Swain v. State , 1991 Alas. App. LEXIS 70 ( 1991 )

817 P.2d 927 (1991)

Swanee F. SWAIN, Appellant,
v.
STATE of Alaska, Appellee.
Mark C. ELLIS, Appellant,
v.
STATE of Alaska, Appellee.

No. A-2913.

Court of Appeals of Alaska.

September 20, 1991.

Rex Lamont Butler, Anchorage, for appellant Swain.

Steven E. Sumida, Law Offices of John C. Pharr, Anchorage, for appellant Ellis.

Shelley K. Chaffin, Asst. Atty. Gen., Office of Sp. Prosecutions and Appeals, Anchorage, and Charles E. Cole, Atty. Gen., Juneau, for appellee.

Before BRYNER, C.J., and COATS, J., and ANDREWS, Superior Court Judge.^[*]

OPINION

BRYNER, Chief Judge.

Swanee F. Swain and Mark C. Ellis were convicted by a jury of robbery in the first degree, burglary in the first degree, and assault in the third degree. They moved for a new trial, alleging misconduct by one of their jurors. Superior Court Judge Beverly W. Cutler denied the motion. On appeal, Swain and Ellis contend that the court erred in failing to order a mistrial. We remand.

The state's evidence at trial indicated that, on November 17, 1986, Swain and Ellis entered the Keck residence near Wasilla; they held John Schreiber, who was housesitting, at gunpoint while they stole money, property, and cocaine. The two men were accompanied by Mark Ellis' sister, Tina Ellis. Before leaving the residence, Swain and Ellis tied Schreiber up and knocked him out by hitting him over the head with a large porcelain object — apparently a flower pot. Schreiber did not immediately report the robbery, evidently because drugs were involved.

About two months later, in January of 1987, Ellis, his sister, and another man — not Swain — committed a similar burglary and robbery at a different Wasilla residence. In the course of the crime, Tina Ellis held the homeowner, Carla Schmelzer, and Schmelzer's infant daughter at gunpoint.

Later in 1987, after being implicated in the Schmelzer robbery, Tina Ellis agreed to cooperate with the state and reported participating with Swain and Ellis in the November 17, 1986, burglary and robbery at the Keck residence. The police contacted Schreiber, who confirmed Tina Ellis' account of the incident. Schreiber was unable to identify Tina Ellis' accomplices, however, because he had had insufficient opportunity to view them.

Swain and Ellis were subsequently charged and jointly tried for breaking into the Keck residence and robbing Schreiber. At trial, Tina Ellis and Schreiber were the primary prosecution witnesses. The trial court allowed the jury to hear that Tina Ellis had been implicated in the subsequent burglary and robbery of the Schmelzer residence. However, the court issued a protective order precluding any mention of Mark Ellis' involvement in that offense.

The jury convicted Swain and Ellis on all charges. Prior to sentencing both men moved for a mistrial, claiming among other things that, during trial, one of the jurors, Darcella Perry, had received prejudicial information from extrinsic sources concerning the Schmelzer robbery.

In support of the mistrial motion, Swain and Ellis submitted an affidavit signed by Carla Schmelzer, victim of the Schmelzer robbery. Schmelzer disclosed that she was acquainted with Perry and that Perry occasionally baby-sat for her. According to the affidavit, Schmelzer had originally told Perry about the robbery at her house shortly after it occurred; she did not know at the time, however, who had committed it.

Schmelzer's affidavit further claimed that Schmelzer spoke with Perry about the robbery again on March 20, 1988, while Swain and Ellis' trial was in progress. According to the affidavit, Perry mentioned that she was on jury duty; Schmelzer, who was on standby as a potential prosecution witness in the case, was aware that the trial was in progress. Schmelzer's affidavit stated: "I asked Darcy [Perry] if it was Tina Ellis' trial on which she was to serve. She buried her hands in her face, shocked that I knew. I told her about the robbery of my house by Tina and Mark Ellis. She did not know that that was the case beforehand or she wouldn't have gotten in."

