CHIN, J.

We granted review to decide whether the so-called rule of consistency that the acquittal of all alleged coconspirators but one requires acquittal of the remaining alleged conspiratorrequires the reversal of a conspiracy conviction in this case, and whether similar principles require the reversal of a finding as to one but not both codefendants that an attempted murder was premeditated. We conclude that the rule of consistency is a vestige of the past with no continuing validity. Many reasons may explain apparently inconsistent verdicts: lenience, compromise, differing evidence as to different defendants, or, possibly, that two juries simply viewed similar evidence differently. If substantial evidence supports a jury verdict as to one defendant, that verdict may stand despite an apparently inconsistent verdict as to another defendant.

In this case, substantial evidence supports the conspiracy conviction and the premeditation finding. Accordingly, we affirm the judgment of the Court of Appeal, which affirmed this conviction and this finding.

On December 11, 1996, codefendants 15 year old Floyd Maurice Palmer and 29 year old Donald Edmond Price engaged in a crime spree, only part of which is relevant to the issues now before us.

As relevant here, the evidence at trial showed that early that morning, Price was driving his blue Chevrolet Nova, with Palmer in the passenger seat. Price drove past Richard Humphries, who was driving a red Ford Escort, moved in front of him, and slowed down so that Humphries had to pass him. Price pulled alongside Humphries, and Palmer fired one shot with a black .25 caliber semiautomatic handgun. The bullet broke Humphries's driver's side window and grazed his scalp. Humphries tried to speed away, but Price pulled alongside him again. Humphries slammed on his brakes. Palmer fired a second shot, breaking Humphries's left rear window and striking a door post. Price stopped, then started to back up. Humphries got away by turning around and driving on the wrong side of the divided highway. Price testified that he did not know Palmer had a gun or was going to shoot Humphries. He said Palmer fired only one shot, after which he slammed on the brakes. He denied pursuing Humphries. Palmer testified that Price gave him the gun and told him to shoot, hitting him across the mouth when he refused. He then fired once or twice but, he said, without aiming at Humphries or intending to kill him.

A couple of hours later, Price, now driving a stolen pickup truck, sideswiped Judith Showalter's 1996 BMW. Both drivers stopped. Price got out of the truck and apologized to Showalter, saying his brakes had locked. Palmer also got out of the truck. He went to Showalter's car, opened the door, said, "take this, bitch," then shot Showalter in the head. She slumped over and played dead. The two men took her pager and purse. One of them pulled her out of the car and left her on the ground. They drove away in the BMW, leaving behind the truck and other property they had stolen that morning. Showalter survived despite suffering major injuries. The bullet had entered her head near the left eye and exited at the right earlobe. Price testified that he and Palmer had agreed to rob Showalter, but he denied knowing Palmer would shoot her. Palmer denied knowing that Price was going to sideswipe Showalter's car or rob her. He testified that Price forced him to shoot Showalter, but he did not intend to kill her.

Price and Palmer each gave statements to the police implicating the other. The prosecution charged them together for these crimes and argued it would be too stressful for Showalter to have to testify at separate trials. For these reasons, the trial court ordered them tried together but by two separate juries. During the trial, Palmer chose to have his jury hear Price's pretrial statements. After Palmer had testified in his own defense, Palmer's pretrial statements were admitted in rebuttal. Accordingly, in the end, both juries heard essentially the same evidence.

Among other charges, the prosecution charged both Price and Palmer with attempted murder of Humphries, alleged to have been premeditated (Pen.Code, §§ 187, 664), and conspiracy to murder Showalter (Pen.Code, §§ 182, 187, subd. (a)). Price's jury found him guilty of all charges and found the premeditation allegation true. Palmer's jury found him guilty of the other charges and of attempting to murder Humphries, but it found the premeditation allegation not true and found him not guilty of conspiracy to murder Showalter. The court sentenced both defendants to state prison.

