MARY FIORILLA, PETITIONER V. UNITED STATES OF AMERICA JOHN FIORILLA, PETITIONER V. UNITED STATES OF AMERICA No. 88-401, 88-467 In the Supreme Court of the United States October Term, 1988 On Petitions For A Writ Of Certiorari To The United States Court Of Appeals For The Third Circuit Brief For The United States In Opposition TABLE OF CONTENTS Question Presented Opinion below Jurisdiction Statement Argument Conclusion OPINION BELOW The opinion of the court of appeals (88-401 Pet. App. 1a-15a) is reported at 850 F.2d 172. JURISDICTION The judgment of the court of appeals was entered on June 30, 1988. Petitions for rehearing were denied on July 26, 1988 (88-401 Pet. App. 16a-17a; 88-467 Pet. App. 2a-3a). The petition for a writ of certiorari in No. 88-401 was filed on August 31, 1988. The petition for a writ of certiorari in No. 88-467 was filed on September 16, 1988. The jurisdiction of this Court is invoked under 28 U.S.C. 1254(1). QUESTION PRESENTED Whether the district court properly continued to poll the jury, revealing its numerical division, after a member of the jury expressed disagreement with the announced verdict. STATEMENT Following a jury trial in the United States District Court for the District of New Jersey, petitioners, a husband and wife, were convicted of conspiring to harbor illegal aliens, in violation of 18 U.S.C. 371; nine substantive counts of harboring illegal aliens, in violation of 8 U.S.C. 1324; and two counts of making false statements to officials of the Immigration and Naturalization Service, in violation of 18 U.S.C. 1001. Petitioners were sentenced on the conspiracy count to a three-year term of imprisonment, the execution of which was suspended in favor of three years' probation. Imposition of sentence on the other counts was suspended, and petitioners were ordered to pay $48,000 in fines. /1/ 1. Following petitioners' trial, the jury deliberated for approximately 6 1/2 hours and informed the court that it had reached a verdict (Pet. App. 3a). /2/ The jury returned to the courtroom, and the jury foreman announced verdicts of guilty against the two petitioners on all counts. The foreman also reported that the jury had found co-defendants Thomas Fiorilla and Frances Scullion guilty on some counts and not guilty on others, and reported that one co-defendant, Charles Hirschkind, had been acquitted on all counts (id. at 4a). After the verdict was read, counsel for petitioner John Fiorilla requested that the individual jurors be polled (ibid.). The first nine jurors concurred with the announced verdict. When the trial judge reached the tenth juror, the following colloquy occurred (id. at 4a-5a): The Court: Mr. Fonseca, do you agree or disagree with the verdict against John Fiorilla as I have just reported it? Juror No. 10: I disagree. The Court: You disagree with the verdict as reported? Juror No. 10: Yes. The Court: Did I report it inaccurately? Juror No. 10: Yes. The Court: Let me reiterate what was stated here. As I told you, your verdict in this case had to be 12 nothing. Correct? In other words, the report of the jury here on this verdict sheet was that the jury had found Mr. John Fiorilla unanimously guilty on all counts with which he was charged. Now, what I'm asking you is do you agree or disagree with this verdict as reported here? Do you understand what I am saying? Juror No. 10: Yes, I understand. Yes, to be honest, I have no peace in me. The trial judge then excused the other 11 jurors and returned to his discussion with juror Fonseca (Pet. App. 5a). The Court: Mr. Fonseca, I'd like to hear what you have to say or what you meant by that comment and we will take it from there. When you said you have no peace in you, could you amplify that for us please? Juror No. 10: Sure. When we were up there in the jury room discussing (sic) the defendants were guilty or not guilty, I gave my opinion and everyone gave their opinion and when I gave mine all of them started giving me their point of view and I kind of saw myself, you know, attaked, you know, and so I agreed. But -- The trial judge then reviewed with juror Fonseca the portion of the charge referring to the obligation of the jurors to deliberate with a view to reaching a consensus after consultation with one another but without doing violence to the individual conscience of each (ibid.). Fonseca explained that the final vote did not reflect his conscientious conclusion concerning petitioner John Fiorilla's guilt or innocence (id. at 5a-6a). All counsel then agreed that Fonseca should be questioned concerning his views with respect to the other defendants. Although he agreed with the verdicts exonerating Hirschkind, Fonseca disagreed with the verdicts of guilt against petitioner Mary Fiorilla, and he disagreed with the verdicts of guilt