The state opposed Swain and Ellis' motion for mistrial. In opposition to Schmelzer's affidavit, the state filed the affidavit of its trial counsel describing an interview he had with juror Perry. The state also submitted a transcript of an interview of Perry conducted by an Alaska State Trooper. These documents disclosed that Perry acknowledged her acquaintance with Schmelzer and admitted being generally aware, just prior to Swain and Ellis' trial, that Schmelzer had been robbed about a year previously. Perry claimed, however, that she did not remember Schmelzer having described the robbery in detail.

Perry further acknowledged speaking with Schmelzer while Swain and Ellis' trial was in progress. She recalled Schmelzer mentioning that Tina Ellis had robbed her. However, Perry claimed no recollection of Schmelzer saying that Mark Ellis had also been involved in that robbery. According to Perry, as soon as Schmelzer mentioned Tina Ellis' participation in the Schmelzer robbery, Perry told Schmelzer that it would be improper to say anything more about the case. Perry admitted asking Schmelzer not to mention their conversation to anyone. She explained that she did not report her conversation to the court because she thought that information concerning the Schmelzer robbery was irrelevant to the charges for which Swain and Ellis were on trial. Perry insisted that her vote as a juror was not influenced by her communication with Schmelzer. She further gave assurances that she never discussed any of the information she learned from Schmelzer with other members of the jury.

To bolster this latter claim, the state submitted additional affidavits from two other jurors, who confirmed that Perry never mentioned, and the jury never discussed, the possibility of Swain and Ellis' participation in the Schmelzer robbery.

After reviewing the pleadings and hearing oral argument, Judge Cutler denied the motion for a mistrial. The judge found credible Perry's claim that her conversation with Schmelzer had not influenced her decision as a juror. Relying on the affidavits of other jurors, Judge Cutler further concluded that Perry had never discussed the Schmelzer conversation with other jurors. On this basis, the judge concluded that no prejudice had been shown and that Swain and Ellis would not be entitled to relief even if Schmelzer's account of her conversation with Perry were taken as true.

Swain and Ellis moved for reconsideration, requesting an opportunity for an evidentiary hearing. Judge Cutler held an evidentiary hearing, at which Schmelzer and Perry testified. Both witnesses essentially reiterated the positions they took in the initial, written pleadings. At the conclusion of the hearing, Judge Cutler reaffirmed her original decision, stating, in relevant part:

Seeing the juror live say that she really doesn't think it affected her, I don't see any reason to believe that it affected her. There's no showing that the fact that she knew Tina Ellis and maybe Mark Ellis had robbed another residence was that that information was given to any of the other jurors, so the only juror who needs to be examined is her, and examining her it really appears that she's saying it just did not affect her verdict in any way, and I believe that.

On appeal, Swain and Ellis contend that Judge Cutler erred in failing to order a new trial. They argue that Perry's ability to render a fair and impartial verdict was impaired by her exposure to prejudicial information concerning Mark and Tina Ellis' participation in the Schmelzer robbery. They also argue that Perry's failure to disclose her acquaintance with Schmelzer and her conversation concerning the Schmelzer robbery amounts to misconduct depriving them of a fair trial.

These arguments require us to consider two related lines of cases: the first deals with situations in which jurors are exposed to potentially prejudicial matter outside the trial record; the second deals with instances of jury misconduct threatening the integrity of the verdict.

The first line of cases is exemplified by this court's recent decision in Ciervo v. State, 756 P.2d 907 (Alaska App. 1988), a case in which several jurors were exposed to a newspaper story disclosing that the defendant had previously entered and withdrawn a no contest plea to the charge at issue.

Relying on Watson v. State, 413 P.2d 22, 25 (Alaska 1966), an Alaska Supreme Court case involving exposure of jurors to publicity, this court recognized that "[e]ach case of juror exposure to publicity ... must turn on its special facts." Ciervo, 756 P.2d at 910. We went on to refer to the ABA Standards for Criminal Justice as an appropriate guide for determining when the facts of a given case require a mistrial:

The ABA Standards for Criminal Justice set forth the following standard for determining when motions for mistrial should be granted:

On motion of the defendant, the verdict of guilty in any criminal case shall be set aside and a new trial granted whenever, on the basis of competent evidence, the court finds a substantial likelihood that the vote of one or more jurors was influenced by exposure to prejudicial matter relating to the defendant or to the case itself that was not part of the trial record on which the case was submitted. Nothing in this recommendation is intended to affect the rules or procedures in any jurisdiction concerning the impeachment of jury verdicts.