Both defendants appealed. Price argued that the premeditation finding and conspiracy conviction had to be reversed in light of the different verdicts the other jury made as to Palmer. In an opinion authored by Justice Richli, the Court of Appeal modified the judgments slightly, but otherwise affirmed them. It rejected Price's arguments regarding the premeditation finding and conspiracy conviction. The court recognized that current California authority supports the general rule that the acquittal of all but one alleged coconspirator requires the acquittal of the remaining one. In a comprehensive analysis, it expressed doubt that the rule remains valid but, partly because the California authority includes dicta from this court, it ultimately concluded that "stare decisis counsels us to acquiesce." Nevertheless, it found that the rule applies only to defendants tried together before a single jury, not to defendants tried separately or, as here, together but before separate juries. Accordingly, it upheld the verdicts in this case. The court also stressed that it acquiesced in the rule of consistency "only reluctantly. We believe the entire topic of inconsistent verdicts against coparticipants is ripe for reexaminationand revisionby the Supreme Court."

Both defendants petitioned for review. We granted Price's petition, limited to whether the verdicts as to codefendant Palmer require reversal of Price's conspiracy conviction and premeditation finding.

Price argues that, because it takes at least two to conspire, the verdict finding Palmer, his only alleged coconspirator, not guilty of conspiracy is inconsistent with his conviction for the same conspiracy. This inconsistency, he further argues, requires us to reverse that conviction. He makes a similar argument regarding the premeditation findings. We disagree. The conspiracy verdicts, if not the premeditation findings, are indeed inconsistent, but both may be given effect. The law generally accepts inconsistent verdicts as an occasionally inevitable, if not entirely satisfying, consequence of a criminal justice system that gives defendants the benefit of a reasonable doubt as to guilt, and juries the power to acquit whatever the evidence.

The United States Supreme Court has embraced this general rule. "Inconsistency in a verdict is not a sufficient reason for setting it aside. We have so held with respect to inconsistency between verdicts on separate charges against one defendant, Dunn v. United States, 284 U.S. 390 , 52 S.Ct. 189 , 76 L.Ed. 356 (1932), and also with respect to verdicts that treat codefendants in a joint trial inconsistently, United States v. Dotterweich, 320 U.S. 277, 279 , 64 S.Ct. 134 , 88 L.Ed. 48 (1943)." ( Harris v. Rivera (1981) 454 U.S. 339, 345 , 102 S.Ct. 460 , 70 L.Ed.2d 530 , fns. omitted.) California law is similar. "An acquittal of one or more counts shall not be deemed an acquittal of any other count." (Pen.Code, § 954; see People v. Santamaria (1994) 8 Cal.4th 903, 911 , 35 Cal. Rptr.2d 624 , 884 P.2d 81 ; People v. Pahl (1991) 226 Cal.App.3d 1651, 1656-1657 , 277 Cal.Rptr. 656 .) "The fact that certain defendants may escape conviction for their crimes is not any legal or logical reason why another defendant, where substantial evidence has been introduced to sustain his conviction, should be exonerated and be permitted to escape punishment for his crime." ( People v. Taylor (1948) 88 Cal. App.2d 983, 987-988 , 199 P.2d 751 .) Accordingly, "The general rule is that acquittal of one codefendant normally will not require acquittal of another." ( People v. Howard (1988) 44 Cal.3d 375, 412 , 243 Cal.Rptr. 842 , 749 P.2d 279 .)

Price relies on an exception to this general rule: the rule of consistency in conspiracy cases. Although the rule has deep common law roots (see, e.g., United States v. Espinosa-Cerpa (5th Cir.1980) 630 F.2d 328, 332-333, fn. 5 ), it first appeared in California in People v. James (1961) 189 Cal.App.2d 14 , 10 Cal.Rptr. 809 . The rule and its rationale were summarized in People v. Nunez (1986) 183 Cal.App.3d 214, 218 , 228 Cal.Rptr. 64 : "When two or more alleged coconspirators are jointly tried, it is generally held that acquittal of all persons with whom a defendant is alleged to have conspired precludes conviction of the remaining defendant. This result is compelled by the rule of consistency of verdicts because, as Justice Cardozo once noted, `It is impossible in the nature of things for a man to conspire with himself.' ( Morrison v. California (1934) 291 U.S. 82, 92 [ 54 S.Ct. 281 , 78 L.Ed. 664 ]; ...)"

