This opinion is subject to formal revision before publication in the Atlantic and Maryland Reporters. Users are requested to notify the Clerk of the Court of any formal errors so that corrections may be made before the bound volumes go to press.

BLACKBURNE-RIGSBY, Associate Judge.

Before us is an appeal challenging the trial court's actions following an aborted jury poll. Appellant Robert Leake was convicted of carrying a pistol without a license, unlawful possession of a firearm, possession of an unregistered firearm, and unlawful possession of ammunition.1 Appellant contends that the trial court abused its discretion by failing to perceive the inherent potential for jury coercion following a juror's dissent in open court, and by relying on improper factors when it denied appellant's motion for a mistrial and required the jury to continue its deliberations. Appellant argues that, taken together, the trial court's actions require reversal. We disagree and conclude that reversal is not required because in this case the potential for coercion was only minimal and the trial judge's actions neutralized the coercive potential.

I. Factual Background

Appellant's charges arise out of a traffic stop that occurred on September 14, 2010, when Metropolitan Police Department Officers required appellant and two other individuals, including the driver and passenger, to step out of the vehicle during the stop. Once one of the officers asked appellant to step out of the vehicle, he noticed a bulge in appellant's waistband. Appellant was subsequently arrested for possessing a handgun.

Appellant's first trial ended in a deadlocked jury that ultimately led to a mistrial. At appellant's second trial, the jury began its deliberations on February 3, 2011, at 4:06 p.m. Thirty minutes into its deliberations, the jury sent its first note to the trial judge requesting fingerprint cards and a map. The judge granted the request without any objection. The jury continued deliberations the next day and sent a second note with two questions: "What is the legality of removing occupants of a vehicle following a traffic stop for a simple moving violation?" and "What legal basis do the officers have for removing a vehicle's occupants and handcuffing them?" The jury foreman stated in the note that these "are questions that we feel need to be answered in order to satisfy some [j]ury members in reaching a verdict." After conferring with counsel, the trial judge responded in writing: "You do not have to decide whether the officers' actions following the traffic stop were lawful. You are permitted to consider the actions of the officers in assessing the credibility of the officers with respect to the question before you: whether the [d]efendant is guilty beyond a reasonable doubt."

At 3:02 p.m., the jury sent a note that it had reached a verdict. The foreperson announced that the jury had found appellant guilty of all charges. Upon appellant's trial counsel's request, the trial judge polled each of the jurors. When asked whether they agreed with the verdict, Jurors One and Two answered in the affirmative, but Juror Three answered: "Sort of yes — I mean, no. Not too much." The trial judge immediately stopped polling the jurors, stating that: "It's important that all of you agree. If there's a question about it, I'm going to ask you — I'm going to excuse you and ask you to continue deliberations until — is there anything else you want — do either of the attorneys want to approach?"

During a bench conference with counsel, both sides agreed not to submit a Winters anti-deadlock instruction because under the circumstances, the instruction would be inappropriate. However, appellant's trial counsel requested that the trial judge provide a "brief instruction to remind [the jury] that [the verdict] has to be unanimous." Ultimately the trial judge instructed the jury as follows: "I'm going to return the verdict form to the foreperson and I'm going to ask you to resume your deliberations and let me know when you've reached a verdict or if you have any more questions." The jury then continued its deliberations.

At 3:10 p.m., several minutes after returning to the jury room, the jury submitted an additional question to the trial judge asking how long fingerprints last. After the trial judge summoned counsel, appellant's trial counsel moved for a mistrial claiming that it would be coercive to send the jurors back to deliberate with the aim of reaching a unanimous verdict after a juror had dissented openly in court. The trial judge denied the motion, observing: "I didn't tell them to reach a unanimous verdict, I told them to go back and continue their deliberations, and that will mean if they've reached a verdict or if they have any additional questions." In response to appellant's trial counsel's claim of inherent coercive potential, the judge found that "[t]he [j]ury is in the same situation as it would have been in if the juror had said in the jury room I agree and then changes his mind there." The judge then responded to the jury's question regarding fingerprints in writing: "The jury must rely on the evidence presented at trial. The jury's recollection of the evidence controls."

