendant or by passion or prejudice against him.... The law forbids you to be governed by mere sentiment, conjecture, sympathy, passion, prejudice, public opinion or public feeling.” This instruction has been referred to in this case as the antisympathy instruction. Clearly such an instruction was proper at the guilt stage. At the sentencing stage, however, the trial court instructed the jury that instructions from the guilt stage continued to “apply in this penalty phase.” As at the guilt stage, the jurors had written copies of all the instructions with them in the' jury room during their sentencing deliberations. As a result, a reasonable juror easily could and would have concluded that it would be inappropriate to feel pity for, or to empathize or sympathize with, defendant when determining whether the death sentence should be imposed. Jurors disposed toward death, for example, could legitimately use the an-tisympathy instruction to disparage the judgment of holdout jurors. Thus, the instruction was improper.
*382In California v. Brown, the United States Supreme Court upheld an antisym-pathy instruction containing language identical to part of the antisympathy instruction in this ease. The instruction at issue in Brown told the jury not to be swayed by “mere sentiment, conjecture, sympathy, passion, prejudice, public opinion or public feeling.” 479 U.S. at 542, 107 S.Ct. at 840. I do not understand nor do I find persuasive the analysis by which the Brown majority concluded, contrary to the California Supreme Court, that reasonable California jurors would not have misunderstood this instruction as interfering with their consideration of mitigating evidence. Furthermore, although the antisympathy instruction in Young’s sentencing hearing contained the same language, it also included a demand not found in the language reviewed in Brown that the jury be “uninfluenced by pity for the defendant.” This additional element changes the context in which we must interpret the charge. The Brown majority concluded that, taken as a whole, the instruction only prohibited the jury from being influenced by “mere sympathy.” Id. This interpretation is baffling because in addition to its obvious and sensible consequence—that jurors properly may be influenced by sympathy “rooted in the aggravating and mitigating evidence,” id.—it similarly suggests that jurors also may properly be influenced by, for example, conjecture rooted in the evidence, prejudice rooted in the evidence, or public opinion rooted in the evidence, and must only avoid the influence of “mere” conjecture, “mere” prejudice, and “mere” public opinion. This is a construction to which I cannot subscribe. The adjective “mere” must be read as applying only to “sentiment,” the first term in the list, in order for the subsequent terms to make sense. See id. at 549, 107 S.Ct. at 843 (Brennan, J., dissenting). The Court also concluded, however, that it was “highly unlikely that any reasonable juror would almost perversely single out the word ‘sympathy’ from the other nouns.” Id. at 542, 107 S.Ct. at 840.
Regardless of whether this conclusion was correct in Brown, the presence in this case of an additional sentence instructing the jurors to avoid pity for defendant presents us with a different question and makes it highly likely that a conscientious juror in fact would notice the word “sympathy” in the list of otherwise impermissible factors. Accordingly, I refuse to view California v. Brown as dispositive of the federal constitutional issue or persuasive regarding the state statutory and constitutional issues relating to the propriety of the antisympathy instruction before us. There is a substantial possibility that the jury may have reached its verdict on improperly limited grounds; a fortiori, there is a consequent risk that Young’s death penalty was imposed in spite of factors that may have called for a less severe penalty. “When the choice is between life and death, that risk is unacceptable and incompatible with the commands of the Eighth and Fourteenth Amendments.” Lockett, 438 U.S. at 605, 98 S.Ct. at 2965. It is also incompatible with article I, section 9 of the Utah Constitution, and we should so hold.
A capital sentencing jury makes a moral determination, not a factual one. Although the determination must be fact based, the evaluation of those facts may and should involve emotional responses of anger, retribution, sympathy, and mercy. In contrast, these responses should not influence the determination of guilt. Jury instructions must not confuse these two roles. The original purpose of requiring a separate penalty phase in capital trials was to avoid arbitrary sentencing decisions based solely on raw emotion, not to strip sentencing of all human feeling. We expect sentencing juries to express the “conscience of the community”:
[0]ne of the most important functions any jury can perform in making such a selection [between life and death] is to maintain a link between contemporary community values and the penal system—a link without which the determination of punishment could hardly reflect “the evolving standards of decency that mark the progress of a maturing society.”
Witherspoon v. Illinois, 391 U.S. 510, 519 & n. 15, 88 S.Ct. 1770, 1775 & n. 15, 20 *383L.Ed.2d 776 (1968) (citations omitted); see Woodson v. North Carolina, 428 U.S. at 295, 96 S.Ct. at 2986-87. A conscience without room for sympathy and pity in meting out punishment is an ailing conscience indeed.
Accordingly, I agree with Young’s challenge to the extent it alleges error in the use at the penalty phase of the otherwise proper antisympathy instruction of the guilt phase. Furthermore, I am persuaded by the reasoning in Justice Stewart’s opinion concluding that juries should be affirmatively instructed regarding the proper function of sympathy (or mercy) in the penalty phase. I therefore join in that conclusion.
IV. GUILTY AND MENTALLY ILL INSTRUCTION AND VERDICT
(lead opinion part X)
The lead opinion erroneously infers that rule 21.5 of the Utah Rules of Criminal Procedure requires a defendant to assert a defense of not guilty by reason of insanity as a condition precedent for the availability of a guilty and mentally ill (GAMI) instruction and verdict. The language and history of the rule do not support this conclusion. Rule 21.5 reads in pertinent part:
(2)(a) If a defendant at trial asserts a defense of “not guilty by reason of insanity,” the court shall instruct the jury that it may find the defendant guilty, not guilty .by reason of insanity, guilty and mentally ill, guilty of a lesser offense, or guilty of a lesser offense due to mental ■illness but not an illness which would warrant full exoneration.
The foregoing language specifies what instructions the trial court must give about available verdicts when a defense of “not guilty by reason of insanity” (NGBRI) has been asserted. It does not preclude the availability of any of those verdict options when the specific NGBRI defense has not been asserted. The availability of instructions and related verdicts supported by the evidence is absolutely contemplated by rule 21(a):
The verdict of the jury shall be either “guilty” or “not guilty,” “not guilty by reason of insanity,” “guilty and mentally ill,” or “not guilty of the crime charged but guilty of a lesser included offense,” or “not guilty of the crime charged but guilty of, a lesser included offense and mentally ill” provided that when the defense of mental illness has been asserted and the defendant is acquitted on the ground that he was insane at the time of the commission of the offense charged, the verdict shall be “not guilty by reason of insanity.”
Utah’s Insanity Defense Act, which introduced the option of a GAMI verdict, was adopted in 1983 as part of a nationwide movement to “reform” the insanity defense. See R.D. Mackay,. Post-Hinckley Insanity in the U.S.A., 1988 Crim.L.Rev. 88, 88-90 [hereinafter R.D. Mackay]; Christopher Slobogin, The Guilty But Mentally III Verdict: An Idea Whose Time Should Not Have Come, 53 Geo. Wash.L.Rev. 494, 494-96 (1985).- Utah was one of only three states to abolish the traditional insanity defense, Utah Code Ann. § 76-2-305 (1978) (repealed 1983), by limiting the mental illness defense to conditions that negate the mens rea necessary for conviction of the crime. R.D. Mackay at 90. Although our rules still refer to “not guilty by reason of insanity” pleas and verdicts, section 76-2-305 as reenacted makes it clear that the defense is limited to an absence of mens rea:
(1) It is a defense to a prosecution under any statute or ordinance that the defendant, as a result of mental illness, lacked the mental state required as an element of the offense charged. Mental illness is not otherwise a defense.
(2) The defense defined in this section includes the defenses known as “insanity” and “diminished mental capacity.”
Utah Code Ann. § 76-2-305 (1992).
Thus, in Utah a defendant found NGBRI is, in effect, not guilty because he or she did not possess the mental state required to commit the offense charged. There is no assessment of a defendant’s capacity to appreciate the wrongfulness of *384his or her conduct or to conform his or her behavior to the requirements of law. A good deal of critical commentary has been directed at this approach, see, e.g., R.D. Mackay at 92, but the legislature has maintained it through recent revisions of the statute.
The concept of GAMI was introduced as part of the 1983 revision eliminating the traditional insanity defense:
“Guilty and mentally ill” is a deliberate variation of the “guilty but mentally ill” provision adopted by some states. The [Utah] legislative committee thought that the words “guilty but mentally ill” implied a causal connection between the mental illness and the crime. That implied connection, however, is inappropriate under the Utah statute because “guilty and mentally ill” focuses on the defendant’s state of mind at the time of sentencing, regardless of his state of mind at the time of the crime. Interview with Ronald N. Boyce, member of Legislature’s Task Force Committee, Salt Lake City, Utah (Sept. 9, 1983).
Utah Legislative Survey—1983, 1984 Utah L.Rev. 115, 156 n. 265 (citations omitted) (emphasis added).
Thus, as we said in State v. DePlonty, 749 P.2d 621, 626 (Utah 1987), “A judgment of guilty and mentally ill does not serve to exonerate or excuse the defendant; rather, the offender found guilty and mentally ill is held accountable for his criminal conduct, yet because of his mental illness, may need specialized treatment.” We went on to state:
A defendant who suffers from a mental disease or defect and, therefore is mentally ill as defined by Utah Code Ann. § 76-2-305 and is found to possess the state of mind necessary to commit the crime charged, despite his illness, should be found guilty and mentally ill as provided by the Act.
Id. at 627.
During Young’s trial, the prosecutor objected to the admission of any evidence regarding defendant’s mental illness. He stated, “[Tjhere are only two areas in which mental health opinions can be entered into evidence, one is on a plea of insanity, and two is intent to rely on diminished capacity, neither of which is present before the court.” Because the GAMI verdict does not relate to the defendant’s state of mind at the time of the crime, this position is wrong. The statute contemplates that evidence of mental illness not amounting to insanity or diminished capacity (now consisting of a “no mens rea due to mental illness” defense) is relevant to sentencing considerations. In this case, the evidence had additional relevance to rebut inferences regarding Young’s mental state raised by some of his own prior statements to the police and in the pro se answer he filed in the wrongful death case brought by the victim’s family.
The trial judge thus properly admitted the testimony of a neuro-psychologist who had examined Young and concluded that he suffers from “an organic brain' syndrome personality syndrome. What that really means is that he has diffuse brain damage.... [Tjhere’s a history of repeated head injuries as a child, some recent head injuries as an adult, in addition to numerous developmental problems that are noted in his medical record.” The psychologist opined that Young has a “mental disease or defect” that places him between the first and fifth percentile “when it comes to ... most areas of cognition [thinking, problem solving, reasoning, being able to employ good judgment, being able to learn from one’s mistakes], specifically his memory.”
Subsequent to the testimony summarized above, defense counsel requested an instruction and verdict form on “guilty and mentally ill.” The prosecution objected, arguing that the statutory structure was such that a GAMI verdict was a form of a “lesser and included” verdict relative to the plea of not guilty by reason of insanity, which defendant had not asserted. Again, the prosecution's position is inconsistent with the GAMI verdict’s focus on sentencing rather than on mens rea. Defense counsel clarified the issue by acknowledging error in the original version of the proposed instruction (which referred to defendant’s mental illness at the time of *385the commission of the offense) and by moving to amend the instruction “to state the jury should consider whether the defendant was guilty and mentally ill at the time of the trial. Because in our view, ... the guilty and mentally ill provisions of the statute are directed toward the issue of sentencing more than the issue of guilt or innocence.” The trial judge erroneously denied the motion and refused to give the instruction.
It may be true, as the trial judge observed, that the GAMI verdict does not “fit” our capital sentencing scheme very well because we separate the guilt and penalty phases. Nevertheless, the statute and rules do not exclude its use in a capital case. In practice, it is usually the state, rather than the defendant, that wants the perceived benefits of the availability of the GAMI verdict as an alternative to a verdict of NGBRI. Rule 21.5 accommodates the state’s position by mandating a GAMI instruction and verdict whenever the defendant has relied on a mental illness defense. The rule does not, however, preclude the defendant from benefitting, for different reasons, from the GAMI option when a mental illness defense has not been used. If the legislature had chosen the “guilty but mentally ill” version, which may imply a causal connection between the mental illness in question and the commission of the crime, the state’s position would be stronger, since the use of a mental illness defense would seem to be a necessary predicate for letting the jury assess the defendant’s mental health as of the time the crime was committed. Because in Utah the GAMI verdict is relevant solely for sentencing, neither logic nor the statute or rules dictate that the defendant assert a mental illness defense to the crime in order to receive whatever benefit the GAMI verdict may afford.
In practical effect, the average criminal defendant in a noncapital case tried to a jury is unlikely to perceive much benefit in asking for a GAMI verdict in the absence of a mental illness defense.6 There is no more real guarantee of mental health assessment and treatment for a defendant found GAMI than for one found guilty who then raises his or her mental illness as an issue at sentencing. Both classes of defendants are eligible for testing and alternative incarceration in mental facilities; the only difference is that the assessment process is automatically triggered for the GAMI defendant, whereas the ordinary “guilty” defendant will have to raise mental illness sua sponte and perhaps carry some additional burden in producing evidence.
In a capital case like the present one, significant psychological advantage may inure to a defendant who can argue to the jury in the guilt phase that a GAMI verdict is justified. Although the law provides that a defendant found GAMI may be sentenced to any punishment lawful for one found simply guilty, including death, a capital defendant who has been officially la-belled “mentally ill” by the fact finder has possibly achieved a tactical advantage as the penalty phase begins, because the jury has found a significant mitigating factor. This tactical advantage, of course, may do no more than counterbalance the similar advantage enjoyed by the state, which has already accomplished a jury finding of at least one aggravating circumstance, but it may be an advantage nonetheless. As discussed above, there is no basis in the statute, our rules, or logic to deny any defendant an instruction on GAMI where the evidence supports it. Furthermore, fundamental fairness requires that such an instruction be available to a capital defendant, for whom sentencing considerations are paramount once a verdict of guilty has been reached.
For the foregoing reasons, I would hold that the trial judge committed reversible error in refusing to instruct the jury that a GAMI verdict was an available option. There can be no question of harmless *386error when the court’s decision deprived the defendant of the possibility of a jury finding that he was mentally ill. The impact that such a verdict would have had on the jury’s penalty deliberations is, of course, incalculable, but as I have discussed more fully in the section on Young’s statutory right, to allocution (part II.A above), given the subjective quality of the moral judgment made by a capital sentencing jury, it cannot be argued that such a fact would be insignificant, especially as applied in the second phase of the Wood analysis. See State v. Wood, 648 P.2d at 71.
Y. ADMISSION OF EVIDENCE DURING GUILT PHASE
(lead opinion part VIII)
I concur in the result the lead opinion reaches on these issues. However, I disagree with its reason for upholding the trial court’s refusal to permit Young’s attorney to cross-examine witness Angela Johnson regarding the victim’s prior sexual activity. The lead opinion asserts that evidence of the victim’s prior sexual activity was irrelevant because Young “admitted raping the victim” and “did not seriously dispute the issue of consent.” Yet a careful appraisal of Young’s entire confession, in light of his obvious mental and emotional deficiencies and the physical evidence, reveals striking ambiguities in the evidence about the nature of defendant’s sexual contact with the victim. Furthermore, in opening argument Young’s counsel acknowledged the rape confession but stated that when the jury looked beyond the words themselves they would be troubled about Young’s actual motivation and extent of control. Young’s attorneys therefore may have had legitimate reasons to contest the issue of consent. The impact of Johnson’s testimony in resolving these ambiguities might have been altered if the court had permitted defendant to rebut her claims of the victim’s “high standards.”
Nevertheless, in my view the trial court acted within its discretion in refusing this line of questioning for two reasons. First, although Young argues that Johnson suggested in direct testimony that the victim “would not voluntarily engage in sexual relations,” in fact the witness never made or implied this claim on direct examination by the prosecution; only in response to defense cross-examination did the witness claim that the victim “had very high standards” and was not promiscuous and would not become “romantically involved with someone.” On direct examination, the witness testified regarding only her appraisal of the victim’s romantic interest in this particular defendant.7 Second, the proffered testimony would have addressed the victim’s relationship with her brother-in-law. Because the direct examination testimony related only to the witness’s assessment of the victim’s specific interest in Young, testimony about the victim’s relationship with her brother-in-law would have had little relevance. I accordingly concur in the result on this point.
VI. DEATH QUALIFICATION OF THE JURY
(lead opinion part IV)
Young appeals from the trial court’s denial of his pretrial “Motion to Preclude Death Qualification of the Jury Venire.” The United States Supreme Court has held that the practice of “death qualification,” or excluding from capital juries an entire class of potential jurors who share attitudes in opposition to the death penalty, does not violate the federal constitution. Lockhart v. McCree, 476 U.S. 162, 106 S.Ct. 1758, 90 L.Ed.2d 137 (1986). As the lead opinion acknowledges, however, this court has not previously analyzed whether death qualification violates the Utah Constitution. The lead opinion then summarily concludes that prior decisions of this court analyzing death qualification under the federal constitution suffice to support the *387holding that death qualification also comports with the Utah Constitution. Because I believe both that our prior analyses were flawed and that we should not tether the meaning of the Utah Constitution to the construction the United States Supreme Court gives the federal constitution, I would give the Utah constitutional issue full and independent treatment.
To defend its decision, the lead opinion relies on the rationale we articulated in State v. Moore, 697 P.2d 233 (Utah 1985), one of our previous analyses of death qualification. I believe we erred in both aspects of our Moore rationale and that the lead opinion today repeats the same two errors. In Moore, we faced the claim that by excluding from the jury all potential jurors who at the penalty phase could never vote to impose the death penalty (Witherspoon-excludables, or “WEs”),8 death qualification produced a jury disposed to convict at the guilt phase. In the face of then-existing empirical support for this claim, we responded that even if the claim were true, Utah counterbalanced this effect at least in part by also excluding from capital juries those individuals who would always vote to impose the death penalty upon a defendant convicted of first degree murder (automatic death penalty jurors, or “ADPs”). Id. at 237. Our first error was in making a factual assumption (about the counterbalancing effect of ADPs) without the benefit of any empirical evidence. In fact, developments in empirical research since Moore continue to show a statistically significant likelihood that death qualification frequently will produce conviction-prone juries regardless of any counterbalancing effect of excluding ADPs. Contrary to this research, the lead opinion persists in the opposite assumption. . In subsections A, B, and C, below, I discuss the empirical evidence in detail.