Id. (quoting II Standards for Criminal Justice § 8-3.7 (Approved Draft 1978 & Supp. 1982)).

The Alaska Supreme Court has formulated a slightly different standard for cases involving juror misconduct. The leading Alaska case is Fickes v. Petrolane-Alaska Gas Service, Inc., 628 P.2d 908 (Alaska 1981), in which a juror failed to reveal his acquaintance with an important witness. During deliberations, the juror relied on his acquaintance to vouch for the reliability of the witness.

In deciding Fickes, the supreme court reiterated the two-prong standard it first set forth in West v. State, 409 P.2d 847, 852 (Alaska 1966):

In West v. State, we established the standard for determining whether juror misconduct requires a new trial:

"Whether the verdict should be set aside and a new trial ordered rests in the sound discretion of the trial judge, but generally the verdict should stand unless the evidence clearly establishes a serious violation of the juror's duty and deprives a party of a fair trial." (emphasis added).

In West, we observed that there is a general policy against allowing a juror to impeach the jury's verdict, but that exceptions to this general rule are recognized when the kind of misconduct claimed is "fraud, bribery, forcible coercion or any obstruction of justice."

628 P.2d at 910 (citations omitted).

Considering the first prong of this test, the court found that the juror's failure to disclose his acquaintance with the witness, and his reliance on this acquaintance during deliberations, was "tantamount to an obstruction of justice." Id. Turning to the second prong of the standard — whether the misconduct deprived a party of a fair trial — the court announced three relevant criteria:

Three considerations provide guidelines for making this determination. First, if the party asserting prejudice had known the true facts, is it probable that it would have challenged the juror? Second, did the improper comment merely go toward a collateral matter, e.g., the general credibility of a witness, or did it go to the essence of a claim or defense? Third, viewed objectively, was the probable effect of the comment prejudicial?

Id. at 911 (citations and footnotes omitted).

The Ciervo and Fickes standards — the former governing juror exposure to extraneous prejudicial matter and the latter governing juror misconduct — share a common objective: to assure a fair trial by protecting the integrity of jury verdicts. Thus, while the two standards differ in form, they are essentially similar in substance.^[1] Both uphold the same fundamental precept: "In order for a defendant to receive a fair trial, the jurors must all be impartial." Ciervo, 756 P.2d at 910 (citing Murphy v. Florida, 421 U.S. 794, 799, 95 S. Ct. 2031, 2037, 44 L. Ed. 2d 589 (1975)).

In the present case, Swain and Ellis' challenge to their jury's verdict is susceptible to analysis under both tests. We turn first to the Ciervo test — the standard primarily relied upon by the trial court.

In declining to order a new trial, Judge Cutler relied strongly on Ciervo, which was newly decided at the time. Although Judge Cutler scrupulously adhered to Ciervo, we are compelled to find error; the error stems from a basic flaw in our Ciervo decision, rather than from Judge Cutler's application of the case.

As we have previously noted, in Ciervo we followed ABA Standard 8-3.7 as a guide for determining whether a mistrial should be granted due to a juror's exposure to matter outside the trial record. The standard calls for a mistrial whenever "the court finds a substantial likelihood that the vote of one or more jurors was influenced by exposure... ." II Standards for Criminal Justice, § 8-3.7.

Applying this standard in Ciervo, we found no substantial likelihood that any juror's vote had been influenced. The record established that only one juror had read and understood the potentially prejudicial newspaper article at issue in that case. As to that juror, we emphasized: "The one juror who remembered reading and understanding the prejudicial portion of the article assured the court that the article did not influence her decision." Ciervo, 756 P.2d at 910 (footnote omitted). This passage from our decision plainly implied that, in determining whether a substantial likelihood of influence exists, the reviewing court may properly consider a juror's testimony concerning the subjective impact that the potentially prejudicial materials actually had.

Understandably relying on Ciervo, Judge Cutler, in the present case, emphasized juror Perry's assurances that the information she received from Schmelzer played no part in her decision on the issue of Swain and Ellis' guilt. Since Judge Cutler found Perry's testimony credible, she concluded that Swain and Ellis had failed to establish any reasonable possibility that Perry's conversation with Schmelzer had been prejudicial.