At 3:44 p.m., the jury alerted the court that it had reached a verdict. Prior to the jury's return to the courtroom, appellant renewed his motion for a mistrial. Appellant informed the court that Juror Three had indicated during voir dire that he had some scheduling concerns due to child care issues. Citing Harris v. United States, 622 A.2d 697 (D.C. 1993), appellant argued that the instruction to the jury to continue deliberations after Juror Three's open dissent in court, combined with the inevitable prolonging of the deliberation process, and its impact on Juror Three's childcare concerns, presented additional elements of jury coercion.

The trial judge delayed ruling on the motion and brought the jury into the courtroom because he did not want to keep the jurors waiting. The jury found appellant guilty on all four counts. The trial judge then polled the jurors, all of whom agreed with the verdict. After dismissing the jury, the trial judge responded to appellant's motion for a mistrial, observing that Juror Three had expressed child care concerns during voir dire but that "he didn't say it wasn't possible" to find child care. The trial judge additionally observed that, when polled the second time, Juror Three "answered yes, kind of straightforward. He didn't seem to be equivocal . . . [and] there was no sign either he felt pressured into reaching a verdict by reasons of time or anything else."

The trial judge recognized that it retained discretion in determining whether to grant a mistrial and noted that "[t]his is clearly a situation where it is — you know, it's clear at least in some substantial degree that the split was, you know, probably eleven to one." The court also stated that it could "take into account the question that preceded by a short period of time the initial note that they'd reached a verdict," which the court observed "was more about the conduct of the police officers than about the conduct of Mr. Leake." Lastly, the court commented on the timing between the jury's return to deliberations and the verdict:

The trial judge denied the second motion for a mistrial and remarked that the jury's second note about the question of fingerprints indicated that "the [j]ury wasn't just focused on the issue of police misconduct," which the judge found to be consistent with "talking through a range of issues."

II. Discussion

Appellant argues that reversal is required because the trial court abused its discretion by failing to recognize the potential for jury coercion and by improperly considering certain factors when it ordered the jury to continue deliberations. Appellant cites to (James) Johnson v. United States, 398 A.2d 354 (D.C. 1979), which provides that in assessing whether the trial court abused its discretion, this court looks to whether the trial judge "failed to consider a relevant factor, . . . relied upon an improper factor, and whether the reasons given reasonably support the conclusion. Id. at 365 (citation omitted). The government contends that the trial court did not abuse its discretion by failing to order a mistrial because the circumstances were not sufficiently coercive, and because the trial judge's reaction ameliorated any potential for coercion. In assessing appellant's claim, we first discuss the legal framework used to assess coercion in jury poll cases.

A.

The jury poll serves as "the primary device for discovering the doubt or confusion of individual jurors." Crowder v. United States, 383 A.2d 336, 340 (D.C. 1978) (citations omitted). "Its purpose is to determine . . . that every juror approves of the verdict . . . and that no juror has been coerced" into agreeing with the verdict. Id. Under Super. Ct. Crim. R. 31 (d), when the jury returns a verdict, the trial judge may poll the jury at the request of one of the parties or upon its own motion. Id. "If upon the poll there is not unanimous concurrence, the jury may be directed to retire for further deliberations or may be discharged."2 Id. The trial court is vested with the discretion to assess any dissent made by a juror during a jury poll because it is in the best position to determine whether the juror freely consented to the verdict and whether to require subsequent deliberations. Green v. United States, 740 A.2d 21, 26 (D.C. 1999).

"An inquiry into jury verdict coercion is made from the perspective of the jurors." Harris, supra, 622 A.2d at 701 (citation omitted). Any alleged coercion "must be evaluated in context and with regard to all of the circumstances of the case." (Tommie) Johnson, 360 A.2d 502, 504 (D.C. 1976) (citation omitted). Evaluation of jury coercion requires this court to inquire into: (1) "the inherent coercive potential before the trial court"; and (2) "the actions of the trial judge in order to determine whether these actions exacerbated, alleviated or were neutral with respect to coercive potential." Harris, supra, 622 A.2d at 701. The two factors must then be viewed jointly "to assess the possibility of any actual coercion on any juror or jurors."3 Id. at 701-02. We will find reversible error "where it is necessary to achieve a proper decision" but will allow the trial court's exercise of discretion to stand where its determination caused no significant prejudice. (James) Johnson, supra, 398 A.2d at 366.

