MEMORANDUM **
In this case, we again address an appeal by Martin and Nanja Rutherford from their convictions under 26 U.S.C. §§ 7203 *590and 7206(1) for tax evasion. We now affirm their convictions.
I.
Following a sixteen-day trial, the Rutherfords were convicted of one count of willfully failing to file an income tax return, see 26 U.S.C. § 7203, and one count of willfully making a false income tax return, see 26 U.S.C. § 7206(1). The district court denied their motion for a new trial, and the Rutherfords appealed. In addition to other claims not relevant to the appeal now before us, the Rutherfords contended that they had been denied their right to a trial by fair and impartial jurors under the Sixth Amendment. They alleged that “the jury was prejudiced because a large number of IRS and government agents sat directly behind the prosecution table throughout the trial and glared at them, intimidating them, and causing some of the jurors to fear that if they acquitted the Rutherfords, the IRS might retaliate against them.” United States v. Rutherford, 371 F.3d 634, 635 (9th Cir.2004).
In a previous decision, we held that the district court had applied the wrong legal standard in evaluating the Rutherfords’ Sixth Amendment claim. Whereas the district court had required the Rutherfords to prove that the IRS intended to intimidate the jury, we held that this was too heavy a burden. See id. at 640, 644. Drawing on recent Ninth Circuit authority, we held that the proper inquiry is whether allegedly improper contact with a juror “ ‘raises a risk of influencing the verdict’ ” or “ ‘ha[s] an adverse effect on the deliberations.’ ” Id. at 644 (quoting Caliendo v. Warden, 365 F.3d 691, 698 (9th Cir.2004), and United States v. Henley, 238 F.3d 1111, 1118 (9th Cir.2001)). We explained that, under Supreme Court precedent, any “ ‘unauthorized intrusions’ ” into the operation of the jury are “ ‘presumptively prejudicial.’ ” Id. at 641 (quoting Remmer v. United States (Remmer II), 350 U.S. 377, 382, 76 S.Ct. 425, 100 L.Ed. 435 (1956), and Remmer v. United States (Remmer I), 347 U.S. 227, 229, 74 S.Ct. 450, 98 L.Ed. 654 (1954)). Once the presumption of prejudice applies, we explained, a conviction is invalid unless the government can demonstrate that the error was harmless beyond a reasonable doubt. Id. (quoting Henley, 238 F.3d at 1118 (quoting United States v. Dutkel, 192 F.3d 893, 897 (9th Cir.1999))). We then remanded the case to the district court to conduct an evidentiary hearing on the issue of jury tampering. Id. at 644.
On remand, the district court conducted the required evidentiary hearing. The hearing involved testimony from all eleven living jurors, an employer of one of the jurors, as well as the Rutherfords’ investigator and his wife. At the conclusion of the hearing, the district court found that the “likely consequence of the staring appears to be that the jurors ... would discuss and consider the IRS, the number of agents present, [and] possible IRS retaliation or audits.” The judge therefore ruled that the presumption of prejudice applied. The district court also ruled, however, that the government had met its burden of proving beyond a reasonable doubt that the agents’ conduct was harmless. The district court noted that all eleven jurors testified “unconditionally, unhesitatingly, and resolutely, and in our view credibly, that the conduct of the IRS agents ... did not affect the jurors’ ability to fairly and impartially receive the evidence, listen to the evidence presented, think about the evidence, pay attention to the judge’s instruction, or [their] ability to deliberate.” Accordingly, it held that the government had proved that any interference with the jury had been harmless and upheld the Rutherfords’ convictions.
The defendants appealed, and we have jurisdiction under 28 U.S.C. § 1291. The Rutherfords advance two arguments on *591appeal: (1) that the district court again applied the wrong legal standard on remand, and (2) that the district court’s holding was erroneous, given the evidence available prior to and presented at the hearing. We address each argument in turn.
II.
The Rutherfords first contend that the district court applied the wrong legal standard on remand. See Appellant’s Opening Brief at 24-25, 35-38. Whether the district court applied the proper standard to evaluate the Rutherfords’ claim of jury tampering is an issue of law that we review de novo. See Rutherford, 371 F.3d at 645.
Our previous opinion established a two-step analysis for the district court. First, we held that the district court had to determine whether the presumption of prejudice applied by evaluating “whether the unauthorized conduct ‘raises a risk of influencing the verdict’ or ‘had an adverse effect on the deliberations.’ ” Id. at 644 (quoting Caliendo v. Warden, 365 F.3d 691, 698 (9th Cir.2004), and United States v. Henley, 238 F.3d 1111, 1118 (9th Cir. 2001)). Second, we held that, if the presumption of prejudice applies, the district court must determine whether the government had met its “heavy burden ... to rebut the presumption by proving that the error was harmless beyond a reasonable doubt.” Id. at 641. In explaining this second step of the inquiry, we cited to our decision in United States v. Henley, in which we described the harmless-error standard as requiring the government to prove that “there is no reasonable possibility that [a juror] was affected in his freedom of action.” Id. (quoting Henley, 238 F.3d at 1118).