After People v. James, supra, 189 Cal. App.2d 14 , 10 Cal.Rptr. 809 , other California decisions have followed, or cited while distinguishing, the rule of consistency, but without independent analysis. (E.g., People v. Reeves (1967) 250 Cal.App.2d 490, 492 , 58 Cal.Rptr. 517 ; People v. Holzer (1972) 25 Cal.App.3d 456, 459 , 102 Cal. Rptr. 11 ; People v. Superior Court (Jackson ) (1975) 44 Cal.App.3d 494, 498 , 118 Cal.Rptr. 702 ; People v. Austin (1994) 23 Cal.App.4th 1596, 1603 , 28 Cal.Rptr.2d 885 .) We have cited it in dicta. ( People v. Taylor (1974) 12 Cal.3d 686, 694-695 , 117 Cal.Rptr. 70 , 527 P.2d 622 ; People v. Chapman (1984) 36 Cal.3d 98, 115 , 201 Cal.Rptr. 628 , 679 P.2d 62 .) Until recently, federal courts generally followed the rule. (E.g., Lubin v. United States (9th Cir.1963) 313 F.2d 419, 423 ; Cofer v. United States (5th Cir.1930) 37 F.2d 677, 680 , cited in People v. James, supra, 189 Cal. App.2d at p. 16 , 10 Cal.Rptr. 809 .)

The United States Supreme Court has never clearly spoken on this precise question. Defendant cites three high court decisions that his claims have recognized the rule of consistency, but none even mention it. Two of the cases merely state that it takes at least two to conspire. ( Morrison v. California, supra, 291 U.S. at p. 92 , 54 S.Ct. 281 ; Gebardi v. United States (1932) 287 U.S. 112, 123 , 53 S.Ct. 35 , 77 L.Ed. 206 .) The statement is correct, but it only shows that the verdicts here were inconsistent, not that the inconsistency requires reversal of the guilty verdict. In the third case, Hartzel v. United States (1944) 322 U.S. 680 , 64 S.Ct. 1233 , 88 L.Ed. 1534 , three defendants were tried together and convicted of certain crimes, including conspiracy. The trial court had set aside two of the defendants' convictions because of insufficient evidence. Without analysis or citation to authority, the high court stated that the codefendants "were the only co-conspirators of petitioner named in the indictment and the setting aside of their convictions makes it impossible to sustain petitioner's conviction upon the basis of ... the conspiracy count." ( Id. at p. 682, fn. 3., 64 S.Ct. 1233 ) The court seemed to be saying that because the evidence was insufficient that either of the two codefendants had conspired with the petitioner, the evidence was likewise insufficient that the petitioner had conspired with either of them. Fairly read, the case simply involved insufficient evidence rather than a silent invocation of the rule of consistency.

In the last two decades, the trend has turned against the rule of consistency, largely due to high court decisions involving inconsistent verdicts generally. In Standefer v. United States (1980) 447 U.S. 10 , 100 S.Ct. 1999 , 64 L.Ed.2d 689 ( Standefer ), the petitioner was convicted of aiding and abetting a crime despite the fact that the alleged actual perpetrator, Niederberger, had previously been acquitted of that crime in a different prosecution. Relying on the doctrine of nonmutual collateral estoppel, the petitioner argued that the prior acquittal precluded the government from relitigating the question of the actual perpetrator's guilt. The high court disagreed. It noted that a jury has the power, if not the legal right, to acquit no matter what the evidence, and the prosecution may never challenge such an acquittal. This circumstance "in criminal cases permits juries to acquit out of compassion or compromise or because of `"their assumption of a power which they had no right to exercise, but to which they were disposed through lenity."` Dunn v. United States[, supra ,] 284 U.S. 390, 393 , 52 S.Ct. 189 , 76 L.Ed. 356 ..." ( Id. at p. 22, 100 S.Ct. 1999 .) "In denying preclusive effect to Niederberger's acquittal, we do not deviate from the sound teaching that `justice must satisfy the appearance of justice.' [Citation.] This case does no more than manifest the simple, if discomforting, reality that `different juries may reach different results under any criminal statute. That is one of the consequences we accept under our jury system.' [Citation.] While symmetry of results may be intellectually satisfying, it is not required. [Citation.] [¶] Here, petitioner received a fair trial at which the Government bore the burden of proving beyond reasonable doubt that Niederberger violated [the statute] and that petitioner aided and abetted him in that venture. He was entitled to no less and to no more." ( Id. at pp. 25-26, 100 S.Ct. 1999 .)