against co-defendants Thomas Fiorilla and Scullion. He agreed with the jury's verdicts of not guilty against the latter two defendants on some of the counts. Pet. App. at 6a. The trial judge then expressed his intention to poll the remaining jurors (Pet. App. 6a & n.3). Petitioners' counsel did not object to the proposed procedure (ibid.). The poll showed that the other jurors agreed with the guilty verdicts against petitioners and their co-defendants, thus revealing an 11 to 1 split. The court then temporarily excused the jury but directed the jurors to return for further deliberations the following morning (id. at 6a-7a). The next morning, the trial judge informed counsel that in United States v. Spitz, 696 F.2d 916 (1983) (per curiam), the Eleventh Circuit had ruled that a trial judge cannot continue a poll once a dissenter has been identified (Pet. App. 7a). Although the judge was not prepared to grant a mistrial on that ground, he explained that he wished to advise counsel of the results of his research (ibid.). The court then directed the jury to continue its deliberations. After two days of additional deliberations, the jury returned verdicts once again convicting petitioners on all counts. Petitioners' co-defendants Thomas Fiorilla and Frances Scullion were acquitted on all counts. A poll of the jury showed that there was no discrepancy between the jury's verdicts and the jurors' individual votes. Ibid. The trial court denied petitioners' post-trial motion for a new trial, in which petitioners argued that a new trial should have been granted because the court had erred in continuing the poll of the jury after Fonseca expressed disagreement with the verdict (Pet. App. 7a). 2. The court of appeals affirmed (Pet. App. 1a-13a). The court rejected petitioners' argument that under Brasfield v. United States, 272 U.S. 448 (1926), it was improper for the court to complete the poll of the jury after one of its members expressed disagreement with the reported verdict (Pet. App. 7a). The court distinguished Brasfield by noting that that case had involved a trial court's inquiry to a jury that had reported a deadlock, rather than a jury poll after an announced verdict as in this case (id. at 9a). The court added that the decision in Brasfield had preceded the adoption of the Federal Rules of Criminal Procedure, which expressly provide in Rule 31(d) for polling the jury after its verdict and returning it for deliberations in the absence of unanimity (Pet. App. 9a). The court also noted that the harmless error rule had been incorporated in the rules governing criminal procedure since Brasfield was decided (id. at 11a-12a). The court disagreed with the Eleventh Circuit's conclusion, expressed in United States v. Spitz, supra, that under Brasfield it is per se reversible error for a court to complete a jury poll after one juror has dissented from the verdict (Pet. App. 12a). The court found that vesting discretion in the trial court in that situation is "consistent with the rules adopted since Brasfield and * * * represent(s) the better approach to an inadvertent disclosure of a jury's division through polling" (ibid.). The court then examined the factors that led the district court to continue the poll after one juror's dissent and found no abuse of discretion (Pet. App. 12a-13a). The court remarked that counsel for petitioners had not objected to the poll, which demonstrated "to some degree the absence of a coercive atmosphere" (id. at 12a). The court also noted that the division in the jury had prompted the judge to dismiss the jurors for the rest of the day, thus removing the dissenter "from the immediate 'attacks' of his peers" (id. at 12a-13a). The following morning, the court observed, the judge gave a cautionary instruction that each juror should weigh carefully the views of his or her peers (id. at 13a). Two days of further deliberations ensued before the jury returned unanimous verdicts of guilty against petitioners (ibid.). The court pointed out that, as the district court had noted, "this lapse of time * * * demonstrated that the lone dissenting juror was not coerced by the poll to capitulate to the views of the majority" (ibid.). The court added that after its further deliberations the jury had acquitted petitioners' two co-defendants. Since juror Fonseca had been the only vote for acquittal of those defendants at the time the verdicts were first announced, the court observed that "(h)is views on these two defendants in the end carried the day." (Ibid.). These factors made it "plain * * * that Mr. Fonseca was not afraid to stand up for his convictions and that his peers took his opinions seriously during their two days of further deliberations" (ibid.). /3/ ARGUMENT 1. Petitioners contend (88-401 Pet. 8-9; 88-467 Pet. 