The second error in the Moore opinion was our assertion that legislative policy provided an additional reason for upholding Utah’s death-qualification process. We stated that “the Legislature has established capital punishment as one of the options in a first degree murder case. Clearly the legislative policy would be undermined if jurors were allowed to sit who by conscience could never impose the death penalty.” Id. I submit that we missed the point. Those challenging the death-qualifi*388cation process, including Young, do not argue that they have a right to present themselves for sentencing before a jury that is not death qualified; rather, they claim a right to have their guilt or innocence determined by a jury not death qualified.9 Properly analyzed, the question in this case is whether under Utah’s capital punishment scheme (which explicitly contemplates the use of a separate sentencing jury in some circumstances, see Utah Code Ann. § 76-3-207(1)), Young and all capital defendants have a constitutional right to a separate sentencing jury. This issue simply does not implicate the policy concern we expressed in Moore. The lead opinion similarly strikes far from the mark when it writes of the “significant state interest in seating jurors willing to abide by the law and follow their oath as jurors.” Although the stated interest undoubtedly is significant, use of separate juries would in no way compromise this interest.
The one plausible significant interest the state does have in death qualifying a capital jury prior to commencing the guilt phase is an interest in avoiding the judicial inefficiency of empaneling two juries. But as I indicate in subsection D below, the state could minimize this inefficiency in several ways, and the remaining state interest—marginal improvements in judicial efficiency—cannot legitimately be called “significant" when compared to the constitutional rights at risk.
Accordingly, I believe the death-qualification issue merits full consideration. I first discuss the potential biasing effect of excluding certain fair and impartial jurors— WEs—from guilt phase juries solely because of their attitudes toward capital punishment. I then examine the complete failure of the accompanying exclusion of ADPs to counterbalance this effect. Finally, I discuss the prejudicial impact of the process of death qualification itself on those jurors who survive the process.
A. Exclusion of WEs
None of our prior opinions regarding death qualification actually address the empirical data establishing the conviction-prone nature of the class of jurors whose attitudes make them “death qualified.” Even in Moore, we merely posited that this data “may well be correct” and then rejected the claim on other grounds. 697 P.2d at 237. Therefore, I deem it appropriate to summarize briefly the accumulated evidence, which now includes over a dozen additional studies not before us in Moore.10
*389For over twenty-five years, empiricists have studied the correlation between an individual’s disposition to impose capital punishment and that individual’s attitudes toward criminal defendants and the criminal justice system. As Justice Marshall observed, “The chief strength of [these studies] lies in the essential unanimity of the results obtained by researchers using diverse subjects and varied methodologies.” Lockhart v. McCree, 476 U.S. 162, 189, 106 S.Ct. 1758, 1773, 90 L.Ed.2d 137 (1986) (Marshall, J., dissenting). Describing these results, Justice Marshall wrote:
The data strongly suggest that death qualification excludes a significantly large subset—at least 11% to 17%—of potential jurors who could be impartial during the guilt phase of trial. Among the members of this excludable class are a disproportionate number of blacks and women.
The perspectives on the criminal justice system of jurors who survive death qualification are systematically different from those of the excluded jurors. Death-qualified jurors are, for example, more likely to believe that a defendant’s failure to testify is indicative of his guilt, more hostile to the insanity defense, more mistrustful of defense attorneys, and less concerned about the danger of erroneous convictions. This pro-prosecution bias is reflected in the greater readiness of death-qualified jurors to convict or to convict on more serious charges. And, finally, the very process of death qualification—which focuses attention on the death penalty before the trial has even begun—has been found to predispose the jurors that survive it to believe that the defendant is guilty.
Id. at 187-88, 106 S.Ct. at 1772-73 (Marshall, J., dissenting) (citations omitted). In essence, some two dozen studies suggest that jurors who withstand death qualification tend to be significantly less solicitous of a defendant’s due process rights and significantly more eager to convict. As one report stated, “From a social scientist’s viewpoint the empirical question [whether death-qualified juries are biased against the defendant on the issue of guilt] has been conclusively answered.” Rick Seltzer et al., The Effect of Death Qualification on the Propensity of Jurors to Convict: The Maryland Example, 29 How.L.J. 571, 581 (1986).
As Justice Marshall also suggested, “The true impact of death qualification on the fairness of a trial is likely even more devastating than the studies show.” McCree, 476 U.S. at 190, 106 S.Ct. at 1774 (Marshall, J., dissenting). This is so, Justice Marshall explained, for two reasons: First, under Wainwright v. Witt,11 courts may actually exclude for cause a broader class of persons than most of the studies assumed would be excluded under Witherspoon. *390Second, the prosecution’s exercise of peremptory challenges to excuse additional jurors not suitably committed to capital punishment will exclude greater numbers of those individuals statistically likely to possess heightened sensitivity to due process protections. Id. at 190-92, 106 S.Ct. at 1773-75 (Marshall, J., dissenting).
After extensively reviewing the evidence, at least three lower federal courts had concluded prior to Lockett that death qualification creates an unjustifiable risk of empaneling a conviction-prone jury. See Grigsby v. Mabry, 758 F.2d 226, 237 (8th Cir.1985), rev’d sub nom. Lockhart v. McCree, 476 U.S. 162, 106 S.Ct. 1758, 90 L.Ed.2d 137 (1986); Keeten v. Garrison, 578 F.Supp. 1164, 1167 (W.D.N.C.1984), rev’d, 742 F.2d 129 (4th Cir.1984); Grigsby v. Mabry, 569 F.Supp. 1273 (E.D.Ark.1983), aff'd, 758 F.2d 226 (8th Cir.1985), rev’d sub nom. Lockhart v. McCree, 476 U.S. 162, 106 S.Ct. 1758, 90 L.Ed.2d 137 (1986). Unfortunately, despite Justice Marshall’s lucid dissent in McCree, the majority of the Supreme Court in that case dismissed the studies, holding that even if the empirical conclusions were correct, the resulting conviction-prone nature of the death-qualified jury did not violate the federal constitution. 476 U.S. at 173, 106 S.Ct. at 1764-65.12 The Court explained:
[T]he Constitution presupposes that a jury selected from a fair cross section of the community is impartial, regardless of the mix of individual viewpoints actually represented on the jury, so long as the jurors can conscientiously and properly carry out their sworn duty to apply the law to the facts of the particular case.
Id. at 184, 106 S.Ct. at 1770-71.
McCree has settled the federal claim for now, but I do not find McCree persuasive with respect to the Utah constitutional issue. Nor am I the first state jurist to believe that state constitutions may still prohibit death qualification of the guilt phase jury. In a 1987 New Jersey Supreme Court case, Justice Handler would have held that death qualification violated state constitutional guarantees of a fair and impartial jury. State v. Ramseur, 106 N.J. 123, 524 A.2d 188, 348 (1987) (Handler, J., dissenting). Justice Handler reviewed the empirical evidence regarding the conviction-prone nature of these juries and then pointedly challenged his colleagues: “The question, ultimately, is whether the state’s putative interest in ‘neutrality’ on the issue of penalty may be vindicated at the expense of the defendant’s interest in a fair trial if both interests may be accommodated by using different procedures.” Id. at 347-48. One of his colleagues, although not ready to hold that death qualification violated the state constitution, insisted that “in the exercise of our judicial supervision over the criminal justice system in New Jersey,” the New Jersey court should prohibit death qualification because it “is inconsistent with New Jersey's traditional sense of fairness and justice.” Id. at 295 (O’Hern, J., concurring in the result). One year later, Justice Handler added: “While no one insists that procedure can be made perfect, in no other context has this Court accepted the proposition that mere prosecu-torial convenience—or any state interest— justifies procedures that render the jury ‘somewhat more “conviction prone.”’” State v. Bey, 112 N.J. 123, 548 A.2d 887, 923 (1988) (Handler, J., dissenting) (quoting McCree, 476 U.S. at 173, 106 S.Ct. at 1764-65). Justice Handler continues to become more adamant as the evidence accumulates. See State v. Marshall, 123 N.J. 1, 586 A.2d 85, 199-204 (1991) (Handler, J., dissenting); State v. Hunt, 115 N.J. 330, 558 A.2d 1259, 1289-97 (1989) (Handler, J., dissenting).
We must examine closely the cost of ignoring the implications of the empirical data. In Witherspoon, the United States *391Supreme Court, restating the obvious, wrote that “a State may not entrust the determination of whether a [defendant] is innocent or guilty to a tribunal ‘organized to convict.’” 391 U.S. at 521, 88 S.Ct. at 1776 (quoting Fay v. New York, 332 U.S. 261, 294, 67 S.Ct. 1613, 1630, 91 L.Ed. 2043 (1947)). Yet the lead opinion in this case allows precisely this result. Young has made a prima facie showing that bias probably will result from death qualifying the guilt phase jury. Until the State rebuts this showing, we must conclude that by excluding an entire class of impartial jurors from ever participating in guilt determination in capital trials (and thereby restricting the class of jurors eligible for capital jury service to a group who, to a statistically significant degree, demonstrate attitudes accurately labeled “conviction-prone”), the State has violated Young’s right to an impartial jury, without a compelling reason for doing so. Given the data, requiring defendants to endure death qualification disadvantages them in an unconstitutional manner. Even if the data turn out in the future to be flawed, abandoning death qualification for guilt phase jurors would in no way disadvantage the state, while it would still protect a capital defendant’s rights. We therefore distort justice when we insist that capital defendants establish their empirical claim more convincingly before we will dispense with death qualification. See Stephen Gillers, Proving the Prejudice of Death-Qualified Juries After Adams v. Texas, 47 U.Pitt.L.Rev. 219, 240-42 (1985).
B. Exclusion of ADPs
To the best of my knowledge, no empirical data support the proposition, first propounded in Hovey v. Superior Court, 28 Cal.3d 1, 168 Cal.Rptr. 128, 170-71 & n. 110, 616 P.2d 1301, 1343 & n. 110 (1980), and then adopted by this court in Moore, that excluding ADPs from death-qualified juries actually counterbalances the risk that a death-qualified jury will be conviction-prone. Rather, recent analyses suggest that these juries still will be conviction-prone. Empirically, we thus face a different set of facts than we did in Moore. Before discussing the empirical findings, however, I first note my apprehension, based on a review of the jury voir dire in this case, that when empaneling a capital jury, Utah trial courts do not always engage in the proper counterbalancing of excluding ADPs that the Moore opinion assumed they would.
The lead opinion cites State v. Schreuder, 726 P.2d 1215, 1225-26 (Utah 1986), to support the proposition that “[t]he Utah practice of death qualification excludes not only those jurors who would never vote to impose the death penalty, but also those jurors who would always vote to impose the death penalty upon a finding of first degree murder.” Unfortunately, in this case I believe the trial court did not satisfy SchreudeYs requirement that it “also ask prospective jurors if they would automatically vote for the death penalty upon a conviction of first degree murder.” 13 Id. In questioning the general panel, the trial court did seek to identify WEs, asking for a show of hands from “any of you who are so unalterably opposed to the imposition of the death penalty that you could not under any set of circumstances ... return a verdict imposing the death penalty.” But when questioning the general panel, the court did not ask the correlative question of whether any of them would automatical*392ly vote for death upon conviction. Instead, all questioning designed to identify ADPs came in the sequestered voir dire of each individual juror that followed.
Moreover, explicit ADP questioning usually came from defense counsel,14 not the court. This was so despite both the trial court’s obligation to examine all jurors uniformly on their ADP status and its consistent practice of supplementing the WE questioning of the general panel with additional WE questioning in the individual voir dire.15 Furthermore, to the extent the trial court did engage in ADP questioning in the individual voir dire, its questioning was inconsistent. For example, when questioning juror Cole, the court did not inquire whether Cole would automatically vote for the death penalty upon conviction of first degree murder. When examining juror Bau-man, the court did not make this inquiry until it sought to clarify a question posed by defense counsel. The court then said, “[The question] is whether or not you would feel that because there had been a homicide, and because it had been intentional, just that alone would make you feel that there must be the death penalty?” Of course, intentional homicides may be other than first degree, so even this “clarification” did not address the proper issue. When questioning juror Petrogeorge, the court explicitly asked, “If the facts and law of this case told you that the appropriate penalty to impose was life imprisonment as opposed to the death penalty, ... would you be able to vote for life imprisonment?” The court asked juror Radon a similar question. These questions come closer to the required inquiry, although by prefacing the question with a statement of what the facts and law suggest, the court substantially undermined the ability of the question to reveal a juror’s potential biases.
To empanel a capital jury properly under our Moore and Schreuder standards, I believe that when the trial court poses a question to the general panel designed to identify WEs, the court also should ask the general panel an ADP question. A proper question (perhaps after explaining the intentional, aggravated nature of first degree murder) would be along the following lines: “Are there any of you who, because of your beliefs in favor of the death penalty, would always vote to impose capital punishment upon every defendant convicted of an intentional, aggravated murder?” More importantly, whether or not the court asks these questions of the general panel, the court should always make this inquiry in the individual voir dire. Furthermore, the court should preface its ADP questions with no more and no less of an explanation of capital sentencing than it uses to preface its WE questions.
In this case, I am disturbed both by the trial court’s failure to ask an ADP question of the general panel as a counterpoint to its WE question and by its failure to ask a uniform ADP question in the individual voir dire of all jurors. I must conclude that the counterbalancing of excluding ADPs, which we presumed in Schreuder and Moore to be occurring, is not in fact always taking place.
Even assuming, however, that the Utah trial courts are properly excluding ADPs from capital juries, the data from several recent studies demonstrate a statistically significant likelihood that death qualification frequently will prejudice juries against the defendant. As one scholar stated, these studies show that when ADPs as well as WEs are excluded, nonetheless, “the procedure of death qualification biases the jury pool against the defense.” Joseph B. *393Kadane, After Hovey: A Note on Taking Account of the Automatic Death Penalty Jurors, 8 L. & Hum.Behav. 115, 119 (1984). The federal district court in Grigsby v. Mabry reviewed several ADP studies, concluding that even after excluding ADPs “the guilt-proneness of the resulting jury remains.” 569 F.Supp. at 1305-08; cf. People v. Middleton, 244 Cal.Rptr. 378, 391 n. 14, 392 (Cal.Ct.App.1988) (describing results of four ADP studies but holding ADP question “still open” in California). I therefore conclude that we erred when we presumed in Moore that excluding ADPs adequately counterbalanced the apparent conviction-proneness of death-qualified juries.
To counter persistent skepticism of these empirical results, I note only that even ignoring these findings, no empirical or intuitive basis exists for concluding that excluding ADPs will resurrect the defendant’s right to a jury drawn from a pool that constitutes a representative cross-section of the community. A criminal defendant is entitled to trial by a jury of “equals and neighbors, indifferently chosen.” Duncan v. Louisiana, 391 U.S. 145, 152, 88 S.Ct. 1444, 1449, 20 L.Ed.2d 491 (1968) (quoting 4 W. Blackstone, Commentaries on the Laws of England, 349-50 (Cooley ed. 1899)); accord Duren v. Missouri, 439 U.S. 357, 99 S.Ct. 664, 58 L.Ed.2d 579 (1979). A jury from which ADPs and WEs have been excluded is hot indifferently chosen. ADPs and WEs are excluded on the basis of characteristics that bear no relationship to their ability to participate impartially in the trial of a defendant’s guilt, and their exclusion compromises the representative quality of the jury. As we said in State v. Ball, excluding all jurors of an identifiable group “ ‘upsets the demographic balance of the venire [and] frustrates the primary purpose of the representative cross-section requirement. That purpose ... is to achieve an overall impartiality by allowing the interaction of the diverse beliefs and values the jurors bring from their group experiences.’” 685 P.2d 1055, 1059 (Utah 1984) (quoting People v. Wheeler, 22 Cal.3d 258, 148 Cal.Rptr. 890, 902-03, 583 P.2d 748, 761 (1978)). I am confident that this court would not uphold the exclusion of all college graduates or all jurors over forty years of age, and yet the lead opinion’s decision today continues to exclude all ADP jurors as well as all WE jurors from ever sitting on a capital jury. Therefore, whether or not the empirical data show that excluding ADPs fails to mitigate the risk that the death-qualified jury will be conviction-prone, the practice only further erodes a criminal defendant’s right to a representative jury.16
C. Effects of Death-Qualification Process on Qualified Jurors
One year after Moore, I wrote the opinion of this court in a case rejecting the related argument, brought under the Utah Constitution, that regardless of the effect of excluding a particular type of juror, the questioning involved in the process of death qualification itself inevitably directs the empaneled jury’s attention prematurely to the penalty phase, thereby eroding the defendant’s presumption of innocence. See State v. Shaffer, 725 P.2d 1301, 1309-12 (Utah 1986). In Shaffer, we dismissed “the only major study that examines the effects of the death-qualification questioning on jurors” because of the “limited value to a single empirical study,” id. at 1311, and because “the side effects of death-qualification voir dire could be minimized by individual sequestered voir dire,” id. at 1312. As in Moore, we again made a factual assumption (about the impact of sequestered voir dire) without the benefit of any empirical support and in the face of uncontradicted, albeit limited, empirical data that demonstrated prejudicial side effects. Because of this unsupported assumption, I would revisit the Shaffer opinion today in response to Young’s argument: I would consider the *394combined effect on the conviction-proneness of death-qualified juries of voir dire questioning itself and the exclusion from the jury of WEs and ADPs.