A year after Standefer , the court upheld inconsistent verdicts by a trial judge sitting without a jury. ( Harris v. Rivera, supra, 454 U.S. 339 , 102 S.Ct. 460 , 70 L.Ed.2d 530 .) Then, in 1984, it considered a case in which a single jury inconsistently found a single defendant guilty of some counts and not guilty of others. ( United States v. Powell (1984) 469 U.S. 57 , 105 S.Ct. 471 , 83 L.Ed.2d 461 ( Powell ).) Reaffirming its earlier decision in Dunn v. United States (1932) 284 U.S. 390 , 52 S.Ct. 189 , 76 L.Ed. 356 , the court gave effect to all the verdicts, including those of guilty that were irreconcilable with those of not guilty. It noted that prior decisions "establish `the unreviewable power of a jury to return a verdict of not guilty for impermissible reasons,'" and that "inconsistencies often are a product of jury lenity." ( Powell, supra, 469 U. S. at pp. 63, 65, 105 S.Ct. 471 .) "The fact that the inconsistency may be the result of lenity, coupled with the Government's inability to invoke review, suggests that inconsistent verdicts should not be reviewable." ( Id. at p. 66, 105 S.Ct. 471 .) The court rejected "a rule that would allow criminal defendants to challenge inconsistent verdicts on the ground that in their case the verdict was not the product of lenity, but of some error that worked against them. Such an individualized assessment of the reason for the inconsistency would be based either on pure speculation, or would require inquiries into the jury's deliberations that courts generally will not undertake.... [W]ith few exceptions [citations], once the jury has heard the evidence and the case has been submitted, the litigants must accept the jury's collective judgment. Courts have always resisted inquiring into a jury's thought processes [citations]; through this deference the jury brings to the criminal process, in addition to the collective judgment of the community, an element of needed finality." ( Id. at pp. 66-67, 105 S.Ct. 471 .)

The Powell court also noted "that a criminal defendant already is afforded protection against jury irrationality or error by the independent review of the sufficiency of the evidence undertaken by the trial and appellate courts. This review should not be confused with the problems caused by inconsistent verdicts. Sufficiency-of-the-evidence review involves assessment by the courts of whether the evidence adduced at trial could support any rational determination of guilt beyond a reasonable doubt. [Citations.] This review should be independent of the jury's determination that evidence on another count was insufficient. The Government must convince the jury with its proof, and must also satisfy the courts that given this proof the jury could rationally have reached a verdict of guilt beyond a reasonable doubt. We do not believe that further safeguards against jury irrationality are necessary." ( Powell, supra, 469 U.S. at p. 67 , 105 S.Ct. 471 .)