We find that the district court correctly followed the prescribed analysis. Finding that the presumption of prejudice applied, the district court nonetheless held that the agents’ staring was “harmless beyond a reasonable doubt” because “there’s no credible basis to infer here in the circumstances that the jurors would have hesitated to act.” The district court further stated that the staring “did not affect” the jurors’ capacity to receive evidence, evaluate evidence, follow instructions, or deliberate. In short, the district court properly applied the harmless-error review set forth by the Supreme Court in Remmer and later explained by the Ninth Circuit in Henley.
The Rutherfords argue that the district court applied the wrong standard because it did not employ the “no reasonable possibility” language used in our decision in Henley. But Henley — and others cases like it, see, e.g., Caliendo, 365 F.3d at 691; United States v. Dutkel, 192 F.3d 893, 899 (9th Cir.1999) — merely rephrase the harmless-error standard set forth by the Supreme Court in the Remmer cases. See Henley, 238 F.3d at 1114-15 (“The Supreme Court announced the standard governing allegations of jury tampering in Remmer v. United States. ”). That is the test we required the district court to apply on remand, and that is the test it actually applied.
The Rutherfords also argue that the district court’s analysis was improper because the judge considered whether the agents’ staring and glaring affected the jury’s verdict. According to the Rutherfords, our previous decision held that the district court should not consider whether the jury tampering affected the outcome of the case. To the contrary, what we held was that a claim of jury tampering does not fail simply because the alleged tampering did not affect the outcome. The district court did not err in considering and concluding that the outcome of the case was not affected by the alleged misconduct of the IRS agents. (In fact, it is necessary, *592though not sufficient, that the district court determine that the alleged tampering did not affect the verdict, for if it had affected the verdict, such tampering could not have been harmless.) While it is true that the district court observed that “the staring ... did not affect the final outcome of the case,” it also noted that this observation was “persuasive, although not determinative.” In short, we conclude that the district court correctly applied the harmless-error standard that we directed it to apply.
III.
The Rutherfords next contend that the district court erred by holding that the government met its burden of proving harmless error. See Appellant’s Opening Brief at 25, 27-28, 37-38. “Because the district court held extensive evidentiary hearings and made findings of fact, we review the findings of fact to determine whether they are ‘clearly erroneous.’ ” United States v. Elias, 269 F.3d 1003, 1020 (9th Cir.2001) (quoting Sea Hawk Seafoods, Inc. v. Alyeska Pipeline Serv. Co., 206 F.3d 900, 911 (9th Cir.2000)).
We are not persuaded that the district court’s factual determinations were clearly erroneous. Seven of the jurors— Kaylor, Washington, Porter, Foster, Guiles, and Campbell — testified that they did not notice IRS agents staring at them, that they had no concern during the trial about IRS retaliation, and that they did not hear any discussion about such retaliation.1 Three other jurors testified that there had been discussion of IRS retaliation, but these jurors viewed that discussion as a “joke” (Juror Clegg), as mere “comments” about the power of “the IRS itself’ (Juror Hoff), or as concern about “the power of the IRS” rather than “concern about being ... stared at” (Juror Hartung). The final juror (Juror Walker), who had been most unnerved by the staring, testified that she initially felt “uncomfortable” but that the “distraction” of staring “didn’t bother [her]” once she grew accustomed to the presence of the agents.
In evaluating the evidence, the district court specifically considered and weighed the testimony that the agents’ staring attracted the jury’s attention. For example, the district court particularly reviewed the testimony of Juror Walker — who was by far the most concerned about retaliation and the most attentive to the IRS agents’ glares — and concluded that she had overcome her initial distraction and was able to concentrate on the evidence as it was presented. Likewise, the district court closely evaluated Juror Hartung’s testimony connecting the agents’ staring to the discussion by some jurors about IRS retaliation but concluded that the jurors agreed that they could nonetheless judge the case fairly and without fear. These are precisely the sort of factual determinations that district courts must make, and the district court found here that the alleged staring and glaring did not affect the capacity of the jurors to receive or evaluate the evidence or to render a verdict based on a fair deliberation. The district court’s findings were not clearly erroneous.
Indeed, the record is more than adequate to support the district court’s finding that the alleged misconduct of the attending IRS agents did not have a reasonable possibility of impeding the jury’s freedom of action. We affirm the district court’s decision that the alleged staring and glaring was harmless beyond a reasonable doubt.
*593For the foregoing reasons, the judgment of the district court is AFFIRMED.
This disposition is not appropriate for publication and may not be cited to or by the courts of this circuit except as provided by 9th Cir. R. 36-3.
. Juror Keifs testimony was contradicted by his former employer, who testified that Keil had in fact reported that "some of the jurors were concerned about being audited.” Keil, however, testified that he did not recall conversations in which jurors said they might be subject to retaliation by the IRS.