The original study before us in Shaffer, now published, confirmed what is intuitively plausible: the process of death qualification, despite any admonitions from the court to the contrary, may lead jurors who survive it to expect to convict the defendant and to assume that the “real” issue will pertain to the penalty. See Craig Haney, On the Selection of Capital Juries: The Biasing Effects of the Death-Qualification Process, 8 L. & Hum.Behav. 121 (1984). In a separate report, Haney went on to consider the cumulative impact of this process effect in addition to the prejudice created by excluding potential jurors. He concluded that the process of subjecting death-qualified jurors to exhaustive questioning prior to trial about the circumstances under which they would sentence a convicted murderer to death compounds the conviction-proneness already created by excluding jurors not death qualified. See Craig Haney, Examining Death Qualification: Further Analysis of the Process Effect, 8 L. & Hum.Behav. 133, 151 (1984).
Furthermore, despite what I said for the court in Shaffer, neither intuitive nor empirical reason suggests that sequestered voir dire can or will eliminate or minimize this danger. In fact, it seems equally plausible that sequestered voir dire will increase the risk, because unsequestered voir dire may best assure that the jurors will have the opportunity to - internalize the hypothetical character of the death-qualification questioning. In unsequestered questioning, jurors will hear the cautionary instruction regarding the defendant’s presumption of innocence numerous times, as it is repeated with each juror, while sequestered jurors will hear it only once or twice. Sequestered jurors also will lack an opportunity to contemplate this instruction and the purpose of the death-qualification process generally over an extended and comparatively relaxed period while they watch the court examine other individuals.
D. Conclusion
Given the dual forms of conviction-proneness that death qualification causes, both from the impact of the process on those jurors who survive it and from the exclusion of an entire class of potential jurors, I would hold that death qualification prior to the guilt phase violates the Utah Constitution, article I, sections 7, 10, and 12. As the United States Supreme Court has indicated, “States are free to provide greater protections in their criminal justice system than the Federal Constitution requires.” California v. Ramos, 463 U.S. 992, 1014, 103 S.Ct. 3446, 3460, 77 L.Ed.2d 1171 (1983). I believe that death qualifying the jury prior to the guilt phase of a capital trial violates a defendant’s right to “trial by an impartial jury,” as guaranteed by article I,- section 12 and as reinforced by article I, section 10, which .requires that “[i]n capital cases the right of trial by jury shall remain inviolate.” I also believe the practice violates the guarantee of article I, section 7 of the Utah Constitution that “[n]o person shall be deprived of life ... without due process, of law.” Given Young’s prima facie showing of the systemic jury bias inherent in our current use of a unitary death-qualified jury for guilt determination, safeguarding his constitutional rights demands that in capital cases such as his we use a bifurcated jury process that employs a nondeath-qualified jury at the guilt phase. Utah already bifurcates most other aspects of the guilt and penalty phases of a capital trial, and there is no compelling justification for failing to take the additional step of employing separate juries to ensure the fairness of Utah’s capital proceedings.
In capital cases, the lengthiest portion of jury voir dire by far pertains to the process of death qualifying the jury, and bifurcation would permit trial courts to reserve this exhausting procedure until needed at the' sentencing phase.17 The need to em*395panel a second jury for sentencing might somewhat lengthen total trial time in those cases in which the defendant is convicted of first degree murder, but in cases of acquittal on the capital charge, avoiding the process of death qualification would substantially shorten total trial time. Many commentators and jurists have discussed the feasibility of administering a dual jury system of some sort. Possible means of employing a nondeath-qualified jury at the guilt phase include (1) seating at sentencing an entirely new death-qualified jury and then using stipulated summaries of the guilt phase testimony in order to avoid repeating the trial before the sentencing jury; (2) seating extra alternates at the guilt phase and then death qualifying the guilt jury when it reaches the penalty phase, replacing ADPs and WEs with qualified alternate jurors; (3) seating a separate death-qualified jury prior to the guilt phase to attend but not participate in the guilt phase and then to undertake sentencing if needed. See Lockhart v. McCree, 476 U.S. 162, 205, 106 S.Ct. 1758, 1781-82, 90 L.Ed.2d 137 (1986) (Marshall, J., dissenting); Grigsby v. Mabry, 758 F.2d 226, 243 (8th Cir.1985); Robert M. Berry, Remedies to the Dilemma of Death-Qualified Juries, 8 U.Ark. Little Rock L.J. 479 (1986); Donald P. Knudsen, Comment, Inequities and Abuses of Death Qualification: Causes and Cures, 32 S.D.L.Rev. 281, 292-98 (1987).
VII. DEFENDANT’S CHALLENGES FOR CAUSE
(lead opinion part VI)
Young has appealed the trial court’s refusal to grant four of his for-cause challenges of potential jurors. Regarding three of these jurors, I agree with the lead opinion that a “review of the entire voir dire record indicates that the trial court did not err in refusing to dismiss the challenged jurors for cause.” However, after a review of the record of the voir dire of juror Cole, I conclude that the trial court did commit clear error in failing to dismiss him.
In addition to Cole’s statements that the death penalty “is like atonement for the [defendant],” that it was less punishment than life in prison, and that he “strongly” favored the death penalty, Cole gave what I consider absolutely disqualifying answers to two voir dire questions from defense counsel, as follows:
Q. Let’s assume we have an intentional, somewhat malicious homicide. Would you be willing, or able, I think you’d probably be willing, but would you be able in all honesty to consider any information, good information about the person to persuade you that something other than death was appropriate?
A. No.
Q. So in your view, once you’ve made that threshold determination that this is a—what you’ve described as a malicious, intentional, planned intentional type of homicide, you wouldn’t be able to consider life?
A. No.
In the context of the entire voir dire record, these responses demonstrate this juror’s inability to weigh mitigating and aggravating circumstances pertaining to a defendant convicted of capital murder. Therefore, the court should have excused Cole for cause. See State v. Norton, 675 P.2d 577, 589 (Utah 1983), cert. denied, 466 U.S. 942, 104 S.Ct. 1923, 80 L.Ed.2d 470 (1984), overruled on other grounds by State v. Hansen, 734 P.2d 421 (Utah 1986).
The State attempted to rehabilitate this juror, as follows:
Q. Let me ask just one more question. You’ve expressed some fairly strong opinions regarding the death penalty when you feel it would be appropriate. If you came to the situation where you subjectively felt the death *396penalty was necessary, or appropriate, and at the same time you felt intellectually that the court’s instructions led you to doubt that under the court’s instructions it should be a death penalty, what would your decision be?
A. You said the court’s instructions that it should be a death penalty?
Q. It should not be a death penalty, but you felt it subjectively should.
A. Like I say, a law is a law, and the man sitting on the bench is the head honcho. I really feel that I’d have to go with the judgment of the Judge.
This colloquy failed to rehabilitate juror Cole for several reasons. First, it created a false dichotomy between the juror’s “subjective” and “intellectual” feelings regarding the appropriate penalty. Even when following correct instructions, a sentencing juror makes a subjective determination, as we made clear in State v. Wood, 648 P.2d 71, 84 (Utah), cert. denied, 459 U.S. 988, 103 S.Ct. 341, 74 L.Ed.2d 383 (1982). The proper dichotomy would have been between a decision made “because of a personal belief that all homicides should be punishable by death” and one “based on the result of the comparison [of aggravating and mitigating factors],” id. at 79; a juror acts subjectively, however, in making' either of these decisions.
Furthermore, the question as posed suggested to Cole that the State was asking what he as a juror would do if the court had concluded that the sentence “should not be a death penalty.” Although the question may not have meant to suggest this, Cole’s response indicates that he interpreted the question precisely this way, as though it were positing that the court had already expressed some “judgment” of its own about the proper penalty. Although in sentencing proceedings a court would never indicate its view of the proper judgment, lay jurors who, as a result of poor voir dire questions, misunderstand this will of course answer that they would follow the court’s instructions if the court has indicated the proper judgment.
Finally, as we have said, “A statement made by a prospective juror that he intends to be fair and impartial loses its meaning in light of other testimony or facts that suggest a bias.” State v. Hewitt, 689 P.2d 22, 26 (Utah 1984); cf. Morgan v. Illinois, — U.S. -, -, 112 S.Ct. 2222, 2233, 119 L.Ed.2d 492 (1992) (holding that general questions on voir dire about fairness and impartiality do not satisfy defendant’s right to identify ADP jurors, because “a juror could, in good conscience, swear to uphold the law and yet be unaware that maintaining such dogmatic beliefs about the death penalty would prevent him or her from doing so”). The voir dire of juror Cole as a whole establishes that Cole was not a qualified capital juror. Accordingly, I conclude that the trial court abused its discretion in denying Young’s challenge of juror Cole.
Although other aspects of this voir dire disturb me, as I have suggested in part VLB above, I have not found any other specific instances in which the trial court abused its discretion in refusing defendant’s challenges for cause of individual jurors.18
VIII. CONSTITUTIONALITY OF UTAH’S STATUTORY AGGRAVATING CIRCUMSTANCES SCHEME
(lead opinion part II)
Young filed a pretrial “Motion to Declare the Death Penalty Statute Unconstitutional *397on Its Face,” in which he argued that the extensive list of aggravating factors in the Utah statute fails to narrow the class of offenders eligible for the death penalty, thereby violating the state and federal constitutions. That motion was denied, and defendant has raised the issue on appeal. The lead opinion dismisses the issue without analysis, asserting that the court has previously considered such challenges. That assertion is incorrect. An examination of each of our death penalty cases since 1983, when Utah’s capital murder statute, Utah Code Ann. § 76-5-202(1), was significantly amended, discloses that we have never addressed the question defendant now raises. I conclude that the current Utah scheme of statutory aggravating circumstances fails to meet federal constitutional requirements and is invalid on its face. I also conclude that the current statute violates sections 7 and 9 of article I of the Utah Constitution.
A. Federal Law
In Furman v. Georgia, 408 U.S. 238, 92 S.Ct. 2726, 33 L.Ed.2d 346 (1972), the United States Supreme Court struck down the death penalty as it then existed in the states, referring frequently to the evils of unlimited sentencing discretion. Justice White, for example, said that the current death penalty practice used “no meaningful basis for distinguishing the few cases in which it is imposed from the many cases in which it is not.” Id. at 313, 92 S.Ct. at 2764 (White, J., concurring). Four years later, the Court issued a series of death penalty cases interpreting Furman and amplifying its analysis. In Gregg v. Georgia, 428 U.S. 153, 188, 96 S.Ct. 2909, 2932, 49 L.Ed.2d 859 (1976), a plurality of the Court interpreted Furman as mandating that “where discretion is afforded a sentencing body on a matter so grave as the determination of whether a human life should be taken or spared, that discretion must be suitably directed and limited so as to minimize the risk of wholly arbitrary and capricious action.” Justice White commented in Gregg on the function of the statutory aggravating factors enacted by the Georgia legislature after Furman:
The Georgia Legislature has plainly made an effort to guide the jury in the exercise of its discretion, while at the same time permitting the jury to dispense mercy on the basis of factors too intangible to write into a statute.... As the types of murders for which the death penalty may be imposed become more narrowly defined and are limited to those which are particularly serious or for which the death penalty is peculiarly appropriate as they are in Georgia by reason of the aggravating-circumstance requirement, it becomes reasonable to expect that juries—even given discretion not to impose the death penalty—will impose the death penalty in a substantial portion of the cases so defined. If they do, it can no longer be said that the penalty is being imposed wantonly or freakishly or so infrequently that it loses its usefulness as a sentencing device.
Id. at 222, 96 S.Ct. at 2947-48 (White, J., concurring) (emphasis added).
In 1980, in Godfrey v. Georgia, 446 U.S. 420, 100 S.Ct. 1759, 64 L.Ed.2d 398 (1980), Justice Stewart, quoting Gregg, wrote, “[I]f a state wishes to authorize capital punishment it has a constitutional responsibility to ... define the crimes for which death may be the sentence in a way that obviates ‘standardless [sentencing] discretion.’” Id. at 428, 100 S.Ct. at 1764. Then, in 1983, the Court decided Zant v. Stephens, 462 U.S. 862, 103 S.Ct. 2733, 77 L.Ed.2d 235 (1983), and focused on the function of statutory aggravating circumstances:
To avoid this constitutional flaw [failure to channel the sentencing decision patterns of juries], an aggravating circumstance must genuinely narrow the class of persons eligible for the death penalty and must reasonably justify the imposition of a more severe sentence on the defendant compared to others found guilty of murder.
[[Image here]]
Our cases indicate, then, that statutory aggravating circumstances play a constitutionally necessary function at the *398stage of legislative definition: they circumscribe the class of persons eligible for the death penalty.
Id. at 877-78, 103 S.Ct. at 2742-43 (emphasis added) (footnote omitted).
The Supreme Court has consistently reaffirmed the constitutional necessity of limiting the pool of defendants who risk imposition of the death penalty. In McCleskey v. Kemp, 481 U.S. 279, 304, 107 S.Ct. 1756, 1773-74, 95 L.Ed.2d 262 (1987), the Court referred to “the carefully defined standards that must narrow a sentencer’s discretion to impose the death sentence” and elaborated on the narrowing process:
First, there is a required threshold below which the death penalty cannot be imposed. In this context, the State must establish rational criteria that narrow the decisionmaker’s judgment as to whether the circumstances of a particular defendant’s case meet the threshold_ Second, States cannot limit the sentencer’s consideration of any relevant circumstance that could cause it to decline to impose the penalty. In this respect, the State cannot channel the sentencer’s discretion, but must allow it to consider any relevant information offered by the defendant.
Id. at 305-06, 107 S.Ct. at 1774.
More recently,- Chief Justice Rehnquist, writing for the Court in Lowenfield v. Phelps, 484 U.S. 231, 108 S.Ct. 546, 98 L.Ed.2d 568 (1988), said:
To pass constitutional muster, a capital sentencing scheme must “genuinely narrow the class of persons eligible for the death penalty and must reasonably justify the imposition of a more severe sentence on the defendant compared to others found guilty of murder.” Zant v. Stephens, 462 U.S. 862, 877[, 103 S.Ct. 2733, 2742.]
[[Image here]]
The use of “aggravating circumstances” is not an end in itself, but a means of genuinely narrowing the class of death-eligible persons and thereby channeling the jury’s discretion....
... [T]he narrowing function required for a regime of capital punishment may be provided in either of these two ways: The legislature may itself narrow the definition of capital offenses, as Texas and Louisiana have done, so that the jury finding of guilt responds to this concern, or the legislature may more broadly define capital offenses and provide for narrowing by jury findings of aggravating circumstances at the penalty phase.
Id. 484 U.S. at 244, 246, 108 S.Ct. at 554, 555 (citations omitted).
The Utah statute purports to perform the narrowing function by the first method described in Lowenfield, namely, by limiting the statutory definition of capital homicide to intentional killings accompanied by at least one aggravating circumstance. The state must prove the intentional killing and the aggravating circumstance at the guilt phase of a capital trial. Thus, all defendants convicted of an intentional killing exacerbated by a statutory aggravating circumstance become death-eligible; they are in the pool of persons subject to the death penalty. During the penalty phase of the trial, the sentencer has very broad discretion to consider virtually any relevant factor in aggravation or mitigation, including but not limited to the specific factors contained in the statute. Pursuant to the standard this court enunciated in State v. Wood, 648 P.2d 71 (Utah 1981) (per cu-riam), the sentencer must balance aggravation and mitigation, be persuaded beyond a reasonable doubt that the former outweighs the latter, and be persuaded to the same degree that death is appropriate in the circumstances. Thus, the decision to impose death on a particular defendant is highly individualized and is characterized by broad discretion in the sentencer. It is essentia] under federal constitutional standards, therefore, that the process by which defendants become eligible for death “genuinely narrow[s] the class of persons eligible” and “reasonably justifies] the imposition of [a death sentence] on the defendant compared to others found guilty of murder.” Zant, 462 U.S. at 877, 103 S.Ct. at 2742. In an article entitled “The New Role of Statutory Aggravating Circumstances in American Death Penalty Law,” 22 Duq. *399L.Rev. 317 (1984) [hereinafter Ledewitz], Bruce S. Ledewitz summarized the post-Zant standard as follows, “Thus, there are two requirements for a valid statutory aggravating circumstance: first, it must limit the class of murderers numerically and, second, it must represent a ‘good reason’ for choosing this defendant to be eligible for death, rather than other defendants who are not eligible for death.” Id. at 351. Based on the federal case law, I conclude that similar requirements apply not only to a single aggravating circumstance, but also to a system of aggravating circumstances, and that Utah’s system does not qualify.
Utah’s statute contains seventeen separate circumstances that “aggravate” an intentional killing, making it eligible for the death penalty. Not only does this definitional scheme include more categories of death-eligible murders than that of any other state, but also it excludes almost no intentional killings at all. I have reviewed the aggravating circumstances used by the other thirty-six death-penalty states to narrow the class of death-eligible defendants, either by statutory definition (as occurs in Utah) or during the sentencing process or both. As one would expect, there is considerable overlap in the language and apparent intent of those statutes. My analysis, summarized in the appendix to this opinion, has identified twenty-two discrete aggravating circumstances extant in death penalty statutes nationally. Utah’s statute contains seventeen of them. Utah Code Ann. § 76-5-202(1). It excludes only the following specific circumstances identified by other states: homicides committed (1) in connection with treason; (2) against an unusually vulnerable victim (e.g., a person under 15, over 62, handicapped, disabled, pregnant, “defenseless”); (3) in a premeditated, cold-blooded fashion (“lying in wait,” “without any pretense of moral or legal justification”); (4) totally at random, with no motive whatsoever; and (5) with a motive predicated on a victim’s race, color, religion, nationality, etc. See Appendix. This last category might in fact be included within the meaning of Utah Code Ann. § 76-5-202(l)(f), which states, “The homicide was committed for pecuniary or other personal gain” (emphasis added), thus further reducing the already very short list of homicides that the Utah Legislature has not defined as death-eligible. Furthermore, the four or five aggravating circumstances not contained in Utah’s definition of capital homicide are- quite likely to be unnecessary to classify any particular homicide as capital, given the extensiveness of the remainder of the list. For example, any one of the excluded categories of killings (vulnerable victims, treason, etc.) might very well be included as capital crimes in Utah pursuant to another category of aggravation, such as killing with a bomb, killing to conceal an offense, killing in an atrocious or cruel manner, knowingly creating risk of death to others, killing for personal gain, and so on. See Utah Code Ann. § 76-5-202(l)(a) through (q).