In the wake of these decisions, numerous federal courts have reconsidered and rejected the rule of consistency. ( U.S. v. Acosta (2d Cir.1994) 17 F.3d 538, 545 ["In light of Powell , one defendant's conspiracy conviction does not become infirm by reason of jury verdicts of not guilty against all of his alleged coconspirators"]; U.S. v. Zuniga-Salinas (5th Cir.1992) 952 F.2d 876, 878 (in bank) [Overruling prior circuit authority and concluding that an "inconsistent verdict should no longer be a bar to conviction where all other co-conspirators are acquitted. For the reasons enunciated in Powell ..., such verdicts should not be subject to review for inconsistency."]; U.S. v. Bucuvalas (1st Cir.1990) 909 F.2d 593, 597 ["the `rule of consistency' ... is no longer viable in light of Powell "]; U.S. v. Thomas (4th Cir.1990) 900 F.2d 37, 40 ; U.S. v. Dakins (D.C.Cir.1989) 277 App. D.C. 91 , 872 F.2d 1061, 1065 [declining to adopt the rule in light of Powell ]; U.S. v. Andrews (11th Cir.1988) 850 F.2d 1557, 1561 (in bank) [overruling prior circuit authority]; U.S. v. Valles-Valencia (9th Cir. 1987) 823 F.2d 381, 382 [rejecting the rule in light of Powell and concluding that contrary language in Lubin v. United States, supra, 313 F.2d 419 , "can no longer be relied upon"].) Others have questioned the rule in light of these cases but not decided its continuing validity. ( U.S. v. Abbott Washroom Systems, Inc. (10th Cir. 1995) 49 F.3d 619, 621-623 ; Cortis v. Kenney (8th Cir.1993) 995 F.2d 838 , 840 & fn. 2 [recognizing that the "rule of consistency has come under substantial attack," but assuming it has continuing force in the circuit while distinguishing it]; U.S. v. Velasquez (3d Cir.1989) 885 F.2d 1076, 1091, fn. 13 [noting the "possibility that [ Standefer and Powell ] have undercut the rule of consistency"]; U.S. v. Mancari (7th Cir. 1989) 875 F.2d 103, 104-105 ; Government of Virgin Islands v. Hoheb (3d Cir. 1985) 777 F.2d 138, 142, fn. 6 [noting that " Standefer and Powell suggest that the rule of consistency may be a vestige of the past"]; see Hoheb, at p. 143 (cone. opn. of Garth, J.) ["the rule of consistency since Standefer and Powell is no longer a viable doctrine"]; but see U.S. v. Boling (6th Cir. 1989) 884 F.2d 924, 926 [simply citing while distinguishing the rule].)

We too believe the rule of consistency should be reconsidered. The general rationale behind the rule has been "that one may not conspire with himself." ( People v. James, supra, 189 Cal.App.2d at p. 16 , 10 Cal.Rptr. 809 .) It does take at least two to conspire. But to go "[f]rom this irrefutable proposition" to a rule requiring reversal of the inconsistent verdict is a "precipitous leap." ( U.S. v. Andrews, supra, 850 F.2d at p. 1560 .) We find the federal decisions persuasive and, like the high court, conclude that we may accept inconsistent verdicts. We agree with the conclusion of the Eleventh Circuit Court of Appeals: "Consistent verdicts are unrequired in joint trials for conspiracy: where all but one of the charged conspirators are acquitted, the verdict against the one can stand." ( Id. at p. 1561.)

Our criminal justice system, which permits a conviction only if the jury unanimously finds beyond a reasonable doubt that a defendant is guilty of the particular charge, gives the defendant the benefit of the doubt. Moreover, a jury clearly has the unreviewable power, if not the right, to acquit whatever the evidence. An inevitable result of this system, and one that society accepts in its quest to avoid convicting the innocent, is that some criminal defendants who are guilty will be found not guilty. This circumstance does not, however, mean that if one person receives lenient treatment from the system, all must. "[I]t is always possible for a jury to exercise lenity and acquit some of the defendants while convicting others who are in fact no more guilty, and when this happens the convicted defendants have no remedy. [Citations.] Such incongruities are built into the American system of criminal justice and can have no weight in our decision whether to reverse the denial of a new trial to the present defendants." ( U.S. v. Williams (7th Cir.1996) 81 F.3d 1434, 1437 .) Here, for example, Palmer's jury may have shown sympathy or lenience because he was 15 years old at the time of the crimes. The other jury may have felt no such sympathy for the 29 year old Price. Price's verdict must stand or fall on its own merit, not in comparison to Palmer's.