8-10) that the district court's continuation of the jury poll following one juror's dissent is contrary to the decision of this Court in Brasfield v. United States, 272 U.S. 448 (1926), and constitutes per se reversible error. In Brasfield v. United States, supra, the trial judge, after deliberations had stalled, inquired how the jury was divided and was informed that it stood nine to three. The jury resumed deliberations and subsequently found the defendants guilty (272 U.S. at 449). This Court held that the inquiry into the jury's division after it had announced its failure to agree required reversal because "in general its tendency is coercive" and it almost always brought to bear "in some degree, serious although not measurable, an improper influence upon the jury" (id. at 450). /4/ The court of appeals properly refused to extend the holding of Brasfield from the deadlocked jury setting to the polling of a jury that has announced a verdict. Identifying the number of holdouts on a deadlocked jury is wrong, the Court explained in Brasfield, because it may bring an "improper influence" to bear on the jurors and because the practice is "never useful" (272 U.S. at 450). /5/ A jury poll, by contrast, has as its purpose to verify publicly that the jury's report of a unanimous verdict reflects the vote of each juror without coercion. Humphries v. District of Columbia, 174 U.S. 190, 194 (1899); United States v. Lopez, 581 F.2d 1338, 1341 n.2 (9th Cir. 1978); Miranda v. United States, 255 F.2d 9, 17 (1st Cir. 1958). The right to a jury poll is guaranteed by Fed. R. Crim. P. 31(d). If a dissent emerges during the poll, it is inevitable that a split among jurors -- and the identity of at least one of the dissenters -- is revealed. In that situation, Fed. R. Crim. P. 31(d) authorizes the trial judge to declare a mistrial and discharge the jury, or to require the jury to resume its deliberations in an effort to reach a conclusive verdict. By authorizing the resumption of deliberations after a poll has revealed a disagreement -- even though the poll may reveal the identity of the dissenter and the fact that the jury is heavily disposed in favor of conviction -- Rule 31(d) indicates that it is inappropriate to extend the per se rule of error adopted in Brasfield to the jury poll setting. /6/ Thus, even if it is the twelfth juror polled who turns out to be the dissenter, the court is not required to declare a mistrial, even though in that situation, as in Brasfield, the division of the jury will have become public, and the potentially coercive effect of that revelation will be at least as great as in the case of a jury that has announced that it is deadlocked. Because Rule 31(d) accords district courts "discretion in assessing the impact of a dissenting vote during a jury poll" (United States v. Brooks, 420 F.2d 1350, 1353 (D.C. Cir. 1969)), it should be permissible for a court to complete the poll of the jurors or to repoll the jury altogether after some disagreement has surfaced, if that procedure would help clarify the confusion regarding the verdict or inform the court's discretion as to whether to discharge the jury or require the jury to resume its deliberations. See United States v. Morris, 612 F.2d 483, 489 (10th Cir. 1979) (repolling the jurors permissible where confusion arises in multi-defendant, multi-count case); Amos v. United States, 496 F.2d 1269 (8th Cir.) (completion of poll on count which dissenter agreed with verdict), cert. denied, 419 U.S. 896 (1974); United States v. Brooks, supra (completion of poll at defense counsel's request following indication of doubts by one juror); Williams v. United States, 419 F.2d 740 (D.C. Cir. 1969) (en banc) (completion of poll following confused response by one juror); Cook v. United States, 379 F.2d 966 (5th Cir. 1967) (repolling proper where one or more jurors indicates he voted for conviction based on a recommendation of leniency). To be sure, the district court is not required to complete the poll once one of the jurors makes an unequivocal expression of dissent, and in most cases the better practice is for the court to terminate the poll at that point. See United States v. Aimone, 715 F.2d 822 (3d Cir. 1983), cert. denied, 468 U.S. 1217 (1984); United States v. Freedson, 608 F.2d 739 (9th Cir. 1979). But there are cases, such as this one, in which continuing the poll does not add significantly to the risk of coercing the dissenting juror, and in which continuing the poll can have benefits, such as permitting the court to take a partial verdict if there is unanimity on the jury with respect to some defendants or some counts. In such a case, we submit that the better rule is to give the district court discretion to continue the poll if, in the court's judgment, the benefits of continuing