In short, Utah’s statutory definition of capital homicide excludes so few categories and so few actual murders that it has in effect returned the state to where it was before Furman was decided; there is no meaningful narrowing of ■ the class of death-eligible murders pursuant to objective, rational standards. As Ledewitz points out, “A death penalty generally applicable to all Enmund-eligible killers [i.e., those with sufficient direct culpability, see Enmund v. Florida, 458 U.S. 782, 102 S.Ct. 3368, 73 L.Ed.2d 1140 (1982)] would risk not disproportion or excessiveness, but rather simple arbitrariness.” Ledewitz at 359 n. 158.
The requirement of one statutory aggravating circumstance may be understood as a way of reconciling nonstatutory aggravating circumstances with Furman v. Georgia. Furman invalidated state death penalty statutes because they were applied in an arbitrary way. Statutory aggravating circumstances as the sole justification for the death penalty were viewed by many states as a way to reduce this arbitrariness. The Court in Barclay and Stephens rejected this approach and approved nonstatutory aggravation. This means essentially that whatever strikes the sentencer as aggravating will justify the death penalty. *400Barclay, where the judge’s experiences in World War II led him to impose the death penalty, illustrates how this process can lead to divergent treatment of similar cases. To counteract this tendency, the Court has apparently fastened on to narrowing statutory aggravating circumstances to limit consideration by this potentially irrational sentencing system to a small number of truly deserving murderers.
Id. at 394 (emphasis added) (footnote omitted).
The foregoing analysis was written in 1984. No language in any United States Supreme Court opinion since then suggests that the Court is prepared to abandon the constitutional function of narrowing the class of death-eligible defendants performed by statutory aggravating factors. “[Tjhere is a required threshold below which the death penalty cannot be imposed.” McCleskey v. Kemp, 481 U.S. 279, 305, 107 S.Ct. 1756, 1774, 95 L.Ed.2d 262 (1987). “[Sjtatutory aggravating circumstances play a constitutionally necessary function at the stage of legislative definition: they circumscribe the class of persons eligible for the death penalty.” Lowenfield v. Phelps, 484 U.S. 231, 244, 108 S.Ct. 546, 554, 98 L.Ed.2d 568 (1988) (quoting Zant v. Stephens, 462 U.S. at 878, 103 S.Ct. at 2743). Unless the Court is now prepared to eliminate the constitutional requirement that statutory aggravating factors establish meaningful distinctions between murderers who are eligible for death and those who are not, I do not see how the Utah statutory scheme, which does not do so, can stand. Our statute is arbitrary on its face; it exposes the overwhelming majority of intentional killers to the death penalty and establishes no objective criteria for distinguishing those few “worst” crimes that the United States Supreme Court has acknowledged are. appropriate for the ultimate penalty. See Zant, 462 U.S. at 876-78 & n. 15, 103 S.Ct. at 2742-43 & n. 15; Gregg, 428 U.S. at 182-84, 96 S.Ct. at 2929-30; cf. State v. Pierre, 572 P.2d 1338, 1356 (Utah 1977) (upholding earlier version of Utah’s death penalty statute because it limited death to only “extreme and unusually serious and shocking crimes”), cert. denied, 439 U.S. 882, 99 S.Ct. 219, 58 L.Ed.2d 194 (1978).
This analysis is not undertaken as a mere “counting” exercise, as Justice Zimmerman’s opinion argues. It is rather an effort to identify the “known universe” of putatively death-worthy homicides, as reflected in the statutory schemes of the thirty-seven states that employ the death penalty. Justice Zimmerman does not suggest, and there is no reason to believe, that any type of homicide not currently included in this list should be reviewed as part of that universe. Thus, Justice Zimmerman’s opinion misses the point. He seems to be discussing the number of death-eligible murderers, rather than the selected class of murderers drawn from the potential universe. My concern is not with the numbers, but with the basis for the distinctions to be drawn between one murder and another. For example, Justice Zimmerman posits a scenario wherein all murders committed in Utah are factually of the same type (i.e., include the same aggravating factor), resulting in death eligibility in all cases. I detect no constitutional problem with such a scenario because the statutory scheme would, notwithstanding the bizarre and coincidental factual result, follow Zant’s mandate that Utah narrow the class of death-eligible murderers. The number of persons within the class is not the issue—the number of classes within the potential universe of murders is.
B. State Law: Article I, Sections 7 (Due Process) and 9 (Cruel and Unusual Punishment, Unnecessary Rigor)
The Utah Constitution’s Declaration of Rights contains a due process guarantee that is this court’s obligation and prerogative to construe. “Due process of law in each particular case means, such an exertion of the powers of government as the settled maxims of law permit and sanction, and under such safeguards for the protection of individual rights as those maxims prescribe for the class of cases to which the one in question belongs.” Thomas Cooley, Constitutional Limitations 741 (8th *401ed.1927). Although we rarely have discussed article I, section 7 without referring to federal due process jurisprudence, we have on occasion predicated holdings solely on state due process principles. For example, in State v. Copeland, 765 P.2d 1266, 1271-72 (Utah 1988), we held that subsections (c) and (d) of Utah Code Ann. § 77-35-21.5(4) violated article I, section 7 of the Utah Constitution because their operation was arbitrary and capricious and their content was not rationally related to the criminal sentencing process at issue. Likewise, we have reviewed statutory sentencing schemes and particular sentences under article I, section 9. See State v. Russell, 791 P.2d 188 (Utah 1990); State v. Gentry, 747 P.2d 1032 (Utah 1987); State v. Bishop, 717 P.2d 261 (Utah 1986). For a Utah due process analysis in the civil context, see Berry v. Beech Aircraft, 717 P.2d 670, 675-76 (Utah 1985).
I submit that this court should independently articulate the minimum standards that apply under the Utah Constitution for our death penalty statutes. Cf. California v. Ramos, 463 U.S. 992, 1014, 103 S.Ct. 3446, 3460, (1983) (“States are free to provide greater protections in their criminal justice system than the Federal Constitution requires.”). A comprehensive review of current state constitutional death penalty jurisprudence is found in James R. Acker & Elizabeth R. Walsh, Challenging the Death Penalty Under State Constitutions, 42 Vand.L.Rev. 1299 (1989) [hereinafter Acker & Walsh]. The authors comment on the differences in perspective and structural placement that affect legal development by state courts in this area:
State supreme courts may not be as willing to tolerate the structural and procedural deficiencies in capital punishment legislation as has a majority of the United States Supreme Court. They may be persuaded for a number of reasons that state constitutions permit or require them to diverge from the judicial gloss Supreme Court precedent attaches to the eighth amendment. The Supreme Court is constrained by principles of federalism, its duty of fixing minimal constitutional standards for fifty individual states with tremendously varied histories and cultures, and by the inhibiting effect that federal constitutional adjudication has upon the states’ abilities to adapt criminal laws and procedures to their unique needs. State supreme courts can be responsive to the values and traditions reflected in state constitutional provisions and to contemporary notions of fairness and decency in their states. Further, their decisions are not so wide reaching and typically not so immutable as the decisions of the Supreme Court.
Id. at 1361 (citations omitted).
As the foregoing article documents, a significant number of states have independently reviewed their death penalty statutes pursuant to the provisions of their state constitutions. Other opinions have been issued since the article was published in 1989. See, e.g., State v. Black, 815 S.W.2d 166 (Tenn.1991) (analyzing the constitutionality of the death penalty under numerous state provisions, including Tennessee’s cruel-and-unusual-punishment clause).
Courts in only two states, California and Massachusetts, have ever declared the death penalty unconstitutional per se, see People v. Anderson, 6 Cal.3d 628, 100 Cal.Rptr. 152, 493 P.2d 880, cert. denied, 406 U.S. 958, 92 S.Ct. 2060, 32 L.Ed.2d 344 (1972); District Atty. for Suffolk Dist. v. Watson, 381 Mass. 648, 411 N.E.2d 1274 (1980), and in both states voters have subsequently amended their constitutions to affirm that death is a permissible penalty. Even after the amendments, however, both courts have asserted that although the death penalty itself cannot violate the amended constitution on a per se basis, any given statutory scheme is subject to judicial review under traditional state protections for due process and against cruel and unusual punishment. See People v. Superior Court of Santa Clara County, 31 Cal.3d 797, 183 Cal.Rptr. 800, 805-06, 647 P.2d 76, 81-82 (1982) (death penalty statute invalidated on state due process grounds); Commonwealth v. Colon-Cruz, 393 Mass. 150, 470 N.E.2d 116, 120-23 (1984) (same).
*402As mentioned above, numerous state courts have reviewed the validity of death penalty statutes under various provisions of their state constitutions. See, e.g., People ex rel. Rice v. Cunningham, 61 Ill.2d 353, 336 N.E.2d 1 (1975) (sentence review method in statute violated direct appeal provision of state constitution); State v. Ramseur, 106 N.J. 123, 524 A.2d 188 (1987) (upholding death penalty statute under state constitutional provisions and analyzing challenges to jury qualification process and jury instructions pursuant to state constitution); State v. Rondeau, 89 N.M. 408, 553 P.2d 688 (1976) (capital punishment statute a deprivation of state due process provision); State v. Wagner, 305 Or. 115, 752 P.2d 1136 (1988) (upholding state death penalty statute under numerous provisions of Oregon constitution), vacated, 492 U.S. 914, 109 S.Ct. 3235, 106 L.Ed.2d 583 (1989). Also, nearly all the states that have considered due process and cruel-and-unusual-punishment challenges have predicated their holdings on their state provisions along with federal provisions, even when they have not undertaken independent analysis. Finally, for an extended discussion of case law regarding challenges to the death penalty predicated on state constitutional separation of powers doctrine and other state provisions without federal counterparts, see Acker & Walsh at 1356—60.
The present challenge to Utah’s capital punishment statute is a question of first impression. No other state has considered the constitutionality of a statute as broad and comprehensive in its definition of death-eligible murders. A similar conceptual challenge was leveled at the Oregon statute in Wagner, 752 P.2d at 1157-58, but there, the statute contained only ten aggravating factors, as compared to Utah's seventeen. Thus, we are in uncharted territory.
The Utah Constitution contains protections against unfairness and cruel and unusual punishment, requiring a scheme which insures that the death penalty is not applied in an arbitrary manner. Such a scheme, this court said in State v. Pierre, 572 P.2d 1338, 1356 (Utah 1977), cert. denied, 439 U.S. 882, 99 S.Ct. 219, 58 L.Ed.2d 194 (1978), must be “structured to provide reasonably that the unique and irretrievable sanction of death will be mandated by its provisions and processes only in extreme and unusually serious and shocking crimes and when mandated—as here—that the risk of discrimination, arbitrariness, caprice, and irrationality is reduced to a minimum.”
The aggravating circumstances in Utah’s statutory scheme ostensibly serve the constitutionally required function of reducing the risks of discrimination, arbitrariness, caprice, and irrationality. Given the nature of Utah’s bifurcated capital trial system, the presence of an aggravating factor, to be determined at the guilt phase of a capital trial, should ensure that only “extreme and unusually serious and shocking crimes” become eligible for the death penalty. The jury in our system has extremely broad discretion to consider during the penalty phase “any matter the court deems relevant to sentence,” whether in aggravation or mitigation, Utah Code Ann. § 76-3-207(2), and the balancing process we require during that phase, although highly individualized, does not contain any objective standards for the jury to follow. The result of the legislature’s overbroad definition of capital murder is a lack of any rational distinction between capital and noncapital homicide. With no objective means to distinguish “extreme and unusually serious and shocking” crimes, the exercise of prosecutorial discretion becomes arbitrary and capricious by definition. The state cannot distinguish between a Mark Hofmann (who intentionally killed two people with bombs that endangered others, apparently for pecuniary gain, thus qualifying as a capital murderer under four separate categories of aggravating circumstances, but was permitted to plead guilty to second degree murder, see State v. Hofmann, No. CR86-834 (Jan. 23, 1987)), and an ElRoy Tillman (who bludgeoned the boyfriend of his former girlfriend into unconsciousness, set a fire that asphyxiated him, and was charged and convicted of capital *403homicide and sentenced to death, see State v. Tillman, 750 P.2d 546 (Utah 1987)).
In State v. Wood, 648 P.2d 71 (Utah), cert. denied, 459 U.S. 988, 103 S.Ct. 341, 74 L.Ed.2d 383 (1982), this court vacated a death sentence, holding that “the sentencing process was flawed because the aggravating factor relied on [“ruthlessness and brutality”] was constitutionally impermissible in this case, since it describes all murders and. therefore fails to provide any guideline for channeling discretion.” Id. at 86. A similar defect now plagues the constitutionality of Utah’s entire statute, which defines nearly every intentional murder as death-eligible. The statute contains no principled way to distinguish death-eligible murderers from nondeath-eligible murderers. The structure of our statute thus permits caprice and arbitrariness to dictate who will be exempted and who will be charged under the statute and fails to isolate only the worst of crimes for death-eligibility. The statute violates the principles of fundamental fairness guaranteed by the due process clause of the state constitution and renders Utah’s death penalty scheme “cruel and unusual” within the meaning of article I, section 9. We should strike it on independent state grounds.
IX.. CONCLUSION
When a state undertakes the deprivation of life as a criminal sentence, it must meet the highest practicable standards of reliability and fairness. See Lockett v. Ohio, 438 U.S. 586, 605, 98 S.Ct. 2954, 2965, 57 L.Ed.2d 973 (1978); Woodson v. North Carolina, 428 U.S. 280, 305, 96 S.Ct. 2978, 2991-92, 49 L.Ed.2d 944 (1976); State v. Pierre, 572 P.2d 1338, 1356 (Utah 1977), cert. denied, 439 U.S. 882, 99 S.Ct. 219, 58 L.Ed.2d 194 (1978); see also Eddings v. Oklahoma, 455 U.S. 104, 117-18, 102 S.Ct. 869, 878-79, 71 L.Ed.2d 1 (1982) (O’Connor, J., concurring) (“Court has gone to extraordinary measures to ... guarantee, as much as is humanly possible, that the [death] sentence was not imposed out of whim, passion, prejudice, or mistake.”); Barefoot v. Estelle, 463 U.S. 880, 913-14, 103 S.Ct. 3383, 3405-06, 77 L.Ed.2d 1090 (1983) (Marshall, J., dissenting) (“Court has always insisted that the need for procedural safeguards is particularly great where life is at stake.”). Utah’s statute does not meet those standards, nor did its application in Young’s trial. I would strike the statute and reverse the judgment of conviction and death.
APPENDIX
Utah’s statutory capital sentencing scheme contains seventeen discrete aggravating factors, any one of which suffices to aggravate an intentional or knowing homicide to an offense punishable by death. Using the Utah scheme as a starting point, I have analyzed the aggravating factors used by the other thirty-six states currently permitting capital punishment. My analysis addresses only the statutory aggravating factors that render a defendant eligible for the death penalty after the state establishes at least one of them. I do not address the totality of aggravating factors the sentencer may .ultimately weigh in determining whether death is the appropriate sentence. I have concluded that Utah’s scheme omits only five additional factors (excluding Georgia’s anomalous category of “treason,” a category unrelated to whether a homicide has occurred)' found in any other state’s scheme and that no other single state’s scheme exposes a broader class of defendants to the death penalty than does Utah’s.
Five states, including Utah, determine death eligibility at the guilt stage by including specific aggravating factors in their statutory definitions of capital offenses. Most states, however, use codified variants of common law capital offenses at the guilt stage and then determine death eligibility at the sentencing stage by requiring the state to establish the existence of a statutory aggravating factor before the sentencer may contemplate whether to impose the death penalty. Seven other states use a combination approach: First, they significantly narrow their class of capital offenses at the guilt stage to those homicides accompanied by certain statutorily defined aggravating factors. Then at *404the sentencing phase, they employ a separate statutory list of aggravating factors, one of which must also be present before the defendant becomes death-eligible. (In most combination states, some overlap exists between the statutory aggravating factors included at the guilt phase and those included at the sentencing phase.)
In this appendix, I have matched the twenty-two total categories of aggravation I have identified with those states employing each category in their scheme. Part I briefly lists all twenty-two categories. Parts II and III cross-reference each category with each death penalty state: Part II is organized by category, and part III is organized by state. In part II, I have grouped the twenty-two total categories of aggravation into five types: those pertaining to the circumstances of the killing, those pertaining to the characteristics of the defendant, those pertaining to the characteristics of the victim, those pertaining to the method of the killing, and those pertaining to the motive for the killing. In both parts II and III, for simplicity in representing the seven combination states (those incorporating significant statutory aggravation both at the guilt stage and again at the sentencing stage), I have included only the aggravating factors relevant at the sentencing stage, thus under-representing the full extent to which these two-stage schemes narrow the class of death-eligible defendants. Regarding all thirty-seven states, although substantial similarity exists in the various aggravating factors used by each individual state, the overlap is not complete, and no single way exists to categorize the composite variety of aggravating factors; an analysis based on fewer or more categories of aggravation is possible, and I have taken some license in categorizing similar factors together. I am confident, however, that under any categorization, the Utah scheme is the most encompassing.
I. A SHORT LIST OF ALL TWENTY-TWO AGGRAVATING FACTORS
(1)Two or more persons killed (or attempted to be killed) in one episode or course of conduct.
(2) Homicide committed during certain enumerated felonies.
(3) Homicide committed during hijacking.
(4) Homicide committed while holding victim as shield, as hostage, or for ransom.
(5) Homicide committed in conjunction with drug offenses.
(6) Perpetrator confined in prison (or otherwise in state custody, including parole, or escaped) at time of offense.
(7) Perpetrator previously convicted of first or second degree murder or other violent felony (or otherwise likely to constitute continuing threat to society).