Unfortunately, Ciervo is mistaken in suggesting the propriety of a subjective standard for determining the influence on a juror of exposure to potentially prejudicial extraneous matter. Although the point was lost on us in Ciervo — perhaps because the parties did not raise it or because it would not have been determinative in any event — there is a wealth of authority holding that, for purposes of determining the likelihood that a juror's vote has been influenced, a reviewing court must apply an objective test and is precluded from considering evidence concerning the subjective impact of the extraneous matter on any juror.

This point is made clear by the commentary to ABA Standard 8-3.7, which states, in relevant part:

[T]his standard was drafted to avoid conflict with the universally recognized rule against the impeachment of jury verdicts. That rule bars any inquiry into the mental processes of the jurors concerning their deliberations on the question of guilt or innocence. Accordingly, standard 8-3.7 directs the court's inquiry only to the fact of exposure to extrajudicial material. The ultimate issue of influence on the juror is resolved by reference to the substantial likelihood test, an objective standard. In effect, the court must examine the extrajudicial material and then judge whether it is inherently likely to have influenced the juror. Though this test may seem unduly speculative, it has significant support in the case law, and there appears to be no more precise way to articulate the standard.

II Standards for Criminal Justice commentary to § 8-3.7 at 58.

This commentary finds strong support in the Alaska Rules of Evidence. A.R.E. 606(b) provides, in relevant part:

Upon an inquiry into the validity of a verdict . .. a juror may not be questioned as to any matter or statement occurring during the course of the jury's deliberations or to the effect of any matter or statement upon his or any other juror's mind or emotions as influencing him to ascent to or dissent from the verdict ... or concerning his mental processes in connection therewith, except that a juror may testify on the question whether extraneous prejudicial information was improperly brought to the jury's attention or whether any outside influence was improperly brought to bear upon any juror.

The plain language of A.R.E. 606(b) establishes that jurors may be questioned only as to whether extraneous matters were brought to their attention, and not as to what the effect of any such matter may have been. The language of the rule is substantiated by the rule's commentary, which provides, in relevant part:

Generally there has been agreement among common law jurisdictions that the mental operations and the emotional reactions of jurors during the deliberative process should not be the subject of later inquiry. There has been substantial disagreement as to whether a juror should be able to impeach a verdict in which he participated by testifying about other matters. See 8 Wigmore §§ 2352, 2353, 2354. This rule, like the Federal Rule after which it is modeled, limits impeachment of jury verdicts to inquiries about extraneous prejudicial information and outside influences which may have been improperly brought to bear upon any juror.

Commentary to A.R.E. 606(b).

Moreover, in the closely related area of juror misconduct, the Alaska Supreme Court's decision in Fickes expressly adopts an objective standard for determining the potential impact of misconduct on a jury's verdict. In articulating its three-factor test for determining when misconduct deprives a party of a fair trial, the court in Fickes described the third factor as follows: "Third, viewed objectively, was the probable effect of the comment prejudicial?" Fickes, 628 P.2d at 911 (footnotes omitted). In connection with this consideration, the court went on to note: "The trial court was correct in its view that it cannot inquire into the actual, that is, subjective, impact of [a juror's] statement upon particular jurors." Id. at 911 n. 2 (citations omitted).

Many other jurisdictions have expressly recognized that an objective test is applicable in cases like the present case. They explain that a reviewing court "must determine whether such extraneous information was prejudicial by determining how it would effect [sic] an objective ``typical juror.'" Urseth v. City of Dayton, 680 F. Supp. 1084, 1089 (S.D. Ohio 1987) (citing Owen v. Duckworth, 727 F.2d 643, 646 (7th Cir.1984); United States v. Bassler, 651 F.2d 600, 603 (8th Cir.1981); Miller v. United States, 403 F.2d 77, 83 n. 11 (2d Cir.1968); 3 Weinstein & Berger, Weinstein's Evidence ¶ 606[5] (1985 & May, 1987 cum. supp.)). See also United States v. Hornung, 848 F.2d 1040, 1045 (10th Cir.1988); United States v. Bruscino, 687 F.2d 938, 940-41 (7th Cir.1982) (en banc); United States v. Gerardi, 586 F.2d 896, 898 (1st Cir.1978); Wiser v. People, 732 P.2d 1139, 1141-42 (Colo. 1987) (citing, inter alia, United States v. Vasquez, 597 F.2d 192 (9th Cir.1979); United States v. Howard, 506 F.2d 865 (5th Cir.1975); Louisell and Mueller, Federal Evidence, § 291, at 55 (1986 Supp.)).