the poll outweigh the risks of coercing the dissenting juror or jurors to acquiesce in the verdict. Even if the district court's action in continuing the poll is found to be error, it should be subject to the harmless error rule. Although the Court in Brasfield refused to apply harmless error principles to inquiries regarding the division of a deadlocked jury, developments in the law since the time of that decision counsel against extending it to a different context, such as the jury poll context at issue in this case. Twenty years after the decision in Brasfield, the Court promulgated Rule 52(a) of the Federal Rules of Criminal Procedure, which has been construed to require harmless error analysis of almost any statutory or constitutional violation that is asserted as the basis for reversal of a conviction. See Pope v. Illinois, No. 85-1973 (May 4, 1987), slip op. 5-6 & n.7; Rose v. Clark, 478 U.S. 570, 576-579 (1986). In particular, where there is no statutory or constitutional prohibition, this Court has held that "a federal court may not invoke supervisory power to circumvent the harmless error inquiry prescribed by Federal Rule of Criminal Procedure 52(a)." Bank of Nova Scotia v. United States, No. 87-578 (June 22, 1988), slip op. 4; see also United States v. Hasting, 461 U.S. 499, 506 (1983). Because extending the rule of Brasfield to the context of jury polls would be justified, if at all, only under the courts' supervisory powers, this Court's recent decisions in Bank of Nova Scotia and Hasting indicate that the court of appeals was correct in applying harmless error principles to the district court's action in this case. 2. Petitioners also contend (88-401 Pet. 8-13; 88-467 Pet. 11-15) that the decision of the court of appeals in this case is contrary to the decision of the Eleventh Circuit in United States v. Spitz, 696 F.2d 916 (1983). In Spitz, the eighth juror in the poll stated that she did not concur in the announced verdict of guilty (id. at 917). After the remaining four jurors assented to the verdict, the judge instructed the eighth juror to stand, verified that she did not concur in the verdict, and gave the jury a supplemental "Allen" charge (ibid.). The jury returned with a unanimous verdict of guilty half an hour later (ibid.). The Eleventh Circuit reversed. The court stated that absent exceptional circumstances, it is "per se error" requiring reversal for a trial judge to continue a jury poll following the expression of dissent by one juror (id. at 917-918). We agree that the per se rule adopted in Spitz conflicts with the holding in the present case that a trial court's continuation of a poll after a juror dissents should be reviewed for abuse of discretion. For the reasons set forth above, we believe that such a per se rule is incorrect, as it misconstrues Brasfield, the Federal Rules of Criminal Procedure, and the precedents on which the court relied. /7/ Nevertheless, the decision in this case is distinguishable from Spitz on its facts and the issue presented in the two cases is a very narrow one of only limited practical importance. Accordingly, we submit that the conflict in the analysis of the two decisions does not warrant this Court's review. On its facts, the Spitz case presented a greater risk of coercion than this case, and the continuation of the poll was harder to justify. First, in Spitz the court followed the completion of the poll with an Allen charge, which could have been interpreted by the sole dissenting juror as specifically directed to that juror to reconsider her views. In this case, by contrast, the district court gave no Allen charge, but merely gave an extremely mild "duty to deliberate" charge (see C.A. App. JA 1809). Second, the jury in Spitz returned with a verdict of guilty after only a half hour of further deliberations, which suggests that the court's procedure may well have had a coercive effect on the dissenting juror. In this case, by contract, the jury deliberated for an additional two days, a clear indication that the court's procedure did not break the dissenting juror's resistance. Third, when the jury in this case returned its verdict, it changed the previously announced verdict in several respects, acquiting two persons who had been found guilty in the jury's initial verdict. Fourth, this case, unlike Spitz, was a multi-defendant case. The district court continued the poll in part because it hoped that it could take a partial verdict as to some defendants on some counts, and in fact it did so. That justification was not present in Spitz. In sum, although we believe that the Spitz court was wrong to adopt a per se rule of reversal for all cases in which a court continues a jury poll after one juror dissents from the verdict, the results in this case and Spitz