(8) Perpetrator under sentence of death or life in prison at time of homicide.
(9) Victim a public official, ex-official, or candidate (or news reporter) killed during or because of the exercise of official duty.
(10) Victim a law enforcement employee, firefighter, judicial officer, or corrections officer killed during or because of the exercise of official duty.
(11) Victim especially vulnerable, e.g., under fifteen years old, over sixty-two years old, disabled, pregnant.
(12) Perpetrator knowingly created great risk of death to persons other than victim.
(13) Homicide committed by means of destructive device or bomb.
(14) Homicide committed by means of poison or lethal substance.
(15) Homicide committed in especially heinous, atrocious, or cruel manner, including torture.
(16) Homicide committed after lying in wait or in cold, calculated fashion without any pretense of moral justification.
(17) Homicide committed to escape from custody or avoid arrest.
(18) Homicide committed for pecuniary or other personal gain.
(19) Contract murders (murder for hire).
(20) Homicide committed to eliminate prospective or former witness or to impede government function.
*405APPENDIX—Continued
(21) Homicide committed because of victim’s race, color, religion, nationality, or country of origin.
(22) Motiveless, random killing.
II. AGGRAVATING FACTORS AND THE STATES USING EACH
A. Circumstances of Offense Factors
(1) TWO OR MORE PERSONS KILLED (OR ATTEMPTED TO BE KILLED) IN ONE EPISODE OR COURSE OF CONDUCT: Utah Code Ann. § 76-5-202(l)(b) (Supp.1992); see Ariz.Rev.Stat.Ann. § 13-703(F)(8) (1989); Cal.Penal Code § 190.-2(a)(3) (West 1988 & Supp.1993); Del.Code Ann. tit. 11, § 4209(e)(l)k (1987 & Supp. 1992); Idaho Code § 19-2515(g)(2) (1987); Ill.Ann.Stat. ch. 38 ¶ 9-l(b)(3) (Smith-Hurd Supp.1992); Ind.Code Ann. § 35-50-2-9(b)(8) (Burns Supp.1992); Ky.Rev.Stat. Ann. § 532.025(2)(a)6 (Michie 1990); Md. Code Ann. art. 27, § 413(d)(9) (1987); Mass. Gen.Laws Ann. ch. 279, § 69(a)(8) (West Supp.1992); Mo.Rev.Stat. § 565.032.2(2) (Supp.1991); Mont.Code Ann. § 46-18-303(5) (1991); Neb.Rev.Stat. § 29-2523(l)(e) (1989); Ohio Rev.Code Ann. § 2929.04(A)(5) (Anderson 1987); Or.Rev.Stat. § ■ 163.-095(l)(d) (1991); S.C.Code Ann. § 16-3-20(C)(a)(9) (Law. Co-op. Supp.1992); Tenn. Code Ann. § 39-13-204(i)(12) (1991); Tex.Penal Code Ann. § 19.03(a)(6) (West 1989); Va.Code Ann. § 18.2-31.7 (Michie Supp.1992); Wash.Rev.Code § 10.95.020(8) (1992).
(2) HOMICIDE COMMITTED DURING CERTAIN ENUMERATED FELONIES: Utah Code Ann. § 76—5—202(1)(d) (Supp. 1992); see Ala.Code § 13A-5-49(4) (Supp. 1992); Cal.Penal Code § 190.2(a)(17) (West 1988 & Supp.1993); Colo.Rev.Stat. § 16-ll-103(5)(g) (Supp.1992); Conn.Gen.Stat. § 53a-46a(h)(l) (1993); Del.Code Ann. tit. II, § 4209(e)(l)j (1987 & Supp.1992); Fla. Stat. § 921.141(5)(d) (1991); Ga.Code Ann. § 17-10-30(b)(2) (1990); Idaho Code §§ 19-2515(g)(7), 18-4003(d) (1987 & Supp.1992); III.Ann.Stat. ch. 38, H 9-l(b)(6) (Smith-Hurd Supp.1992); Ind.Code. Ann. § 35-50-2-9(b)(1), (12) (Burns Supp.1992); Ky.Rev. Stat.Ann. § 532.025(2)(a)2 (Michie 1990); La.Code Crim.P.Ann. art. 905.4.A(1) (West Supp.1992); Md.Code Ann. art. 27, § 413(d)(10) (1987); Mass.Gen.Laws Ann. ch. 279, § 69(á)(10) (West Supp.1992); Miss. Code Ann. § 99-19-101(5)(d) (Supp.1992); Mo.Rev.Stat. § 565.032.2(11) (Supp.1991); Mont.Code Ann. § 46-18-303(7), (9) (1991); Nev.Rev.Stat. Ann. § 200.033.4 (Michie 1992); N.J.Stat.Ann. § 2C:ll-3 c(4)(g) (West Supp.1992); N.M.Stat.Ann. § 31-20A-5.B (Michie 1990); N.C.Gen.Stat. § 15A-2000(e)(5) (1988); Ohio Rev.Code Ann. § 2929.04(A)(7) (Anderson 1987); 42 Pa.Cons.Stat.Ann. § 9711(d)(6) (Supp.1992); S.C.Code Ann. § 16-3-20(C)(a)(l) (Law.Coop. Supp.1992); Tenn.Code Ann. § 39-13-204(i)(7) (1991); Tex.Penal Code Ann. § 19.-03(a)(2) (West 1989); Va.Code Ann. § 18.2-31.1, .4, .5, .9 (Michie Supp.1992); Wash. Rev.Code § 10.95.020(9) (1992); Wyo.Stat. § 6-2-102(h)(iv) (1988).
(3) HOMICIDE COMMITTED DURING HIJACKING: Utah Code Ann. § 76-5-202(l)(m) (Supp.1992); see Fla.Stat. § 921.141(5)(d) (1991); Ga.Code Ann. § 17-10-30(a) (1990)*; Ill.Ann.Stat. ch. 38, 119-1(b)(4) (Smith-Hurd Supp.1992); Miss.Code Ann. § 99-19-101(5)(d) (Supp.1992); Mo. Rev.Stat. § 565.032.2(14) (Supp.1991); N.C.Gen.Stat. § 15A-2000(e)(5) (1988); 42 Pa.Cons.Stat.Ann. § 9711(d)(4) (Supp.1992); Tenn.Code Ann. § 39-13-204(i)(7) (1991); Wyo.Stat. § 6-2-102(h)(iv) (1988).
(4) HOMICIDE COMMITTED WHILE HOLDING VICTIM AS SHIELD, AS HOSTAGE, OR FOR RANSOM: Utah Code Ann. § 76-5-202(1)(o) (Supp.1992); see Colo.Rev.Stat. § 16-ll-103(5)(d) (Supp. 1992); Del.Code Ann. tit. 11, § 4209(e)(l)e, f (1987 & 1992); Ind.Code Ann. § 35-50-2-9(b)(12)(B), (C) (Burns Supp.1992); Md.Code Ann. art. 27, § 413(d)(4) (1987); Mont.Code Ann. § 46-18-303(7) (1991); 42 Pa.Cons. Stat-Ann. § 9711(d)(3) (Supp.1992); Va. Code Ann. § 18.2-31.1, .8 (Michie Supp. 1992).
(5) HOMICIDE COMMITTED IN CONJUNCTION . WITH DRUG OFFENSES: See Ind.Code Ann. § 35-50-2-9(b)(l)(H) (Burns Supp.1992); N.H.Rev.Stat.Ann. § 630:5.VII(d) (Supp.1992); 42 Pa.Cons. *406StatAnn. § 9711(d)(14) (Supp.1992); S.D.Codified Laws Ann. § 23A-27A-1(10); Va.Code Ann. § 18.2-31.9 (Michie Supp. 1992).
B. Characteristics of Perpetrator Factors
(6) PERPETRATOR CONFINED IN PRISON (OR OTHERWISE IN STATE CUSTODY, INCLUDING PAROLE, OR ESCAPED) AT TIME OF OFFENSE: Utah Code Ann. § 76-5-202(l)(a) (Supp. 1992); see Ala.Code § 13A-5-49(l) (Supp. 1992); Ariz.Rev.Stat.Ann. § 13-703(F)(7) (1989); Ark.Code Ann. § 5-4-604(1), (2) (Michie Supp.1991); Colo.Rev.Stat. § 16-ll-103(5)(a) (Supp.1992); Del.Code Ann. tit. II, § 4209(e)(l)a (1987 & Supp.1992); Fla. Stat. § 921.141(5)(a) (1991); Ga.Code Ann. § 17-10-30(b)(9) (1990); Idaho Code §§ 19-2515(g)(7), 18-4003(e) (1987 & Supp.1992); III.Ann.Stat. ch. 38, ¶ 9-l(b)(2) (Smith-Hurd Supp.1992); Ind.Code Ann. § 35-50-2-9(b)(9) (Burns Supp.1992); Ky.Rev.Stat. Ann. § 532.025(2)(a)5 (Michie 1990); La. Code Crim.P.Ann. art. 905.4.A(6) (West Supp.1992); Md.Code Ann. art. 27, § 413(d)(2) (1987); Mass.Gen.Laws Ann. ch. 279, § 69(a)(2) (West Supp.1992); Miss. Code Ann. § 99-19-101(5)(a) (Supp.1992); Mo.Rev.Stat. § 565.032.2(9) (Supp.1991); Mont.Code Ann. § 46-18-303(1), (8) (1991); Nev.Rev.Stat.Ann. § 200.033.1 (Michie 1992); N.M.Stat.Ann. § 31-20A-5.C-.E (Mi-chie 1990); N.C.Gen.Stat. § 15A-2000(e)(l) (1988); Ohio Rev.Code Ann. § 2929.04(A)(4) (Anderson 1987); Okla.Stat.Ann. tit. 21, § 701.12.6 (West 1983); Or.Rev.Stat. § 163.095(2)(b), (f) (1991); S.D.Codified Laws Ann. § 23A-27A-1(8) (Supp.1992); Tenn.Code Ann. § 39—13—204(i)(8) (1991); Tex.Penal Code Ann. § 19.03(a)(5) (West 1989); Va.Code Ann. § 18.2-31.3 (Michie Supp.1992); Wash.Rev.Code § 10.95.020(2), (3) (1992); Wyo.Stat. § 6-2-102(h)(i) (1988).
(7) PERPETRATOR PREVIOUSLY CONVICTED OF FIRST OR SECOND DEGREE MURDER OR OTHER VIOLENT FELONY (OR OTHERWISE LIKELY TO CONSTITUTE CONTINUING THREAT TO SOCIETY): Utah Code Ann. § 76-5-202(l)(h) (Supp.1992); see Ala.Code § 13A-5-49(2) (Supp.1992); Ark.Code Ann. § 5-4-604(3) (Michie Supp.1991); Ariz.Rev. Stat.Ann. § 13-703(F)(2) (1989); Cal.Penal Code § 190.2(a)(2) (West 1988 & Supp. 1993); Colo.Rev.Stat. § 16-ll-103(5)(b) (Supp.1992); Conn.Gen.Stat. § 53a-46a(h)(2) (1993); Del.Code Ann. tit. 11, § 4209(e)(l)i (1987 & Supp.1992); Fla.Stat. § 921.141(5)(b) (1991); Ga.Code Ann. § 17-10-30(b)(l) (1990); Idaho Code § 19-2515(g)(1), (8) (1987); Ill.Ann.Stat. ch. 38, ¶ 9-l(b)(3) (Smith-Hurd Supp.1992); Ind. Code Ann. § 35-50-2-9(b)(7), (8) (Burns Supp.1992); Ky.Rev.Stat.Ann. § 532.-025(2)(a)l (Michie 1990); La.Code Crim. P.Ann. art. 905.4.A(3) (West Supp.1992); Mass.Gen.Laws Ann. ch. 279, § 69(a)(4) (West Supp.1992); Miss.Code Ann. § 99-19-101(5)(b) (Supp.1992); Mo.Rev.Stat. § 565.032.2(1) (Supp.1991); Mont.Code Ann. § 46-18-303(2) (1991); Neb.Rev.Stat. § 29-2523(l)(a) (1989); Nev.Rev.Stat.Ann. § 200.033.2 (Michie 1992); N.H.Rev.Stat. Ann. § 630:5.VII(c) (Supp.1992); N.J.Stat. Ann. § 2C:ll-3 c(4)(a) (West Supp.1992); N.C.Gen.Stat. § 15A-2000(e)(2), (3) (1988); Ohio Rev.Code Ann. § 2929.04(A)(5) (Anderson 1987); Okla.Stat.Ann. tit. 21, § 701.12.1, .7 (West 1983); Or.Rev.Stat. § 163.095(l)(c) (1991); 42 Pa.Cons.Stat. Ann. § 9711(d)(9) (Supp.1992); S.C.Code Ann. § 16-3-20(C)(a)(2) (Law. Co-op. Supp. 1992); S.D.Codified, Laws Ann. § 23A-27A-1(1) (Supp.1992); Tenn.Code Ann. § 39-13-204(i)(2) (1991); Wyo.Stat. § 6-2-102(h)(ii) (1988).
(8) PERPETRATOR UNDER SENTENCE OF DEATH OR LIFE IN PRISON AT TIME OF HOMICIDE: Utah Code Ann. § 76—5—202(l)(p) (Supp.1992); see Ariz.Rev.Stat.Ann. § 13-703(F)(1) (1989); Del.Code Ann. tit. 11, § 4209(e)(l)n (1987 & Supp.1992); Idaho Code §§ 19-2515(g)(7), 18-4003(c) (1987 & Supp.1992); Md.Code Ann. art. 27, § 413(d)(8) (1987); N.H.Rev. Stat.Ann. § 630:5.VII(b) (Supp.1992); 42 Pa.Cons.Stat.Ann. § 9711(d)(10) (Supp. 1992).
C. Characteristics of Victim Factors
(9) VICTIM A PUBLIC OFFICIAL, EX-OFFICIAL, OR CANDIDATE (OR NEWS *407REPORTER) KILLED DURING OR BECAUSE OF THE EXERCISE OF OFFICIAL DUTY: Utah Code Ann. § 76-5-202(l)(j) (Supp.1992); see Cal.Penal Code § 190.2(a)(13) (West 1988 & Supp.1992); Colo.Rev.Stat. § 16-ll-103(5)(c) (Supp. 1992); Ky.Rev.Stat.Ann. § 532.025(2)(a)5, (a)7 (Michie 1990); Mass.Gen.Laws Ann. ch. 279, §'69(a)(1), (3) (West Supp.1992); Mo. Rev.Stat. § 565.032.2(5) (Supp.1991); N.J.Stat.Ann. § 2C:ll-3 c(4)(h) (West Supp. 1992); Ohio Rev.Code Ann. § 2929.04(A)(1) (Anderson 1987); 42 Pa.Cons.Stat.Ann. § 9711(d)(1) (Supp.1992); Tenn.Code Ann. § 39—13—204(i)(l 1) (1991); Wash.Rev.Code § 10.95.020(10) (1992).
(10)VICTIM A LAW ENFORCEMENT EMPLOYEE, FIREFIGHTER, JUDICIAL OFFICER, OR CORRECTIONS OFFICER KILLED DURING OR BECAUSE OF THE EXERCISE OF OFFICIAL DUTY: Utah Code Ann. § 76—5—202(l)(k) (Supp. 1992); see Ariz.Rev.Stat.Ann. § 13-703(F)(10) (1989); Cal.Penal Code § 190.-2(a)(7)-(9), (11), (12) (West 1988 & Supp. 1993); Colo.Rev.Stat. § 16-ll-103(5)(c) (Supp.1992); Del.Code Ann. tit. 11, § 4209(e)(l)c, d (1987 & Supp.1992); Ga. Code Ann. § 17-10-30(b)(5), (8) (1990); Idaho Code § 19-2515(g)(9) (1987); Ill.Ann. Stat. ch. 38, ¶ 9-l(b)(l), (2) (Smith-Hurd Supp.1992); Ind.Code Ann. § 35-50-2-9(b)(6) (Burns Supp.1992); Ky.Rev.Stat. Ann. § 532.025(2)(a)5, (a)7 (Michie 1990); La.Code Crim.P.Ann. art. 905.4.A(2), (9) (West Supp.1992); Md.Code Ann. art. 27, § 413(d)(1) (1987); Mass.Gen.Laws Ann. ch. 279, § 69(a)(1), (3) (West Supp.1992); Mo. Rev.Stat. § 565.032.2(5), (8), (13) (Supp. Í991); MontCode Ann. § 46-18-303(6) (1991); Neb.Rev.Stat. § 29-2523(l)(g) (1989); Nev.Rev.Stat.Ann. § 200.033.7 (Mi-chie 1992); N.J.Stat.Ann. § 2C:ll-3 c(4)(h) (West Supp.1992); N.M.Stat.Ann. § 31-20A-5.A, .E (Michie 1990); N.C.Gen.Stat. § 15A-2000(e)(8) (1988); Ohio Rev.Code Ann. § 2929.04(A)(6) (Anderson 1987); Okla.Stat.Ann. tit. 21, § 701.12.8 (West 1983); Or.Rev.Stat. § 163.095(2)(a) (1991); 42 Pa.Cons.Stat.Ann. § 9711(d)(1) (Supp. 1992); S.C.Code Ann. § 16-3-20(C)(a)(5), (7) (Law.Co-op. Supp.1992); S.D.Codified Laws Ann. § 23A-27A-1(4), (7) (Supp.1992); Tenn.Code Ann. § 39—13—204(i)(9), (10) (1991); Tex.Penal Code Ann. § 19.03(a)(1) (West 1989); Va.Code Ann. § 18.2-31.6 (Mi-chie Supp.1992); Wash.Rev.Code § 10.95.-020(1), (6) (1992); Wyo.Stat. § 6-2-102(h)(viii) (1988).