The historical basis for the rule also provides a "damning indictment of its continued use in our system." ( United States v. Espinosa-Cerpa, supra, 630 F.2d at p. 332, fn. 5 .) "The rule undoubtedly was originally simply transplanted from the English system in which it had been applied since the time of Henry IV.... [¶] The history and development of the doctrine that the conviction of a single conspirator may not stand if all his alleged coconspirators are acquitted was reviewed at some length in the relatively recent decision of the House of Lords in R. v. Shannon, [(1974)] 2 All E.R. 1009 (H.L.). The Shannon opinions disclosed that this rule requiring the quashing of such convictions apparently developed at a time, long since past, when review of such proceedings (under writs of error) was restricted to a formal `record' of the indictment, plea, verdict and the like, to determine if there was such an apparent inconsistency or `repugnancy' on the face of the record as required reversal. [Citations.] Most significantly, English appellate tribunals had no capacity to review the evidence to determine whether it was sufficient to support the conviction of the lone conspirator. [Citation.] Thus, the rule emerged merely as a product of this review procedure based on the theory that such an inconsistency more often than not signalled that an error had been made by the jury in the conviction. [Citation.] The original basis and rationale for the rule largely vanished, of course, with the onset of pervasive evidentiary review just after the beginning of the twentieth century. [Citation.] Nonetheless, the rule itself has stubbornly persisted in various forms even to the present. [Citations.] [¶] This exposition in R. v. Shannon of the historical basis for the rule makes abundantly clear its inappropriateness to a modern American criminal justice system in which all verdicts obviously are ... subject to independent review for evidentiary support and where apparently inconsistent verdicts are almost uniformly tolerated whether explainable by variances in proof or merely as jury license." ( Ibid. )

Defendant argues that our decision in People v. Taylor, supra, 12 Cal.3d 686 , 117 Cal.Rptr. 70 , 527 P.2d 622 , compels the rule of consistency. In Taylor, we held that principles of collateral estoppel prohibited the murder trial of a person alleged to be vicariously liable for a killing he did not personally commit when the alleged direct perpetrator had been acquitted of that same murder in an earlier trial. Taylor has no application here. It carefully limited its holding to a case "where an accused's guilt must be predicated on his vicarious liability for the acts of a previously acquited [sic] confederate." ( Id. at p. 698, 117 Cal.Rptr. 70 , 527 P.2d 622 .) The Taylor court identified three purposes for its collateral estoppel rule: "(1) to promote judicial economy by minimizing repetitive litigation; (2) to prevent inconsistent judgments which undermine the integrity of the judicial system; and (3) to provide repose by preventing a person from being harassed by vexatious litigation." ( Id. at p. 695, 117 Cal.Rptr. 70 , 527 P.2d 622 .) None of these purposes applies here. Reversing Price's conviction would not serve judicial economy; this case was litigated once, not repetitively. The rule of consistency cannot prevent an inconsistent verdict that has already occurred; the only question that remains is what to do with that verdict. Nothing about this case hints at vexatious litigation; defendant was tried but once on good, credible evidence. Moreover, occasional inconsistent verdicts do not undermine the integrity of our criminal justice system but are an inevitable consequence of that system. Indeed, a "rule that could promote the duplication of an erroneous acquittal to all persons who participate in a criminal transaction" ( People v. Nunez, supra, 183 Cal.App.3d at p. 224 , 228 Cal. Rptr. 64 ) might itself undermine the system more than accepting inconsistent verdicts once they have occurred.

Price does not claim that any of his jury's verdicts, including the conspiracy conviction, lacks evidentiary support. Accordingly, the Court of Appeal correctly affirmed the conspiracy conviction even though it is logically inconsistent with Palmer's acquittal of that conspiracy. The same rule tolerating inconsistent verdicts also disposes of Price's additional argument that we must reverse the finding that he premeditated Humphries's attempted murder in light of the other jury's finding that Palmer did not premeditate that attempted murder. The Court of Appeal expressed doubt that the premeditation verdicts were actually inconsistent rather than merely different. Like the Court of Appeal, however, we need not decide the question, for even if we assume the verdicts were inconsistent, we will give effect to both for the reasons already stated.