The function of the "objective" formula is succinctly explained in United States v. Allen, 736 F. Supp. 914, 918 (N.D.Ill. 1990): "The determination is objective precisely because the trial judge cannot rely upon any testimony of the jurors as to their subjective assessment of the actual impact of the extraneous evidence or influence on their deliberations." In other words, inquiry to determine whether extraneous contact affected the juror is "limited to identification of those extraneous sources of information — once the existence of external influences has been established, neither the Court nor counsel may inquire into the subjective effect of these external influences upon particular jurors." Urseth, 680 F. Supp. at 1089 (citing Owen, 727 F.2d at 646; Bassler, 651 F.2d at 603; Miller, 403 F.2d at 83 n. 11; 3 Weinstein & Berger, Weinstein's Evidence ¶ 606[5] (1985 & May, 1987 cum. supp.)).

In short, we are convinced that our decision in Ciervo is misleading and must be disavowed to the extent that it suggests that courts may properly consider evidence of the actual, subjective impact of extraneous prejudicial matter on a juror's deliberations. The court erred in so doing.^[2]

In the present case, the trial court's decision to deny Swain and Ellis' motion for a mistrial turned precisely on this aspect of Ciervo. As a basis for concluding that no reasonable possibility of juror influence had been established, Judge Cutler placed strong reliance on juror Perry's assurances that her vote had not actually been swayed by her conversation with Schmelzer.

In some situations, mistaken reliance on a juror's subjective statements might well be inconsequential. For example, in the present case, Judge Cutler's emphasis on juror Perry's assurances would at worst have amounted to harmless error if it were nevertheless objectively clear that no substantial likelihood of juror influence existed — that is, if it were apparent that no reasonable juror would have been influenced by the information Perry gained from Schmelzer.

No such objective clarity emerges from the record in this case, however, because it is not certain what Schmelzer told Perry. Apparently because Judge Cutler relied on Perry's subjective assurances to negate the likelihood of any improper influence, the judge did not definitively resolve the primary factual dispute emerging from Schmelzer's and Perry's testimony. Schmelzer unequivocally claimed to have told Perry not only that she had been robbed by Tina Ellis, but further that Mark Ellis had also been a participant in the crime. Perry, in contrast, was more equivocal. She acknowledged being told that Tina had robbed Schmelzer; she further acknowledged that Mark Ellis' name had been mentioned during the mid-trial conversation with Schmelzer. Perry claimed, however, that she did not recall Schmelzer ever specifically saying that Mark Ellis had actually participated with Tina Ellis in the robbery.

The question of whether Schmelzer told Perry of Mark Ellis' involvement in the Schmelzer robbery is, in our view, of crucial significance. During the course of trial, Swain and Ellis' jury was apprised that Tina Ellis had been implicated in a burglary and robbery at the Schmelzer residence and that, in committing the offense, Tina had held Carla Schmelzer at gunpoint. Accordingly, if Schmelzer told Perry only of Tina Ellis' involvement, without describing Mark Ellis' role, Perry would have learned little that was not independently revealed at trial. There would be little objective basis for finding a substantial likelihood that Perry's vote "was influenced by exposure to prejudicial matter that was not part of the trial record on which the case was submitted to the jury." II Standards for Criminal Justice § 8-3.7.

On the other hand, if Schmelzer did expressly disclose Mark Ellis' role in the Schmelzer robbery, Perry would have been privy to significant information that was not presented at trial. Disclosure of Ellis' role in the Schmelzer robbery was specifically precluded by a protective order. Given the similarities between the Schmelzer robbery and the robbery for which Swain and Ellis were on trial, disclosure of Ellis' participation in the Schmelzer case would have strongly suggested his propensity to engage in conduct similar to that with which he was charged. From an objective point of view, it is difficult to imagine that exposure to this type of information would have no influence on a typical juror's vote, particularly when the information comes from an acquaintance of the juror who claims to have been a victim of the defendant's misconduct.^[3]

Moreover, one of the primary lines of defense at trial was that Tina Ellis had fabricated the entire story concerning the burglary and robbery at the Keck residence — a defense rendered at least facially plausible by Schreiber's failure to report the crime. Disclosure that both Tina and Mark Ellis had subsequently committed a similar robbery together would have rendered it far more difficult to accept this line of defense as plausible.