(11) VICTIM ESPECIALLY VULNERABLE, E.G., UNDER FIFTEEN YEARS OLD, OVER SIXTY-TWO YEARS OLD, DISABLED, PREGNANT: See Ariz.Rév. Stat.Ann. § 13-703(F)(9) (1989); Del.Code Ann. tit. 11, § 4209(e)(l)p-r (1987 & Supp. 1992); Idaho Code §§ 19-2515(g)(7), 18-4003(d) (1987 & Supp.1992); Ill.Ann.Stat. ch. 38, 11 9-l(b)(7) (Smith-Hurd Supp.1992); Ind.Code Ann. § 35-50-2-9(b)(ll) (Burns Supp.1992); La.Code Crim.P.Ann. art. 905.-4.A(10) (West Supp.1992); Md.Code Ann. art. 27, § 413(d)(5) (1987); Mont. Code Ann. § 46-18-303(9) (1991); N.H.Rev.StatAnn. § 630:5.VII(g) (Supp.1992); S.C.Code Ann. § 16-3-20(C)(a)(10) (Law.Co-op. Supp.1992); Tenn.Code Ann. § 39—13—204(i)(l) (1991); Va.Code Ann. § 18.2-31.8 (Michie Supp. 1992).
D. Method of Killing Factors
(12) PERPETRATOR KNOWINGLY CREATED GREAT RISK OF DEATH TO PERSONS OTHER THAN VICTIM: Utah Code Ann. § 76-5-202(l)(c) (Supp.1992); see Ala.Code § 13A-5-49(3) (Supp.1992); Ariz.Rev.Stat.Ann. § 13-703(F)(3) (1989); Ark. Code Ann. § 5-4-604(4) (Michie Supp. 1991); Colo.Rev.Stat. § 16-ll-103(5)(i) (Supp.1992);' Conn.Gen.Stat. § 53a-46a(h)(3) ■ (1993); Fla.Stat. § 921.141(5)(c) (1991); Ga.Code Ann. § 17-10-30(b)(3) (1990); Idaho Code § 19-2515(g)(3), (6) (1987) (risk to many persons or utter disregard for human life); La.Code Crim.P.Ann. art. 905.4.A(4) (West Supp.1992); Mass. Gen.Laws Ann. ch. 279, § 69(a)(9) (West Supp.1992); Miss.Code Ann. § 99-19-101(5)(c) (Supp.1992); Mo.Rev.Stat. § 565.-032.2(3) (Supp.1991);' Neb.Rev.Stat. § 29-2523(l)(f) (1989); Nev.Rev.Stat.Ann. § 200.-033.3 (Michie 1992); N.H.Rev.Stat.Ann. § 630:5.VII(e) (Supp.1992); N.J.Stat.Ann. § 2C:ll-3 c(4)(b) (West Supp.1992); N.C.Gen.Stat. § 15A-2000(e)(10) (1988); Okla.Stat.Ann. tit. 21; § 701.12.-2 (West *4081983); 42 Pa.Cons.Stat.Ann. § 9711(d)(7) (Supp.1992); S.C.Code Ann. § 16-3-20(C)(a)(3) (Law.Co-op. Supp.1992); S.D.Codified Laws Ann. § 23A-27A-1(2) (Supp. 1992); Tenn.Code Ann. § 39-13-204(i)(3) (1991); Wyo.Stat. § 6-2-102(h)(iii) (1988).
(13) HOMICIDE COMMITTED BY MEANS OF DESTRUCTIVE DEVICE OR BOMB: Utah Code Ann. § 76-5-202(1X1) (Supp.1992); see Cal.Penal Code § 190.-2(a)(4), (6) (West 1988 & Supp.1993); Colo. Rev.Stat. § 16-ll-103(5)(f)(I)-(III) (Supp. 1992) (including assault weapons); Del. Code Ann. tit. 11, § 4209(e)(1)/ (1987 & Supp.1992); Fla.Stat. § 921.141(5)(d) (1991); Ga.Code Ann. § 17-10-30(b)(3) (1990); Ind. Code Ann. § 35-50-2-9(b)(2) (Burns Supp. 1992); Ky.Rev.Stat.Ann. § 532.025(2)(a)3 (Michie 1990); Mass.Gen.Laws Ann. ch. 279, § 69(a)(9) (West Supp.1992); Miss. Code Ann. § 99-19-101(5)(d) (Supp.1992); N.C.Gen.Stat. § 15A-2000(e)(5) (1988); Or. Rev.Stat. § 163.095(2)(c) (1991); Tenn.Code Ann. § 39—13—204(i)(7) (1991); Wyo.Stat. § 6-2-102(h)(iv) (1988).
(14) HOMICIDE COMMITTED BY MEANS OF POISON OR LETHAL SUBSTANCE: Utah Code Ann. § 76-5-202(l)(n) (Supp.1992); see Cal.Penal Code § 190.2(a)(19) (West 1988 & Supp.1993); Del.Code Ann. tit. 11, § 4209(e)(1)/ (1987 & Supp.1992); S.C.Code Ann. § 16-3-20(C)(a)(l)(f) (Law.Co-op. Supp.1992).
(15) HOMICIDE COMMITTED IN ESPECIALLY HEINOUS, ATROCIOUS, OR CRUEL MANNER, INCLUDING TORTURE: Utah Code Ann. § 76-5-202(l)(q) (Supp.1992); see Ala.Code § 13A-5-49(8) (Supp.1992); Ariz.Rev.Stat.Ann. § 13-703(F)(6) (1989); Ark.Code Ann. § 5-4-604(8) (Michie Supp.1991); Cal.Penal Code § 190.2(a)(14), (18) (West 1988 & Supp. 1993); Colo.Rev.Stat. § 16-11-103(5)0) (Supp.1992); Conn.Gen.Stat. § 53a-46a(h)(4) (1993); Del.Code Ann. tit. 11, § 4209(e)(1)/ (1987 & Supp.1992); Fla.Stat. § 921.141(5)(h) (1991); Ga.Code Ann. § 17-10-30(b)(7) (1990); Idaho Code § 19-2515(g)(5) (1987); Ind.Code Ann. § 35-50-2-9(b)(10) (Burns Supp.1992); La.Code Crim.P.Ann. art. 905.4.A(7) (West Supp. 1992); Mass.Gen.Laws Ann. ch. 279, § 69(a)(7) (West Supp.1992); Miss.Code Ann. § 99-19-101(5)(h) (Supp.1992); Mo. Rev.Stat. § 565.032.2(7) (Supp.1991); Mont. Code Ann. § 46-18-303(3) (1991); Neb.Rev. Stat. § 29-2523(l)(d) (1989); Nev.Rev.Stat. Ann. § 200.033.8 (Michie 1992); N.H.Rev. Stat.Ann. § 630:5.VII(h) (Supp.1992); NJ.Stat.Ann. § 2C:ll-3 c(4)(c) (West Supp. 1992); N.C.Gen.Stat. § 15A-2000(e)(9) (1988); Okla.Stat.Ann. tit. 21, § 701.12.4 (West 1983); Or.Rev.Stat. § 163.095(l)(e) (1991); 42 Pa.Cons.Stat.Ann. § 9711(d)(8) (Supp.1992); S.C.Code Ann. § 16—3— 20(C)(a)(l)(h) (Law.Co-op. Supp.1992); S.D.Codified Laws Ann. § 23A-27A-1(6) (Supp.1992); Tenn.Code Ann. § 39-13-204(i)(5) (1991); Wyo.Stat. § 6-2-102(h)(vii) (1988).
(16) HOMICIDE COMMITTED AFTER LYING IN WAIT OR IN COLD, CALCULATED FASHION, WITHOUT ANY PRETENSE OF MORAL JUSTIFICATION: See Cal.Penal Code § 190.2(a)(15) (West 1988 & Supp.1993); Colo.Rev.Stat. § 16-ll-103-(5)(f) (Supp.1992); Fla.Stat. § 921.-141(5)(i) (1991); Ind.Code Ann. § 35-50-2-9(b)(3) (Burns Supp.1992); Mont.Code Ann. § 46-18-303(4) (1991); N.H.Rev.Stat.Ann. § 630:5.VII(f) (Supp.1992).
E. Motive for Killing Factors
(17) HOMICIDE COMMITTED TO ESCAPE FROM CUSTODY OR AVOID ARREST: Utah Code Ann. § 76-5-202(l)(e) (Supp.1992); see Ala.Code § 13A-5-49(5) (Supp.1992); Ark.Code Ann. § 5-4-604(5) (Michie Supp.1991); Cal.Penal Code § 190.-2(a)(5) (West 1988 & Supp.1993); Colo.Rev. Stat. § 16-ll-103(5)(k) (Supp.1992); Del. Code Ann. tit. 11, § 4209(e)(l)b (1987 & Supp.1992); Fla.Stat. § 921.141(5)(e) (1991); Ga.Code Ann. § 17-10-30(b)(10) (1990); Idaho Code §§ 19-2515(g)(7), 18-4003(f) (1987 & Supp.1992); Md.Code Ann. art. 27, § 413(d)(3) (1987); Mass.Gen.Laws Ann. ch. 279, § 69(a)(6) (West Supp.1992); Miss. Code Ann. § 99-19-101(5)(e) (Supp.1992); Mo.Rev.Stat. § 565.032.2(10), (15), (16) (Supp.1991); Neb.Rev.Stat. § 29-2523(l)(b) (1989); Nev.Rev.Stat.Ann. § 200.033.5 (Michie 1992); N.H.Rev.Stat.Ann. § 630:5.VII(j) (Supp.1992); N.J.Stat.Ann. *409§ 2C:ll-3 c(4)(f) (West Supp.1992); N.M.Stat.Ann. § 31-20A-5.C (Michie 1990); N.C.Gen.Stat. § 15A-2000(e)(4) (1988); Ohio Rev.Code Ann. § 2929.04(A)(3) (Anderson 1987); Okla.Stat.Ann. tit. 21, § 701.12.5 (West 1983); Or.Rev.Stat. § 163.095(2)(e) (1991); S.D.Codified Laws Ann. § 23A-27A-1(9) (Supp.1992); Tenn. Code Ann. § 39-13-204(i)(6) (1991); Tex.Penal Code Ann. § 19.03(a)(4) (West 1989); Wash.Rev.Code § 10.95.020(7) (1992); Wyo. Stat. § 6-2-102(h)(v) (1988).
(18) HOMICIDE COMMITTED FOR PECUNIARY OR OTHER PERSONAL GAIN: Utah Code Ann. § 76-5-202(l)(f) (Supp.1992); see Ala.Code § 13A-5-49(6) (Supp.1992); Ariz.Rev.Stat.Ann. § 13-703(F)(5) (1989); Ark.Code Ann. § 5-4-604(6) (Michie Supp.1991); Cal.Penal Code § 190.2(a)(1) (West 1988 & Supp.1993); Colo.Rev.Stat. § 16-ll-103(5)(h) (Supp. 1992); Conn.Gen.Stat. § 53a-46a(h)(6) (1993); Del.Code Ann. tit. 11, § 4209(e)(l)o (1987 & Supp.1992); Fla.Stat. § 921.-141(5)(f) (1991); Ga.Code Ann. § 17-10-30(b)(4) (1990); Idaho Code § 19-2515(g)(4) (1987); Ky.Rev.Stat.Ann. § 532.025(2)(a)4 (Michie 1990); Miss.Code Ann. § 99-19-101(5)(f) (Supp.1992); Mo.Rev.Stat. § 565.-032.2(4) (Supp.1991); Neb.Rev.Stat. § 29-2523(l)(c) (1989); Nev.Rev.Stat.Ann. § 200.-033.6 (Michie 1991); N.H.Rev.StatAnn. § 630:5.VII(i) (Supp.1992); N.J.Stat.Ann. § 2C:ll-3 c(4)(d) (West Supp.1992); N.C.Gen.Stat. § 15A-2000(e)(6) (1988); Okla.Stat.Ann. tit. 21, § 701.12.3 (West 1983); S.C.Code Ann. § 16-3-20(C)(a)(4). (Law. Co-op. Supp.1992); S.D.Codified Laws Ann. § 23A-27A-1(3) (Supp.1992); Tenn.Code Ann. § 39-13-204(i)(4) (1991); Tex.Penal Code Ann. § 19.03(a)(3) (West 1989); Wyo.Stat. § 6-2-102(h)(vi) (1988).
(19) CONTRACT MURDERS (MURDER FOR HIRE): Utah Code Ann. § 76-5-202(l)(g) (Supp.1992); see Ariz.Rev.Stat. Ann. § 13-703(F)(4), (5) (1989); Colo.Rev. Stat. § 16-ll-103(5)(e) (Supp.1992); Conn. Gen.Stat. § 53a-46a(h)(5) (1993); Del.Code Ann. tit. 11, § 4209(e)(l)h, (l)m (1987 & Supp.1992); Ga.Code Ann. § 17-10-30(b)(6) (1990); Idaho Code § 19-2515(g)(4) (1987); Ill.Ann.Stat. ch. 38, 119-l(b)(5) (Smith-Hurd Supp.1992); Ind.Code Ann. § 35-50-2-9(b)(4), (5) (Burns Supp.1992); Ky.Rev.Stat. Ann. § 532.025(2)(a)4 (Michie 1990); La. Code Crim.P.Ann. art. 905.4.A(5) (West Supp.1992); Md.Code Ann. art. 27, § 413(d)(6), (7) (1987); Mass.Gen.Laws Ann. ch. 279, § 69(a)(5) (West Supp.1992); Mo. Rev.Stát. § 565.032.2(6) (Supp.1991); Neb. Rev.Stat. § 29-2523(l)(c) (1989); Nev.Rev. StatAnn. § 200.033.6 (Michie 1992); N.J.Stat.Ann. § 2C:ll-3 c(4)(e) (West Supp. 1992); N.M.Stat.Ann. § 31-20A-5.F (Michie 1990); Ohio Rev.Code Ann. § 2929.04(A)(2) (Anderson 1987); Okla.Stat.Ann. tit. 21, § 701.12.3 (West 1983); Or.Rev.Stat. § 163.095(l)(a), (b) (1991); 42 Pa.Cons.Stat. Ann. § 9711(d)(2) (Supp.1992); S.C.Code Ann. § 16-3-20(C)(a)(4), (6) (Law.Co-op. Supp.1992); S.D.Codified Laws Ann. § 23A-27A-1(3), (5) (Supp.1992); Tenn. Code Ann. § 39-13-204(i)(4) (1991); Tex.Penal Code Ann. § 19.03(a)(3) (West 1989); Va.Code Ann. § 18.2-31.2 (Michie Supp. 1992); Wash.Rev.Code § 10.95.020(4), (5) (1992).
(20) HOMICIDE COMMITTED ' TO ELIMINATE PROSPECTIVE OR FORMER WITNESS OR TO IMPEDE GOVERNMENT FUNCTION: Utah Code Ann. § 76-5-202(1)0) (Supp.1992); see Ala.Code § 13A-5-49(7) (Supp.1992); Ark.Code Ann. § 5-4-604(7) (Michie Supp.1991); Cal.Penal Code § 190.2(a)(10) (West 1988 & Supp. 1993); Colo.Rev.Stat. § 16-ll-103(5)(k) (S.upp.1992); Del.Code Ann. tit. 11, § 4209(e)(l)g (1987 & Supp.1992); Fla.Stat. § 921.141(5)(g) (1991); Idaho Code § 19-2515(g)(10) (1987); Ill.Ann.Stat. ch. 38 119-1(b)(8) (Smith-Hurd Supp.1992); Mass.Gen. Laws Ann. ch. 279, § 69(a)(3) (West Supp. 1992); Miss.Code Ann. § 99-19-101(5)(g) (Supp.1992); Mo.Rev.Stat. § 565.032.2(12), (16) (Supp.1991); Neb.Rev.Stat. § 29-2523(l)(h) (1989); N.M.Stat.Ann. § 31-20A-5.G (Michie 1990); N.C.Gen.Stat. § 15A-2000(e)(7), (8) (1988); Ohio Rev.Code Ann. § 2929.04(A)(8) (Anderson 1987); Or.Rev. Stat. § 163.095(2)(a)(E) (1991); 42 Pa.Cons. Stat.Ann. §' 9711(d)(5), (15) (Supp.1992); Wash.Rev.Code § 10.95.020(6) (1992).
(21) HOMICIDE COMMITTED BECAUSE OF VICTIM’S RACE, COLOR, RELIGION, NATIONALITY, OR COUNTRY *410OF ORIGIN: See Cal.Penal Code § 190.-2(a)(16) (West 1988 & Supp.1993).
(22) MOTIVELESS, RANDOM KILLING- See Nev.Rev.Stat.Ann. § 200.033.9 (Michie 1992).
III. STATES AND THE AGGRAVATING FACTORS EACH USES
[[Image here]]
*411ZIMMERMAN, Justice:I join the Chief Justice’s opinion in part and Justice Durham’s opinion in part. I would reverse the guilt determination and remand for a new trial and sentencing.
I join parts III, V, VII, VIII, IX, XI, XIII, XIV, XVI, XVII, XVIII, and XIX of the Chief Justice’s opinion. With respect to the remaining parts of the Chief Justice’s opinion, I concur in the result or dissent in whole or in part, all as specified below.
I
With respect to part II of the Chief Justice’s opinion, I join its result, dismissing defendant’s claims of unconstitutionality lodged against the Utah death penalty scheme. However, I cannot agree with the Chief Justice, that our prior decisions have disposed of defendant’s facial challenge to Utah’s death penalty statute. Defendant argues that the statute makes so many intentional murders eligible for the death penalty that in operation, the statute fails to limit the penalty’s availability to a distinct class of murders that are different in kind from those not punishable by death. Justice Durham correctly observes that this is a new claim and, like her, I would address it. Having done so, however, I am led to a conclusion different from Justice Durham’s: I think the Utah death penalty scheme is not invalid under either the state or the federal constitution.