We affirm the judgment of the Court of Appeal and disapprove People v. James, supra, 189 Cal.App.2d 14 , 10 Cal.Rptr. 809 , and its progeny ( ante, 103 Cal.Rptr.2d at pp. 16-17, 15 P.3d at pp. 237-238) to the extent they are inconsistent with this opinion.

GEORGE, C.J., KENNARD, J., BAXTER, J., WERDEGAR, J., and BROWN, J., concur.

Concurring Opinion by MOSK, J.

I concur in the result.

This is an easy case that makes good law. If we are careful.

The first of two general scenarios for inconsistent verdicts is this: A single jury returns inconsistent verdicts either against a single defendant on multiple crimes (e.g., the jury finds the defendant not guilty of selling narcotics, but also finds him guilty of using a telephone to sell narcotics) or against multiple defendants for a single crime (e.g., the jury finds the first of two defendants not guilty of conspiring with the second, but also finds the second guilty of conspiring with the first).

Such inconsistency of verdicts necessarily entails error, albeit apparently not of federal constitutional dimension, on the part of the jury in question, according to decisions including United States v. Powell (1984) 469 U.S. 57 , 105 S.Ct. 471 , 83 L.Ed.2d 461 , and Dunn v. United States (1932) 284 U.S. 390 , 52 S.Ct. 189 , 76 L.Ed. 356 . It is "`error,' in the sense that the jury has not followed the court's instructions" and its own oath. ( United States v. Powell, supra, 469 U.S. at p. 65 , 105 S.Ct. 471 ; see Dunn v. United States, supra, 284 U.S. at p. 393 , 52 S.Ct. 189 .)

The source of such error by a jury may evidently be found in mistake, compromise, or nullification. (See United States v. Powell, supra, 469 U.S. at p. 65 , 105 S.Ct. 471 ; see generally Muller, The Hobgoblin of Little Minds? Our Foolish Law of Inconsistent Verdicts (1998) 111 Harv. L.Rev. 771, 782-786, 803-806 (hereafter Muller).) Mistake by a jury may be somewhat troubling, but hardly greatly, since it is not a willful violation by the panel of the court's instructions or its own oath. (Muller, supra, 111 Harv. L.Rev. at p. 782.) More troubling is compromise on the part of a jury, which is indeed a willful violation by the panel of both instructions and oath, involving, by definition, the jury's return of a guilty or not guilty verdict in spite of the honestly held view of one or more jurors to the contrary. ( Id. at pp. 782-784.) At least as troubling is jury nullification, a "`power'" that the panel, according to Powell and Dunn , has "`no right to exercise'" ( Dunn v. United States, supra, 284 U.S. at p. 393 , 52 S.Ct. 189 ; accord, United States v. Powell, supra, 469 U.S. at p. 66 , 105 S.Ct. 471 ), which similarly amounts to a willful violation by the panel of both instructions and oath. (Muller, supra, 111 Harv. L.Rev. at pp. 784-786, 803-806.) As Powell and Dunn emphasize, nullification may run in favor of the defendant, in the form that is called lenity. ( United States v. Powell, supra, 469 U.S. at p. 65 , 105 S.Ct. 471 ; accord, Dunn v. United States, supra, 284 U.S. at p. 393 , 52 S.Ct. 189 .) But, as they fail to acknowledge, nullification may also run against the defendant, in the form that may be called hostility. (Muller, supra, 111 Harv. L.Rev. at pp. 803-806.) In either direction, it amounts to "illegality." ( United States v. Powell, supra, 469 U.S. at p. 66 , 105 S.Ct. 471 .)

The second of two general scenarios for inconsistent verdicts is this: Multiple juries return inconsistent verdicts against multiple defendants for a single crime (e.g., one jury finds the first of two defendants not guilty of conspiring with the second, but another jury finds the second guilty of conspiring with the first). Such inconsistency does not necessarily entail error on the part of any of the juries in question. The source need be found in nothing more than the possibility that each drew differing inferences from differing evidence. But, inasmuch as mistake, compromise, and nullification cannot be ruled out entirely, neither can error.