Finally, disclosure of Mark Ellis' participation with Tina in the Schmelzer robbery would immeasurably have enhanced the possibility of prejudice to Swain. Faced with information that both Mark and Tina Ellis had participated in another similar robbery with an unnamed third person, a typical juror might well be tempted to draw the inference that Swain was that third person. This, in turn, would enhance the likelihood that the determination of both Swain and Ellis' guilt might be colored by a perception that they had a propensity for teaming up with Tina Ellis to commit precisely the type of misconduct for which they were charged.

Applying ABA Standard 8-3.7 objectively, we believe that denial of a motion for mistrial in this case would be an abuse of discretion if, as a factual matter, the trial court found that Schmelzer told Perry of Mark Ellis' role in the Schmelzer robbery.

We reach the same conclusion when we apply the alternative Fickes standard governing juror misconduct. The first prong of the Fickes test requires proof of a serious violation of a juror's duty — misconduct akin to fraud, bribery, coercion, or obstruction of justice. Fickes, 628 P.2d at 910. The requirement appears to be satisfied here. Just as a juror's failure to reveal his acquaintance with a witness was found to be tantamount to an obstruction of justice in Fickes, so too was it an obstruction of justice for Perry not to disclose her friendship with Schmelzer or the fact that she had talked to Schmelzer about the Schmelzer robbery. That Perry's reticence may have been inadvertent or well-intentioned does not alter the situation. In Fickes, the supreme court was disposed to find conduct amounting to an obstruction of justice regardless of whether the juror acted intentionally or inadvertently. Id. Here, as in Fickes, regardless of Perry's motivation, her failure to disclose was unfair to the parties in that it deprived them of the opportunity to challenge Perry or to question her further on the issue of actual bias. See id. at 911.

Whether juror Perry's failure to disclose her conversation with Schmelzer deprived Swain and Ellis of a fair trial — the second prong of the Fickes standard — requires consideration of the three factors enumerated in that case. In our view, the outcome of this process depends on how much Schmelzer told Perry.

The first Fickes factor is whether Swain and Ellis would probably have challenged Perry had they been apprised of her conversation with Schmelzer. Id. While the issue might be debatable if Schmelzer and Perry had spoken only of Tina Ellis' participation in the Schmelzer robbery, it seems apparent that a challenge would almost certainly have been exercised if Schmelzer had also told Perry of Mark Ellis' involvement.

Factor two concerns the importance of the information related by Schmelzer to Perry — that is, did the information merely involve collateral issues or did it go to the essence of the case? Id. The general topic of the Schmelzer/Perry conversation falls somewhere between the two extremes of being collateral and essential. Precisely where it falls, however, depends on exactly what was communicated. Disclosure that Perry had learned of Mark Ellis' participation in the Schmelzer robbery would have been far less collateral than mere disclosure of her knowledge of Tina's role in the crime.

The final Fickes factor is the probable effect, in objective terms, of the Schmelzer/Perry conversation. Id. Consideration of this factor leads to the same conclusion that we reached in applying the Ciervo standard. If Perry was specifically told by Schmelzer of Mark Ellis' role, it would, in our view, be difficult to objectively conclude that the information would not have tended to be prejudicial.

In summary, applying both the Ciervo and Fickes approaches to the circumstances in this case, we conclude that the need for a mistrial turns on how much Schmelzer told Perry. Because this issue requires resolution of disputed facts that were not definitively addressed below, we find it necessary to remand the case to the superior court for further findings. We hold that, on remand, a mistrial should be declared unless the superior court determines that Perry was not told of Mark Ellis participation in the Schmelzer robbery.^[4]

This case is REMANDED for further proceedings in conformity herewith.

MANNHEIMER, J., not participating.

NOTES

[*] Sitting by assignment made pursuant to article IV, section 16 of the Alaska Constitution.