B.

1. Potential for Coercion

We begin by assessing the degree of inherent coercive potential facing the trial court. In making this assessment, we look to a series of indicators, including: (1) the extent of isolation of a dissenting juror; (2) whether the identity of a dissenting juror is revealed in open court; (3) whether the exact division of the jury's verdict is disclosed; (4) whether the judge is aware of the identity of the dissenting juror; (5) whether the dissenting juror knows of the judge's awareness; (6) whether other jurors feel "bound" by a verdict they announced; and (7) whether the trial court issues an "anti-deadlock" instruction. Harris, supra, 622 A.2d at 705.

In the present case, the third juror in the jury poll announced his dissent in open court. We have recognized that the potential for coercion is minimal in cases where the juror makes his dissent early in the polling because the positions of the remaining jurors are not revealed, thus minimizing any degree of isolation that might otherwise attach to the dissenting juror. Elliott v. United States, 633 A.2d 27, 36 (D.C. 1993). In Harris, for example, where the twelfth juror dissented, we observed that:

622 A.2d at 703 (citing Crowder, supra, 383 A.2d at 343 n.14).

Additionally, the trial judge ended the jury poll immediately after Juror Three dissented, which prevented disclosing a clear division within the jury. Cf. In re Pearson, 262 A.2d 337, 338-39 (D.C. 1970) (concluding that the trial judge's continued polling after the first juror dissented in open court served no "useful purpose" and instead revealed the jury's split in an already "magnified" coercive atmosphere). The trial judge also carefully instructed the jury to answer only yes or no when polled, avoiding any potential for a juror to provide a basis for his dissent in open court. See Green, supra, 740 A.2d at 24, 30 (reduced potential for coercion where the trial judge carefully instructed the jury to only indicate whether the juror agreed with the verdict).

Here, after the aborted jury poll, the judge posited to counsel that he thought the split was "probably eleven to one." Notably, however, the judge's observation regarding the jury's division was made outside the presence of the jury. For that reason, the jury had no way to know that the judge felt Juror Three was the only juror who had dissented to the verdict, which reduced any possible isolation Juror Three might have otherwise experienced. See Artis v. United States, 505 A.2d 52, 58 (D.C. 1986) (observing that despite the trial court's decision to continue polling on some counts, there was minimal inherent coercive potential because the dissent came first in line and all further polling ceased on that count).

Although the inherent coercive pressure on Juror Three after he dissented in open court may not have been as slight as it would have been had he changed his mind in the jury room, in this case the degree of inherent coercive potential was not as great as some of our other jury coercion cases. Compare Green, supra, 740 A.2d at 29 (recognizing only minimal inherent coercive potential where the jury poll was aborted after the eighth juror made his dissent known in open court), and Elliott, supra, 633 A.2d at 36 (noting how the seventh juror's early dissent in the polling avoided revealing a jury split and isolating the dissenting juror), with Harris, supra, 622 A.2d at 705-06 (acknowledging the increased potential for coercion where the twelfth juror openly dissented and revealed herself as the lone dissenter but ultimately finding no actual coercion), and Crowder, supra, 383 A.2d at 343 (high degree of coercive potential where the twelfth juror registered her dissent in open court, revealing the jury's numerical split, and unequivocally dissented from the verdict as to a specific charge because of the lack of evidence).

Additionally, the judge polled the jury on the verdict rather than on each charge, thus making it impossible to know whether Juror Three dissented on one, some, or all four of appellant's counts. Juror Three also did not provide a basis for his dissent. In Harris, this court ultimately determined that, notwithstanding the twelfth juror's dissent in open court, there was no prejudice because the twelfth juror indicated that she disagreed with a verdict that pertained to two defendants. 622 A.2d at 706 (noting that it was unclear whether the twelfth juror's dissent related to appellant's verdict, as opposed to appellant's co-defendant's verdict, which had not yet been polled). In sum, on the continuum of inherent coercive potential, the circumstances of this case place it closer to the less coercive cases that this court has encountered.