As for the first general scenario for inconsistent verdicts, I agree that we should abandon the so-called rule of consistency, which requires virtually automatic reversal of an inconsistent conviction. It is true that, under this scenario, we know that error, arising from mistake, compromise, or nullification, tainted either the inconsistent acquittal or the inconsistent conviction. But we simply do not know whether error tainted the inconsistent conviction rather than the inconsistent acquittal. Lacking such knowledge, we should not reverse the inconsistent conviction automatically. Instead, we should attempt to determine whether error tainted the inconsistent conviction in fact. To find the answer, we must do something morenotwithstanding Powell and Dunn than merely ask the question whether the evidence was sufficient for the inconsistent conviction. (See generally Muller, supra, 111 Harv. L.Rev. at pp. 812-820.) For what we learn from such an inquiry is only whether any rational jury, viewing the evidence in the light most favorable to the government, could have returned the verdict resulting in the inconsistent conviction. What we need to learn, by contrast, is whether the defendant's actual jury did in fact return the verdict in question out of mistake, compromise, or nullification. Of course, we cannot subjectively look into the mind of any individual juror. But we may objectively consider the circumstances under which all of the jurors acted. Hence, the question that we should ask is whether the record on appeal presents such a picture of what transpired both inside of the courtroom, including the trial, and also outside of the courtroom, including the events that were the trial's subject, as to cause us to conclude that mistake, compromise, or nullification had a marginal effect on the verdict resulting in the inconsistent conviction.

As for the second general scenario for inconsistent verdicts, I agree that we should abandon the rule of consistency with its requirement of virtually automatic reversal of an inconsistent conviction. A fortiori. For, under this scenario, we do not know whether error, arising from mistake, compromise, or nullification, tainted either the inconsistent acquittal or the inconsistent conviction or neither. But we also do not know that it did not. As stated, since we cannot rule out the source of error entirely, neither can we rule out error itself. Here, as above, we should ask whether the record on appeal presents such a picture of what transpired within and without the courtroom as to cause us to conclude that mistake, compromise, or nullification had a marginal effect on the verdict resulting in the inconsistent conviction.

The case at bar implicates the second general scenario for inconsistent verdicts. At the time of the crimes, Price was 29 years of age and Palmer was only 15. Palmer's jury returned a verdict finding him guilty of attempting to murder Humphries, and made a finding determining that the attempted murder was not willful, deliberate, and premeditated; it also returned a verdict finding him not guilty of conspiring to murder Showalter altogether with Price. Subsequently, Price's jury returned a verdict finding him guilty of attempting to murder Humphries, and made a finding, arguably inconsistent with Palmer's, determining that the attempted murder was indeed willful, deliberate, and premeditated; it also returned a verdict, surely inconsistent with Palmer's, finding him guilty of conspiring to murder Showalter altogether with Palmer. On Price's challenge, we should ask whether the record on appeal presents such a picture of what transpired within and without the courtroom as to cause us to conclude that mistake, compromise, or nullification had a marginal effect on the verdict resulting in the surely inconsistent conviction of conspiracy to murder Showalter or on the finding resulting in the arguably inconsistent determination that the attempted murder of Humphries was willful, deliberate, and premeditated. We would have to answer in the negative. The presence of overwhelming evidence that Price conspired to murder Showalter and that the attempted murder of Humphries was willful, deliberate, and premeditated would likely bar a conclusion that mistake, compromise, or nullification had a marginal effect on the verdict and finding adverse to Price. (See Muller, supra, 111 Harv. L.Rev. at p. 824.) Evidence there was. But it does not seem overwhelming. With that said, however, there is nothing in the record of appeal that would support a conclusion that mistake, compromise, or nullification had a marginal effect on the verdict and finding against Price. We could no doubt speculate about the matter. Speculation, however, is not enough. Were we to indulge nevertheless, we might conjecture not mistake or compromise, but only nullificationand then only in the form of lenity toward the much younger Palmer on the part of his jury as opposed to hostility toward the much older Price on the part of his.