are reconcilable on their facts. In light of the special dangers of coercion and the absence of any justification for continuing the poll in Spitz, the district court's action in that case could be viewed as reversible error in any circuit. Because the district court's action in this case was justified by a legitimate interest in obtaining a partial verdict, and because the facts disclosed no evidence of coercion, the continuation of the poll in this case can be regarded as within the district court's discretion or, at most, harmless error. Even if the decisions in this case and in Spitz were truly in conflict, we submit that the conflict would not be of sufficient importance to warrant this Court's review. To our knowledge, no other court of appeals has faced the precise question presented here and in Spitz since the adoption of Rule 31(d) in 1946. /8/ Although the District of Columbia and Eighth Circuits have dealt with related issues -- and declined to reverse convictions where jury polls were continued after an expression of dissent -- those cases involved special circumstances that distinguish them from this case and Spitz. In the Eighth Circuit case the court continued the poll only on a count as to which there was no expression of dissent, see Amos v. United States, supra, and in the District of Columbia Circuit cases, the poll was continued where defense counsel specifically requested it (see United States v. Brooks, supra), or where the juror in question gave a confused response that did not make clear whether that juror meant to dissent from the verdict (see Williams v. United States, supra). None of the other cases cited by petitioners (88-401 Pet. 9; 88-467 Pet. 14) are even arguably in conflict with the decision in this case or in Spitz. /9/ There are several reasons why the issue presented in this case arises so infrequently. First, dissents during jury polls are quite rare. Obviously, once a jury has announced agreement, it is unusual for some juror to change his mind during the poll. Second, the usual practice when a dissent emerges during the poll is for the court to terminate the poll and require the jurors to resume their deliberations. Only in rare instances do district courts continue the poll after one juror dissents. Third, the issue does not arise in cases in which the reviewing court, applying the Third Circuit's rule, finds the continuation of the poll to have been coercive, as the Third Circuit might well have found on the facts of Spitz. The competing rules adopted by the Third and Eleventh Circuits will produce a different result only in those rare instances in which a court continues a poll after an unequivocal expression of dissent, on a count as to which the dissenting juror disagreed with the verdict, and where defense counsel has not requested the continuation of the poll. In those rare cases, any resulting conviction is per se invalid in the Eleventh Circuit, while in the Third Circuit it may be affirmed if the reviewing court concludes that the practice did not result in coercion on the facts of the particular case. As a consequence of those competing rules, district courts in the Eleventh Circuit may be more reluctant to continue jury polls than district courts in the Third Circuit. But unless the issue begins to rise with more frequency than it has to date, it does not affect enough cases to justify this Court's intervention. Finally, the decision in this case works no unfairness to petitioners because, even if the inquiry that resulted in revelation of the jury's alignment was improper, the error was plainly harmless. As the court of appeals observed (Pet. App. 12a-13a), immediately following the disclosure of the one juror's dissent the jury was excused for the day, which gave the jury and the dissenting juror a "cooling-off" period. After reconvening, the jury deliberated for an additional two days before returning guilty verdicts against petitioners. The fact that the dissenting juror felt no coercive effect as a result of earlier disclosure of his lone dissent is confirmed by his success in convincing his peers to acquit two of petitioners' co-defendants on all charges -- as he had favored in open court. Ibid. Finally, petitioners' counsel did not object to the continued poll at the time the court proposed it. While the court of appeals did not decide this case on waiver grounds (id. at 6a), this Court has noted that the "omission (of an objection of counsel) indicates that the potential for coercion argued now was not apparent to one on the spot" (Lowenfield v. Phelps, No. 86-6867 (Jan. 13, 1988), slip op. 9 (footnote omitted)). On this record, any alleged error by the trial court in continuing the poll could not have resulted in denial of a fair trial to petitioners. CONCLUSION