2. Trial Court's Actions

We next assess the trial judge's actions to determine "whether these actions exacerbated, alleviated or were neutral with respect to coercive potential." Harris, supra, 622 A.2d at 701. Specifically, we look to (1) whether the judge made any affirmative efforts to dispel the coercive potential, (2) whether the judge's actions took a middle ground, (3) whether the judge's actions exacerbated the problem by effectively contributing to the potential for jury coercion, and (4) whether the judge's reaction independently created a coercive atmosphere for the jury. Id. at 705. Then, we view the coercive potential and trial judge's actions jointly to determine whether there is any actual coercion. Id. at 701-02.

Here, the trial judge's instruction to the jury was neutral: "I'm going to ask you to resume your deliberations and let me know when you've reached a verdict or if you have any more questions." This case is therefore similar to Green, where we concluded that the trial judge's instruction sending the jury back for further deliberations without further comment was appropriate and did not warrant an additional instruction4 to the jury because the judge took an "essentially neutral course of action" when faced with a "run-of-the-mill polling breakdown." Green, supra, 740 A.2d at 30.

Appellant characterizes the trial court's instruction as insisting that "reaching a verdict was the paramount if not the only goal." However, the language used by the trial judge does not indicate that he was requiring further deliberations in order to eliminate the third juror's dissent. Had the court done so, we would have cause for concern. See Crowder, supra, 383 A.2d at 342 n.11 (observing that further deliberations are particularly problematic where a twelfth juror reveals his dissent in open court because the lone juror may perceive the judge's instruction as a means of eliminating his dissent). Instead, the judge left room for an outcome other than reaching a unanimous verdict because it charged the jury with either trying to reach a verdict or coming back to the court with questions.5 Appellant contends that the trial judge was required to either declare a mistrial or take additional action to safeguard appellant's rights, such as issuing a Crowder instruction.6 We disagree. A Crowder instruction is not required in every case involving a jury poll breakdown. See Green, supra, 740 A.2d at 28-29 (stating that a Crowder instruction is not intended for routine use after a breakdown in a jury poll). Rather, as the Commentary to Instruction 2.603 (formerly 2.93) in the District of Columbia's Red Book notes, a Crowder charge is recommended by this court "for use in cases where there is a particularly high likelihood of juror coercion."7 Criminal Jury Instructions for the District of Columbia, No. 2.603 (5th ed. 2012). Though we recognize that a Crowder instruction could have reduced the inherent coercive potential, here, a Crowder instruction was not required because the trial judge faced only a nominal level of coercive potential. See Brown v. United States, 59 A.3d 967, 974 (D.C. 2013) (explaining that a Crowder instruction is "[t]he best instruction" a trial court can give when the potential for coercion is high); Harris, supra, 622 A.2d at 704-05 (recognizing that the purpose of a Crowder instruction is to alleviate coercive potential (citing Perkins v. United States, 473 A.2d 841, 846-47 (D.C. 1984)).8

Lastly, we note that following the jury's second verdict, the trial judge specifically observed Juror Three and found that he "didn't seem to be equivocal" when he agreed with the verdict. This court has observed that the trial judge's "on-the-spot perception" of whether a juror was coerced is entitled to some deference. Harris, supra, 622 A.2d at 701 n.6. Additionally, over thirty minutes had passed between the time the court sent the jury back to deliberate and its final verdict, which the trial court felt provided an "adequate opportunity to have talked things through and to give any concerns or [sic] chance to be voiced and addressed. It doesn't suggest that the juror just got rolled over . . . ." See Green, supra, 740 A.2d at 25, 32 (concluding that the jury was not coerced following a forty minute gap between re-deliberations and the guilty verdict). The jury also sent a second note during re-deliberations, which indicated to the trial court that the jury was "talking through a range of issues."

We agree with the trial judge's assessment of the situation and conclude that the second verdict was arrived at "freely" and "fairly," particularly because the trial court used neutral language in instructing the jury to resume its deliberations. The facts in this case are closer to Green and Elliott, where the dissenting jurors were polled early in line and the jury's exact numerical division was unknown, than to Crowder and Harris, where we found a great degree of inherent coercive potential. For that reason, and because the judge's actions defused the coercive potential, we find no reversible error or actual coercion. Harris, supra, 622 A.2d at 707. Accordingly, we affirm.

So ordered.