Justice Durham reasons as follows in finding the statute unconstitutional. First, the constitutional function performed by aggravating circumstances is to “ ‘narrow the class of persons eligible for the death penalty and [the aggravating circumstances] must reasonably justify the imposition of a more severe sentence on the defendant compared to others found guilty of murder.’” (Quoting Zant v. Stephens, 462 U.S. 862, 877-78, 103 S.Ct. 2733, 2742-43, 77 L.Ed.2d 235 (1983)). Second, Utah’s statute lists seventeen separate circumstances that “aggravate” an intentional killing and make the perpetrator eligible for the death penalty. Third, the thirty-six other states with death penalty statutes have a cumulative total of twenty-two aggravating circumstances, but no single state statute includes as many of these as Utah’s. Based upon these facts, Justice Durham concludes that the effect of Utah’s extensive list of aggravating circumstances is that the statute “excludes almost no intentional killings at all” from the reach of the penalty, or stated differently, “excludes so few categories and so few actual murders that it has in effect returned the state to where it was before Furman was decided; there is no meaningful narrowing of the class of death-eligible murderers pursuant to objective, rational standards.” *412Therefore, she reasons, the Utah statute violates the federal constitution. By parallel reasoning, she concludes that it also violates the Utah Constitution.
I cannot agree with the assertions and assumptions underlying Justice Durham’s conclusion that the statute performs no meaningful narrowing of the class of murders that are exposed to the death penalty. There are two basic flaws in Justice Durham’s reasoning. First, she asserts as a factual matter that the Utah scheme excludes few actual murders from eligibility for the death penalty. Second, she assumes that simply by counting the aggravating circumstances listed in a statute and comparing the length of that list with similar lists from other states, one can conclude that the statute does not narrow the class of murders eligible for death and does not reasonably distinguish between those that are eligible for death and those that are not. I will address these points in order.
First, as a factual matter, we have before us no information about the number of murders committed in Utah over any relevant period, and we certainly do not know anything about whether the circumstances of those murders might bring some of them within the reach of the death penalty. Therefore, there is no empirical basis for Justice Durham’s assertion that, factually, few Utah murders are not eligible for the death penalty., And, I might add, while courts need not have proof positive of legislative facts upon which they base legal rulings, see Utah R.Evid. 201 & Advisory Committee Note; Fed.R.Evid. 201 & Advisory Committee Note, a healthy skepticism about the ability of judges to intuit the true state of the world around us should make us cautious in founding the law on intuition about empirical facts, especially facts so readily available as those upon which Justice Durham’s analysis is premised. See McCormick on Evidence § 331, at 929 (3d ed. 1984) (intellectual legitimacy of legal decisions founded on legislative facts turn on actual truth content of those facts); cf. State v. Long, 721 P.2d 483 (Utah 1986) (requirement of cautionary eyewitness instruction founded on nearly unanimous scientific judgments).
Second, even if we did know that most of the murders actually committed in Utah over some relevant period of time were eligible for the death penalty, that fact alone would not support the conclusion that the Utah statute does not pass constitutional muster. I find nothing in the United States Supreme Court’s death penalty cases holding that the federal constitution requires that only some small percentage of murders actually committed can be eligible for death, although one can find isolated bits of language that can be so construed. See, e.g., McCleskey v. Kemp, 481 U.S. 279, 304-05, 107 S.Ct. 1756, 1773-74, 95 L.Ed.2d 262 (1987); Zant v. Stephens, 462 U.S. 862, 877-78, 103 S.Ct. 2733, 2742-43, 77 L.Ed.2d 235 (1983); Gregg v. Georgia, 428 U.S. 153, 222, 96 S.Ct. 2909, 2947-48, 49 L.Ed.2d 859 (1976) (White, J., concurring). In my view, the principle the Court’s cases articulate is that there must be a qualitative distinction between murders that are eligible for death and those that are not. See State v. Tuttle, 780 P.2d 1203, 1217 (Utah 1988) (citing Gregg, 428 U.S. at 188, 96 S.Ct. at 2932 (citing Furman v. Georgia, 408 U.S. 238, 313, 92 S.Ct. 2726, 2764, 33 L.Ed.2d 346 (1972) (White, J., concurring))); State v. Tillman, 750 P.2d 546, 590 (Utah 1987) (Durham, J., concurring and dissenting, joined by Zimmerman, J.) (citing People v. Green, 27 Cal.3d 1, 164 Cal.Rptr. 1, 38-39, 609 P.2d 468, 505-06 (1980)). Therefore, the sheer number or percentage of murderers exposed to the death penalty, even if we knew those facts, would not tell us much in determining whether the statute imposing the death penalty was constitutional.
For example, during some period of time, Utah might have the unfortunate distinction of being the site of only murders of the most aggravated type, all of which are eligible for the death penalty. That fact alone would not create constitutional problems, however, if the statutory aggravating circumstances making these murders subject to the death penalty narrowed the death penalty’s availability so that those murders could be reasonably distinguished from the universe of all possible murders, *413even though no noncapital murders happened to be committed during the period in question. And to determine whether Utah’s statute performs this discriminating function, one could not simply total the number of aggravating circumstances listed in our statute and compare them to similar lists in other states’ statutes, as Justice Durham does. Nor can the question of whether Utah’s statute performs this discrimination function be determined by trying to imagine the universe of all possible murders, as evidenced by the number of aggravating circumstances in all states, and then ask how large a proportion of that universe Utah’s statute comprehends. Nothing in the Supreme Court’s rather vague death penalty jurisprudence requires such an abstract undertaking! Instead, a determination of whether capital murders are qualitatively distinguishable from noncapital murders requires an evaluation of the operation of each aggravating circumstance. Yet Justice Durham offers no such analysis of the aggravating circumstances charged against Young or of other aggravating circumstances listed in the statute, probably because this court has performed such an analysis of some of the broadest aggravating circumstances and found them valid. We specifically considered the rationality of the aggravating circumstance as it qualifies the murder for death in at least four cases. See State v. James, 819 P.2d 781, 796-97 (Utah 1991) (subparagraph (h)); Tuttle, 780 P.2d at 1217 (subparagraph (q)); State v. Holland, 111 P.2d 1019, 1022-23 (Utah 1989) (subparagraph (h)); State v. Brown, 607 P.2d 261, 266-67 (Utah 1980) (subparagraph (h)). And Justice Durham signals no retreat from her votes in these cases.
It should be clear that if discrimination among murders is the crux of the narrowing of aggravating circumstances required by the federal constitution, judging a statute by the sheer number of aggravating circumstances is, at base, arbitrary. What would be the impact on Justice Durham’s purely quantitative analysis if Utah’s statute were amended to eliminate five or even ten of the aggravating circumstances that seldom, if ever, are actually used to make a murder eligible for death? Would that make the statute constitutional because a number of states then would have more aggravating circumstances than Utah, even though the number and percentage of total murders actually eligible for death would not have been materially diminished? Merely proposing the question suggests the untenability of an affirmative answer and of her position.
One further point: Justice Durham is prepared to declare the Utah statute unconstitutional under both the state and federal constitutions, but she offers no standards to guide a conscientious legislature in crafting a constitutional statute. She does not fix a specific number of aggravating circumstances that are permissible, she does not fix a percentage test for determining the number of murders that can be eligible for death, nor does she offer any test for determining what makes one murder qualitatively different from another. In my view, it would be irresponsible to strike down the Utah statute without articulating a standard for a statute that would pass constitutional muster.
Having said that I would reject Young’s facial challenge to the Utah death penalty statute, I should note that I do share some of the general concerns Justice Durham expresses. Our statute began in 1973 with eight categories of aggravating circumstances. See 1973 Utah Laws ch. 196 (codified as Utah Code Ann. § 76-5-202 (Supp.1973)). Over the intervening years, the legislature has made more and more categories of murder eligible for the death penalty. This is perhaps inevitable. I doubt very much that any Utah legislator has lost an election because he or she voted to increase the penalty on some crime. This legislative lengthening of the list of aggravating circumstances has created a real danger that some of these factors will not make reasonable qualitative distinctions between those murders that are eligible for the death penalty and those that are not, thus violating the standards fixed by the federal constitution for imposing the death penalty.
*414The remedy for this problem, however, is not to strike the entire statute and hope the legislature does better next time. Rather, it falls to this court and the other courts of this state to scrutinize these statutory aggravating circumstances and their application on a case-by-case basis and, when necessary, to give the statute a limiting construction to ensure that it does not sweep an indiscriminately broad category of murders within the reach of the death penalty. We did this in Tuttle. 780 P.2d at 1216-19; see also State v. Wood, 648 P.2d 71, 85-86 (Utah 1981), cert. denied, 459 U.S. 988, 103 S.Ct. 341, 74 L.Ed.2d 383 (1982). And two of us suggested that the rather broad aggravating circumstance relied upon in Tillman, 750 P.2d at 568, was unconstitutionally overinclusive when applied to the facts of that case. Id. at 590-91 (Durham, J., concurring and dissenting, joined by Zimmerman, J.) (citing Utah Code Ann. § 76-5-202(l)(d)). This course may not make us popular. See, e.g., Michael Morris, Utah Supreme Court Justices Seem Determined to Turn the Streets over to Murderers, Deseret News, May 17, 1989, at B4 (Utah Central ed.); Utah Supreme Court Wins “Dumb Quote of the Year” Award, Davis County Clipper, April 26, 1989, at 11. On occasion, it may require overturning the imposition of the death penalty, but that is the only way we can be assured that the penalty’s application in Utah stays within federal constitutional limits. In the present case, it cannot be contended seriously that Young’s actions do not make him constitutionally eligible for the death penalty. Therefore, I find no basis for overturning its imposition on grounds that the statute is overinclusive as applied to him.
For the foregoing reasons, I would reject defendant’s facial attack on the Utah death penalty statute under both the state and federal constitutions.
II
I next address the contention that death qualifying a jury so biases the jury in favor of guilt that the procedure violates the Utah Constitution. The Chief Justice rejects this claim in part IV of his opinion, and Justice Durham accepts it in part VI of hers. Justice Durham relies- upon empirical research that she says has found that death qualifying jurors before the guilt phase makes the resulting jury more prone to convict than would otherwise have been the case and that this is fundamentally unfair to defendant. The Chief Justice relies upon the federal courts’ rejection of the same claim under the federal constitution and upon our decision in State v. Moore, 697 P.2d 233, 237-38 (Utah 1985), which concluded that as a matter of state constitutional law, any guilt-pronen'ess that results from death qualifying a jury is counterbalanced by the requirement that the trial judge also exclude those who would automatically impose the death penalty upon a defendant convicted of first degree murder. Justice Durham’s response is to observe that there is no empirical support for Moore’s assumption that removing automatic-death-penalty jurors counterbalances any bias that may be produced by death qualification and some for the contrary proposition.
I agree with Justice Durham that the empirical evidence raises serious questions about the impact of death qualification upon the fairness of the jury. However, from my reading of the scientific literature she cites, I find no scientific consensus on either the degree of pro-prosecution bias that death qualification introduces into the jury panel or the extent to which this bias skews the resulting jury verdicts. See John A. Wasleff, Note, Lockhart v. McCree: Death Qualification as a Determinant of the Impartiality and Representativeness of a Jury in Death Penalty Cases, 72 Cornell L.Rev. 1075, 1099-100 & n. 151 (1987) [hereinafter Wasleff]. See generally Rogers Elliott, Social Science Data and the APA: The Lockhart Brief as a Case in Point, 15 Law & Hum. Behav. 59 (1991); Rogers Elliott & Robert J. Robinson, Death Penalty Attitudes and the Tendency to Convict or Acquit, 15 Law & Hum. Behav. 389 (1991) [hereinafter Death Penalty Attitudes ]. Absent strong agreement on the empirical question, I am loath to declare a practice that has persisted for *415a number of years violative of the state constitution on empirical grounds alone.
I have two other bases for disagreeing with Justice Durham’s opinion that using a death-qualified jury in the guilt phase of a case denies a defendant a state constitutional right. First, I see no reason for singling out the uncertain effect of death qualification for constitutional condemnation when we tolerate, indeed, have always tolerated, other procedures used in seating juries that have a biasing effect of one kind or another. Without going into the minutia of how juries are selected, I will give one obvious example of a procedure that may— and in fact is calculated to—result in a jury that is more or less conviction-prone than would be the case if the first twelve persons called for jury duty were permitted to try the case: peremptory challenges. There can be no doubt that each side in a jury trial uses its peremptory challenges to attempt to stack the jury in favor of the result sought by that side. It is no response to this observation to say that because both sides have the opportunity to use peremptory challenges, each will operate to cancel out the other’s attempted biasing of the jury. As is true throughout the law, the theory that equal adversaries produce a balanced and “true” result gives way upon close examination to the cold fact that all lawyers, like all judges and all auto mechanics, are not equally skilled or equally diligent. There is no reason to believe that any particular jury is going to be impartial simply because both sides have exercised their peremptory challenges.
Indeed, there is every reason to believe that those challenges, when used skillfully, can have a significant biasing effect. Both an empirical study, see Bruce J. Winick, Prosecutorial Peremptory Challenge Practices in Capital Cases: An Empirical Study and a Constitutional Analysis, 81 Mich.L.Rev. 1, 5, 28-35 (1982), and press accounts of attorneys who have taken advantage of professional jury selection con-, sultants, see Gail Diane Cox, Experts Helped Pick King Jury, Nat’l L.J., May 25, 1992, at 3; Roy E. Black, Subvert Damaging Testimony from Winning: Successful Strategies from 10 Litigators Who Stand Apart from the Crowd, Nat’l L.J., Feb. 3, 1992, at S3; Fred Strasser, Smith Jury Consultant Reflects, Nat’l L.J., Dec. 30, 1991, at 3, suggest that the canny use of peremptories can introduce significant bias into a jury panel. See State v. Cantu, 750 P.2d 591, 596 (Utah 1988). One commentator has noted the logical similarity between the operation of peremptory challenges and the death qualification of jurors:
Holding death qualification unconstitutional would threaten the use of ... peremptory challenges.... A prosecutor’s peremptory challenges, in the very nature of the practice, would alter the attitude pattern of the jury in favor of the prosecution. There is no material difference between removing a Witherspoon-excludable for cause or by peremptory challenge. Either way, the prosecutor removes a juror that [sic] may be unsympathetic to the prosecution’s case.
Wasleff at 1102-03 (footnotes omitted).
I do not say all this in defense of death qualification, but only to explain why I decline to find that the unquantified biasing supposedly produced by death qualification is unconstitutional. It seems to me that if our concern is biasing the jury either way, we should be no more ready to accept the use of peremptory challenges than we are the process of death qualification. The real issue should be whether the jury is biased through the procedures used to seat them, not which procedure produced the bias. If we accept the one, I cannot see how we can find it fundamentally unfair to accept the other.
Second, on a more fundamental level, I think that the constitutional standard of jury impartiality Justice Durham would embed in the Utah Constitution is unrealistic. The jury system as it actually functions in death penalty cases, as in all others, is not a scientifically constructed, finely tuned balancing instrument for determining absolute truth. Rather, like any other human institution, it is a rough approximation of an articulated ideal. One particular jury may more closely approximate that ideal than another, but that does not make the one that falls further from *416perfection unconstitutional. See State v. Andrews, 843 P.2d 1027, 1034-35 (Utah 1992) (Zimmerman, J., concurring). In my-view, Justice Durham applies a standard of constitutionality that is too close to the ideal, one that, is unrealistic. I suspect that the United States Supreme Court has made a similar judgment in ruling that the degree of bias produced by death qualification does not violate the federal constitution.
Having said why I cannot join in a finding that death qualification violates the Utah Constitution, I should note that if I were convinced that there was reliable scientific evidence demonstrating the degree of pro-prosecution bias that death qualification introduces into the jury panel or the extent to which this bias skews the resulting jury verdicts, my-view would change. I would entertain the idea of exercising our inherent supervisory authority over the lower courts, see State v. Thurman, 846 P.2d 1256 (1993); State v. James, 767 P.2d 549, 557 (Utah 1989); In re Criminal Investigation, 754 P.2d 633, 642 (Utah 1988); State v. Bishop, 753 P.2d 439, 499 (Utah 1988) (Zimmerman, J., concurring); State v. Lafferty, 749 P.2d 1239, 1260 (Utah 1988); Long, 721 P.2d at 492; In re Clatterbuck, 700 P.2d 1076, 1081 (Utah 1985), to require that in future dedth penalty cases, sufficient alternative jurors be seated at the beginning of the guilt phase of the trial so that if a guilty verdict were returned, the court could then death qualify the penalty jury as provided by Moore and proceed with the penalty determination.1 As Justice Durham notes in her opinion, such an exercise of supervisory authority was suggested by Justice Daniel O’Hern of the New Jersey Supreme Court in State v. Ramseur, 106 N.J. 123, 524 A.2d 188, 295 (1987) (O’Hern, J., concurring in the result).
This course would have the advantage of allowing us to change an administrative practice that conveniently permits the use of- only one jury but apparently has unacceptable side effects, without first deciding that its effects are so uniform and so severe as to violate the constitutional command of fairness and, perhaps, unnecessarily trivialize that command in the process. It would also permit us to abolish prospectively a questionable practice that has gone on for years without requiring us to reverse convictions of all those tried by death-qualified juries who still await execution or who were convicted of a capital offense but sentenced to life. Such wholesale reversal would be a consequence of Justice Durham’s reasoning, a consequence that she does not mention.
Although this exercise of our supervisory power would have the aforementioned salutary effects, I would hot undertake this course unless I were first convinced of the scientific soundness of our intervention. The evidence on which Justice Durham relies does not appear to me to be sufficiently definitive to show that such a step does not have a reasonable likelihood of producing unforeseen deleterious consequences. See Death Penalty Attitudes at 401-03 & n. 1.