The petition for a writ of certiorari should be denied. Respectfully submitted. CHARLES FRIED Solicitor General EDWARD S.G. DENNIS, JR. Assistant Attorney General JOHN F. DE PUE Attorney OCTOBER 1988 /1/ Three co-defendants, Thomas Fiorilla, Frances Scullion, and Charles Hirschkind, were also charged with conspiracy and harboring illegal aliens. The three were acquitted on all counts. /2/ "Pet. App." refers to the appendix to the petition in No. 88-401. /3/ The court also indicated that appellate courts should accord discretion to trial judges to respond to exigencies arising in complex, multi-defendant cases, and accepted the district court's explanation that its continued poll was an attempt to take a partial verdict "in a relatively complex multi-count, multi-defendant criminal prosecution" (Pet. App. 13a). /4/ The Court noted that in Burton v. United States, 196 U.S. 283, 307 (1905), it had "condemned the practice of inquiring of a jury unable to agree, the extent of its numerical division" (Brasfield, 272 U.S. at 449). The Court had explained in Burton that "(c)ases may easily be imagined where a practice of this kind might lead to improper influences" (196 U.S. at 308). In Brasfield, the Court resolved a conflict amont courts of appeals as to whether the inquiry criticized in Burton was reversible error (Brasfield, 272 U.S. at 449-450). /5/ Courts have noted that inquiries about the division of the jury after a reported deadlock may exert an undue influence on minority jurors, who could interpret the judge's supplemental instruction for jurors to reconsider their views as being directed specifically to them. United States v. Rogers, 289 F.2d 433, 435 (4th Cir. 1961); United States v. Sae-Chua, 725 F.2d 530, 532 (9th Cir. 1984). /6/ As the court of appeals noted (Pet. App. 10a n.5), this Court in Lowenfield v. Phelps, No. 86-6867 (Jan. 13, 1988), did not extend Brasfield to jury polls conducted under Fed. R. Crim. P. 31(d) or other analogous authority. The Court noted the "potential dangers of jury polling" (slip op. 8), but it did not suggest that those dangers should prohibit a poll where the rules require it, nor did it suggest that the per se rule of Brasfield should be extended beyond the context of deadlocked juries. Lowenfield, like Brasfield itself, involved inquiries to a jury that had reported an inability to reach a decision. /7/ No case cited by the court in Spitz reversed a conviction because a trial judge had continued a poll after a juror dissented. In Sincox v. United States, 571 F.2d 876 (5th Cir. 1978), the court of appeals reversed because the trial court had entered a conviction despite a juror's dissent (571 F.2d at 876). The court never discussed the propriety of completing a poll. United States v. Edwards, 469 F.2d 1362 (5th Cir. 1972), like Sincox, involved the entry of a verdict over a dissent, not a poll completed after a dissent, followed by further deliberations. /8/ Petitioners assert (88-404 Pet. 11; 88-467 Pet. 13) that the holding of the court below conflicts with its earlier decision in United States v. Aimone, 715 F.2d 822 (3d Cir. 1983), cert. denied, 486 U.S. 1217 (1984). Even if there were a conflict within the circuit, that would provide no basis for granting certiorari. But in any case, as the court below observed (Pet. App. 10a n.5), the question presented in this case was not at issue in Aimone, and the passing reference to the Spitz rule was merely dictum, which the court of appeals chose not to follow. /9/ Most of the cases cited by petitioners involve inquiries into the division of a jury that has reported an inability to agree -- the same situation faced in Brasfield v. United States, supra. See Cook v. United States, 254 F.2d 871 (5th Cir. 1958); Jacobs v. United States, 279 F.2d 826 (8th Cir. 1960); United States v. Rogers, 289 F.2d 433 (4th Cir. 1961). The decision in United States v. Sae-Chua, 725 F.2d 530 (9th Cir. 1984), involved a variation on that theme. In that case, after the jury foreman had advised the trial judge that there was only one juror who persisted in voting not guilty, the trial judge gave a supplemental charge, which the court of appeals found to be coercive. Two cases cited by petitioners involve state convictions, to which Brasfield's supervisory rule is inapplicable. Cornell v. Iowa, 628 F.2d 1044 (8th Cir. 1980), cert. denied, 449 U.S. 1126 (1981); United States ex rel. Anthony v. Sielaff, 552 F.2d 588 (7th Cir. 1977) (upholding conviction after state judge polled deadlocked jury and instructed the jurors to resume deliberations). Finally, United States v. Essex, 734 F.2d 832 (D.C. Cir. 1984), had nothing to do with polling or inquiring about numerical division, as it involved a conviction by an 11-person jury when the 12th juror failed to show up.