[1] Indeed, clear distinctions have not always been drawn between claims of juror misconduct and claims of exposure to potentially prejudicial extrajudicial materials. For example, in Watson v. State, 413 P.2d 22, 24 (Alaska 1966), the supreme court, in considering a case involving exposure of jurors to prejudicial newspaper articles, relied on the standard for juror misconduct that it had previously formulated in West v. State, 409 P.2d 847, 852 (Alaska 1966). The court apparently assumed that both categories of cases were amenable to the same standard.

[2] We emphasize that A.R.E. 606(b) expressly permits the use of juror testimony to determine whether and to what extent "extraneous prejudicial information was improperly brought to the jury's attention or whether any outside influence was improperly brought to bear upon any juror." Thus, in the current case, Judge Cutler acted properly in considering Perry's testimony insofar as Perry indicated that she did not mention her conversation with Schmelzer to any other juror. Similarly, Judge Cutler properly considered the affidavits of other jurors confirming that Perry never mentioned the Schmelzer conversation. Judge Cutler did not err in relying on this evidence as a basis for the objective conclusion that the effect of the Schmelzer/Perry conversation was necessarily limited to Perry and had no influence on any other juror. This determination, however, is not dispositive on the issue of whether a mistrial was called for in the present case, since, as we recognized in Ciervo, "In order for a defendant to receive a fair trial, the jurors must all be impartial." 756 P.2d at 910. The relevant issue here is whether, in objective terms, a substantial likelihood exists that Perry's conversation with Schmelzer influenced Perry's vote.

[3] We note that other courts have found reversible error stemming from exposure of jurors to information relating to a defendant's participation in prior similar misconduct. See, e.g., Dickson v. Sullivan, 849 F.2d 403 (9th Cir.1988). The Alaska Supreme Court has found harmless error when jurors were improperly allowed to hear evidence of other misconduct by the defendant that was dissimilar to the misconduct charged and not otherwise inherently prejudicial. See, e.g., Fields v. State, 629 P.2d 46 (Alaska 1981). Here, however, the similarities between the Schmelzer and Keck robberies clearly make Fields distinguishable.

[4] Our decision to remand makes it unnecessary at this time to resolve Swain's claim that the trial court erred in restricting his cross-examination of Tina Ellis. Swain raises one additional issue, however, that requires our consideration. He argues that the trial court erred in denying his motion for judgment of acquittal. The motion was based on AS 12.45.020, which requires that the testimony of an accomplice be "corroborated by other evidence that tends to connect the defendant with the commission of the crime... ." Swain argues that the testimony of Tina Ellis, an accomplice, was uncorroborated as to his own participation.

We are inclined to agree with Swain that the evidence discussed by the parties in their briefs would not, standing alone, be sufficient to corroborate Tina Ellis' testimony implicating Swain. The parties' briefs, however, have overlooked at least two areas in which evidence corroborates Tina Ellis' testimony. First, the state presented telephone records establishing a series of telephone calls between Ellis' residence in Anchorage and Swain's in Eagle River on the date of the alleged offense. The timing of the calls coincided with that of the robbery. Second, Ellis' father, who testified for the defense and attempted to establish an alibi, indicated that Swain, Tina Ellis, and Mark Ellis had all been together at his home in Anchorage on the night of the offense. Although this evidence was presented as part of the defense case, it may properly be considered in determining the sufficiency of the evidence at trial. Martin v. Fairbanks, 456 P.2d 462 (Alaska 1969), overruled on other grounds, Whitton v. State, 479 P.2d 302 (Alaska 1970). Viewing the testimony of Ellis' father in the light most favorable to the state, the jury could well have relied on it as linking Swain to Tina and Mark Ellis on the night of the offense, thus lending credence to Tina Ellis' testimony that Swain was a participant in the crime.

Considering this evidence in conjunction with the other evidence discussed by the parties in their briefs, we conclude that the totality of the evidence in the record would suffice to induce "a rational belief that the accomplice was speaking the truth" in implicating Swain. Brown v. State, 693 P.2d 324, 329 (Alaska App. 1984) (quoting Oxenberg v. State, 362 P.2d 893, 897 (Alaska 1961)). We find no error in the denial of Swain's motion for a judgment of acquittal.