As a procedural matter, I observe that in past cases in which we have relied upon our own reading of the results of scientific studies as a basis for promulgating a new rule of law or exercising our inherent supervisory authority, the empirical evidence has been readily understandable by the relatively unsophisticated reader. There also has been a substantial consensus on the operation of the legal rule or procedure in question and on the impact of that particular rule or procedure on the judicial process. See, e.g., State v. Ramirez, 817 P.2d 774, 779-80 (Utah 1991); Tuttle, 780 P.2d at 1210; Kofford v. Flora, 744 P.2d 1343, 1348 (Utah 1987); Long, 721 P.2d at 488. But where, as here, the scientific literature is abstruse and lacking in apparent consensus on critical issues, we should not act to change a rule of law or procedure upon little more than assertions by advocates that the evidence proves something. See generally Death Penalty Attitudes. Instead, if, as appears to be the case here, the science is conflicting but raises serious *417questions as to the fairness of a procedure or a rule of law, we should request expert amicus briefs and supplemental briefing from both parties to clarify the complex and often contradictory scientific studies on this issue. We should give the parties a chance to enlighten us. See, e.g., Robert E. Keeton, Legislative Facts and Similar Things: Deciding Disputed Premise Facts, 73 Minn.L.Rev. 1 (1988); Laurens Walker & John Monahan, Social Frameworks: A New Use of Social Science in Law, 73 Va.L.Rev. 559 (1987). Once this is done, we would be in a position to better evaluate the scientific evidence and act upon it.2 But until that is done, I would not change the present method for impaneling death-penalty juries.
Accordingly, I reject Young’s constitutional challenge to the death qualification of his jury. I concur in the result with part IV of the Chief Justice’s opinion.
Ill
The next issue is the for-cause challenge to the jurors, dealt with in part VI of the Chief Justice’s opinion and part VII of Justice Durham’s. I concur in the Chief Justice’s analysis, except as to juror Cole. As to him, I concur with Justice Durham that he should have been stricken for cause. He displayed a clear view that if defendant was convicted of a murder that was at all aggravated, death was the only appropriate penalty. The prosecutor’s attempt to rehabilitate him was ineffective because the hypothetical instruction upon which he based his questions was one that would never be given by a trial judge in submitting the penalty question to a jury. The failure to strike this juror constitutes harmful error and requires vacation of the verdict and death penalty. I would therefore remand this case for a new trial. See State v. Jones, 734 P.2d 473, 475 (Utah 1987); State v. Hewitt, 689 P.2d 22, 26 (Utah 1984).
IV
In addition, regarding defendant’s entitlement to a guilty and mentally ill instruction and verdict, I join Justice Durham in part IV of her opinion and dissent from part X of the Chief Justice’s opinion. Justice Durham would hold that defendant was entitled to such an instruction and verdict form, and I agree that under rule 21.5 of the Utah Rules of Criminal Procedure, such a verdict and instruction form are available upon request without regard to whether a defendant offers an insanity defense. See Utah R.Crim.P. 21.5. I also agree that the failure to give such an instruction and verdict form was harmful error. However, in finding harmfulness, I rely on the cumulative effect of this error and the other errors identified below that occurred during the penalty, phase.
V
Regarding the shackling of defendant during the penalty phase, I dissent from part XII of the Chief Justice’s opinion and I join part I of Justice Durham’s opinion to the extent she finds that the trial court failed to show that the shackling was a necessity, thus abusing its discretion.
VI
On the question of the denial of allocution to defendant, I dissent from part XV of the Chief Justice’s opinion and join part II of Justice Durham’s opinion to the extent it holds that defendant had a statutory right to allocution that was wrongfully denied him.
I do not find it necessary to determine whether a denial of this right warrants an automatic reversal or whether it is to be appraised under our usual harmless error rule. Even if the usual harmless error standard applies, given the errors discussed in the preceding paragraphs that occurred in connection with the penalty phase, I conclude that the cumulative effect of these errors undermines my confi*418dence in the outcome of the penalty phase and requires reversal of the death penalty.
VII
I would vacate the conviction and remand for a new trial. Given this conclusion, I have no occasion to consider the other issues addressed by Justice Durham.
Because of the viciousness of the crime of which Young was convicted and his past record of homicide, our decision today is bound to arouse adverse comment. Obviously, anticipation of that comment has not deterred us from voting as we think the law requires. However, it is worth noting that in voting to reverse defendant’s conviction, no member of this court has suggested that he is innocent of the appalling crime of which he was convicted or that the commission of that crime by one with defendant’s past record of violent crime cannot be punished by death. But we are saying that the processes by which such a conviction is entered and such a sentence is imposed must comply with Utah’s statutes and the federal and state constitutions, something that did not occur here.
. Even Duckett v. State, 104 Nev. 6, 752 P.2d 752, 755 (1988), upon which the lead opinion relies, admits that physical restraints at sentencing may not be imposed absent necessity. Furthermore, although Duckett upheld a sentencing-stage shackling order, it did not, as the lead opinion suggests, hold that the constitutional right to be free of shackles did not exist at sentencing; it held only that the constitutional right to be free of prison garb, established in Estelle v. Williams, 425 U.S. 501, 96 S.Ct. 1691, 48 L.Ed.2d 126 (1976), did not exist at sentencing.
. In contrast, the defendant in Elledge had made threats upon the court, 823 F.2d at 1450, and the defendant in Spain had repeatedly disrupted the pretrial proceedings, 883 F.2d at 719. Nevertheless, the reviewing courts in both cases deemed those shackling orders unwarranted. 823 F.2d at 1452; 883 F.2d at 728.
. Young has raised no state constitutional challenge to his shackling.
. I am aware that the United States Supreme Court seems to be making exactly this suggestion in Saffle v. Parks, discussed below. I can only hope that the Supreme Court’s apparent readiness to undertake an unacknowledged yet complete retrenchment from Brown is as transitory as was the McGautha Court’s unwillingness to require that death penalty statutes provide specific guidelines for determining who could be put to death. And I emphasize that the retrenchment has not yet occurred, as I explain later.
. The quoted language originated in Justice O’Connor’s concurring opinion in California v. Brown. Her concurrence clearly implied that an improper antisympathy instruction could mislead jurors into ignoring the full import of the mitigating evidence. 479 U.S. at 545-46, 107 S.Ct. at 841-42.
. The availability of a GAMI verdict, like a guilty but mentally ill verdict, may serve to elicit more guilty pleas. See Ira Mickenberg, A Pleasant Surprise: The Guilty But Mentally III Verdict Has Both Succeeded in its Own Right and Successfully Preserved the Traditional Role of the Insanity Defense, 55 U.Cin.L.Rev. 943, 989 (1987).
. I note, however, that because the prosecution did not elicit testimony about the victim’s general sexual behavior (and suggested as much to the trial court in opposing defendant’s proffered cross-examination), I view with great disfavor the prosecution’s use of this testimony in its closing argument at the penalty phase.
. In Witherspoon v. Illinois, 391 U.S. 510, 88 S.Ct. 1770, 20 L.Ed.2d 776 (1968), the United States Supreme Court held that in capital cases, the prosecution could not challenge for cause jurors merely scrupled against or hesitant to impose the death penalty. Id. at 522, 88 S.Ct. at 1777. The opinion did not disturb the right of the prosecution to challenge two categories of jurors: those whose "reservations about capital punishment would prevent them from making an impartial decision as to the defendant's guilt” and those who "could never vote to impose the death penalty." Id. at 513-14, 522 n. 21, 88 S.Ct. at 1772-73, 1777 n. 21. Those jurors still removable for cause after Witherspoon have come to be known as “Witherspoon-exdudables.” See, e.g., McCree, 416 U.S. at 167 n. 1, 106 S.Ct. at 1761 n. 1. The two categories often are distinguished by referring to jurors in the first category as "nullifiers.” Id. at 172, 106 S.Ct. at 1764. In light of Wainwright v. Witt, 469 U.S. 412, 105 S.Ct. 844, 83 L.Ed.2d 841 (1985), the second category now includes all those jurors whose attitudes about capital punishment "would prevent or substantially impair the performance of [the juror’s] duties as a juror." Id. at 424, 105 S.Ct. at 852.
This opinion uses the term " Witherspoon-ex-cludable,” or WE, to refer only to the second category. Critics of the use of death-qualified juries at the guilt phase of capital trials condemn only the exclusion of this second category of WEs. Individuals in this category are perfectly able to serve as fair and impartial jurors in the guilt phase, yet are excluded from the guilt phase because of their inability later to serve at the penalty phase. It is this practice that I conclude violates capital defendants’ constitutional rights.
I would continue to permit, however, the exclusion from all phases of capital trials those jurors in the first category. To this end, I would also permit carefully limited questioning, designed to identify these nullifiers. Nullifiers are by definition unable to fulfill their obligation as jurors to make an impartial decision at either the guilt or penalty phase, and therefore, their exclusion from all stages of a capital trial is proper. But in light of the discussion in subsection C below, I stress that courts must identify nullifiers with a minimum of inquiry concerning their attitudes toward the death penalty. Justice Marshall and others have suggested the ease with which courts can do this. See McCree, 476 U.S. at 203, 106 S.Ct. at 1780-81; Grigsby v. Mabry, 569 F.Supp. 1273, 1310 (E.D.Ark.1983); Bruce J. Winick, Prosecutorial Peremptory Challenge Practices in Capital Cases: An Empirical Study and a Constitutional Analysis, 81 Mich.L.Rev. 1 (1982).
. If defendant Moore in fact asserted a right to be sentenced by a nondeath-qualified jury, then in the instant case we face a new issue. If, however, Moore did not assert such a right, our legislative policy argument in Moore missed the mark.
. For recent studies supporting the claim that death-qualified juries may be conviction-prone, see Claudia L. Cowan et al., The Effects of Death Qualification on Jurors’Predisposition to Convict and on the Quality of Deliberation, 8 Law & Hum.Behav. 53 (1984); Robert Fitzgerald & Phoebe C. Ellsworth, Due Process vs. Crime Control: Death Qualification and Jury Attitudes, 8 Law & IIumlBehav. 31 (1984); Craig Haney, On the Selection of Capital Juries: The Biasing Effects of the Death-Qualification Process, 8 Law & Hum.Behav. 121 (1984); Irwin A. Horowitz & David G. Seguin, The Effects of Bifurcation and Death Qualification on Assignment of Penalty in Capital Crimes, 16 J. Applied Soc.Psychol. 165 (1986); Joseph B. Kadane, After Hovey: A Note on Taking Account of the Automatic Death Penalty Jurors, 8 Law & Hum.Behav. 115 (1984); James Luginbuhl & Kathi Middendorf, Death Penalty Beliefs and Jurors' Responses to Aggravating and Mitigating Circumstances in Capital Triais, 12 Law & Hum.Behav. 263 (1988); Marilyn D. McShane et ah, Eligibility for Jury Service in Capital Trials: A Question of Potential Exclusion and Bias, Texas Bar J., April 1987, at 365; Gary Moran & John C. Comfort, Neither “Tentative” nor “Fragmentary": Verdict Preference of Impaneled Felony Jurors as a Function of Attitude Toward Capital Punishment, 71 J. Applied Psy-chol. 146 (1986); Michael L. Neises & Ronald C. Dillehay, Death Qualification and Conviction Proneness: Witt and Witherspoon Compared, 5 Behav. Sci. & the L. 479 (1987); Rick Seltzer et ah, The Effect of Death Qualification on the Propensity of Jurors to Convict: The Maryland Example, 29 How.L.J. 571 (1986); David G. Se-guin & Irwin A. Horowitz, The Effects of "Death Qualification" on Juror and Jury Decisioning: An Analysis from Three Perspectives, 8 L. & Psychol.Rev. 49 (1984); William C. Thompson et ah, Death Penalty Attitudes and Conviction Proneness: The Translation of Attitudes Into Verdicts, 8 L. & Hum.Behav. 95 (1984). But see Rogers Elliott & Robert J. Robinson, Death Penalty Attitudes and the Tendency to Convict or Acquit, 15 L. & Hum.Behav. 389 (1991); J.L. Bernard & W.O. Dwyer, Witherspoon v. Illinois: *389The Court Was Right, 8 L. & Psychol.Rev. 105 (1984).
For earlier studies supporting the claim that death-qualified juries may be conviction-prone, see Virginia R. Boehm, Mr. Prejudice, Miss Sympathy, and the Authoritarian Personality: An Application of Psychological Measuring Techniques to the Problem of Jury Bias, 1968 Wis. L.Rev. 734; Edward J. Bronson, On the Conviction Proneness and Representativeness of the Death-Qualified Jury: An Empirical Study of Colorado Veniremen, 42 U.Colo.L.Rev. 1 (1970); Edward J. Bronson, Does theExclusion of Scrupled Jurors in Capital Cases Make the Jury More Likely to Convict? Some Evidence from California, 3 Woodrow Wilson LJ. 11 (1980); Faye Goldberg, Toward Expansion of Witherspoon: Capital Scruples, Jury Bias, and Use of Psychological Data to Raise Presumptions in the Law, 5 Harv.C.R.—C.L.L.Rev. 53 (1970); Louis Harris & Assocs., Study No. 2016 (1971), in Welsh S. White, The Constitutional Invalidity of Convictions Imposed by Death Qualified Juries, 58 Cornell L.Rev. 1176, 1185-86 (1973); George L. Jurow, New Data on the Effect of a “Death Qualified" Jury on the Guilt Determination Process, 84 Harv.L.Rev. 567 (1971); see also MJ. Lerner, The Belief in a Just World, New York: Plenum, 1981 and Z. Rubin & L.A. Peplau, Who Believes in a Just World?, 31 J.Soc.Issues 65 (1975), cited in Luginbuhl & Middendorf, supra; W.C. Wilson, Belief in Capital Punishment and Jury Performance (unpubl. manuscript, University of Texas, 1964) and H. Zeisel, Some Data on Juror Attitudes Toward Capital Punishment, Center for Studies in Criminal Justice, University of Chicago Law School (monograph 1968), cited in Witherspoon, 391 U.S. at 516 n. 10; unpubl. studies summarized in Grigsby v. Mabry, 569 F.Supp. 1273, 1298-1308 (E.D.Ark.1983).
. For an explanation of Witt’s modification of Witherspoon, see supra note 8.
. Although the Supreme Court's opinion also included several pages of dicta challenging some of the empirical conclusions, 476 U.S. at 168-73, 106 S.Ct. at 1762, Justice Marshall’s dissent more than responded to these challenges, id. at 189-92, 106 S.Ct. at 1773-75, and many other commentators have echoed Justice Marshall’s criticism. See, e.g., Ian T. Moar, Death Qualified Juries in Capital Cases: The Supreme Court’s Decision in Lockhart v. McCree, 19 Colum.Hum.Rts.L.Rev. 369, 387-93 (1988); William C. Thompson, Death Qualification After Wainwright v. Witt and Lockhart v. McCree, 13 L. & Hum.Behav. 185, 194-204 (1989).
. Both Schreuder, 726 P.2d at 1225, and Moore, 697 P.2d at 237, claim that we established this requirement in State v. Norton, 675 P.2d 577 (Utah 1983), cert. denied, 466 U.S. 942, 104 S.Ct. 1923, 80 L.Ed.2d 470 (1984), overruled on other grounds by State v. Hansen, 734 P.2d 421 (Utah 1986). In fact, Norton held only that when a defendant requested it, the court had to inquire whether any prospective jurors would feel compelled to vote to impose a death sentence upon all persons convicted of murder. 675 P.2d at 589. Norton thus prefigured the United States Supreme Court’s opinion in Morgan v. Illinois, — U.S. -, 112 S.Ct. 2222, 119 L.Ed.2d 492 (1992). The more rigorous demands articulated in Schreuder and Moore, however, reflect our evolving recognition that when death qualifying a jury, the trial court must level the playing field from the outset for both the state and the defendant and must ensure that it remains level throughout the death-qualification process. My review of the voir dire in this case suggests that this obligation needs further clarification.
. This questioning process, designed to establish or rebut an ADP challenge for cause, typically involved both sets of attorneys in an extensive, complex, and frequently suggestive line of questioning. This type of adversarial endeavor creates precisely the process problem I discuss in subsection C.
. As part of its initial follow-up to the WE inquiry, during the individual voir dire the trial court appears to have usually asked each juror a question along the lines of, "Would you say you’re in favor of the death penalty, somewhat in favor of the death penalty, somewhat opposed to the death penalty, or strongly opposed to the death penalty?” Because a juror could be merely somewhat in favor of the death penalty and yet want to impose it automatically on all those convicted of capital murder, this question is not an adequate ADP question.
. That ADPs and WEs are in fact in the pool is irrelevant, given that they can never sit on a capital jury. Although the point of the fair cross-section requirement is not to ensure that each petit jury contains a representative sample of the community, it is to ensure at least the possibility that a petit jury could be representative. Death qualification eliminates this possibility.
. I also note that in the current system the defense must choose between naming all its potential witnesses at the voir dire prior to the guilt phase, including those it may call only for *395mitigating testimony at the sentencing phase, or risk having an undisclosed witness disqualified at the sentencing phase if it turns out that an empaneled juror has some prior relationship with that witness. After objecting to this dilemma, Young chose to risk losing his mitigating witnesses rather than naming them all prior to the guilt phase. Using separate juries would eliminate this problem as well.
. Nor do I find that the trial court abused its discretion in granting the State’s challenge for cause regarding juror Hoffman. The lead opinion accurately indicates that one legitimate interpretation of her voir dire testimony, particularly if confirmed by her demeanor, was that her emotional condition may have made it difficult for her to perform her duties as a juror, especially in a lengthy capital trial. Applying a deferential standard of review, I also vote to affirm the trial court. It dismays me, however, that the lead opinion supports this conclusion by reciting rather benign excerpts from the voir dire transcripts. In my experience, an inability to recall the previous page of a book or the location of one’s sunglasses simply does not indicate an unusual memory problem sufficient to disqualify one for legal reasoning. We do the law a disservice by reciting these isolated passages as grounds for dismissal for cause.
. In impaneling this jury, the court could, of course, exclude any "nullifiers,” or people who refuse to find a defendant guilty if there is any chance that he or she might receive the death penalty.
. For the degree of rigor with which we should examine empirical evidence upon which changes in the law are made, see State v. Rimmasch, 775 P.2d 388, 400-03 (Utah 1989).