United States v. Perrault

FILED United States Court of Appeals PUBLISH Tenth Circuit

UNITED STATES COURT OF APPEALS April 21, 2021

Christopher M. Wolpert FOR THE TENTH CIRCUIT Clerk of Court _________________________________

UNITED STATES OF AMERICA,

Plaintiff - Appellee, v. No. 19-2184

ARTHUR PERRAULT,

Defendant - Appellant. _________________________________

Appeal from the United States District Court for the District of New Mexico (D.C. No. 1:17-CR-02558-MV-1) _________________________________

Aric Elsenheimer, Assistant Federal Public Defender, Albuquerque, New Mexico, for Defendant-Appellant.

Sean J. Sullivan, Assistant United States Attorney (John C. Anderson, United States Attorney, with him on the briefs), Albuquerque, New Mexico, for Plaintiff-Appellee. _________________________________

Before PHILLIPS, SEYMOUR, and CARSON, Circuit Judges. _________________________________

PHILLIPS, Circuit Judge. _________________________________

In the early 1990s, a television reporter in Albuquerque, New Mexico, began investigating reports that a well-known Catholic priest had sexually abused numerous boys during his decades-long tenure at several local parishes. Upon learning of the imminent, breaking-news story, that priest, Arthur Perrault, departed early for a *2 previously planned sabbatical. Foregoing his original plans to stay in the United States, he instead opted for an international destination. After a two-week sojourn in Canada, Perrault made a new life in Morocco—a country, as it so happens, that doesn’t share an extradition treaty with the United States.

Fast-forward twenty-five years. In 2017, a federal grand jury sitting in the District of New Mexico charged Perrault with seven counts of sexual abuse, all stemming from Perrault’s relationship with John Doe 1 in the early 1990s. Doe 1 was about age ten at the time of the charged abuse. The Moroccan government agreed to expel Perrault, the FBI brought him back from Morocco, and Perrault stood trial in the community he had fled over two decades earlier. During the trial, seven other victims testified that Perrault had sexually abused them when they were just boys. The jury convicted Perrault of all seven counts.

To hear Perrault tell it, his trial was more character assassination than solemn search for the truth. On that score, he faults the district court for permitting so many former victims to testify. He also argues that the district court plainly erred in seating the jurors that convicted him, who, according to Perrault, had predetermined his guilt before hearing any evidence. Perrault also disputes certain jury instructions as well as his sentence.

We don’t share Perrault’s view of the proceedings in the trial court. Rather, after reviewing the record, the parties’ briefing, and the relevant law, we are convinced that Perrault received a fundamentally fair trial in compliance with his constitutional rights. Exercising jurisdiction under 28 U.S.C. § 1291, we affirm.

BACKGROUND

I. Events Leading Up to Perrault’s Trial

From the mid-1960s to 1992, Perrault was a Catholic priest at several parishes in Albuquerque, New Mexico. Most of Perrault’s acts underlying the Indictment’s charges occurred while he served as the pastor of St. Bernadette’s Church, though other victims encountered Perrault while he was a priest at Our Lady of Guadalupe Church and Our Lady of the Assumption. As detailed below, numerous victims testified that Perrault sexually abused them over the three decades Perrault was a Catholic priest in Albuquerque.1

In 1992, a local television reporter began investigating allegations that Perrault had sexually abused young boys. Almost immediately after learning of the investigation, Perrault told his congregation that he would be taking a sabbatical in Canada, not in Denver, Colorado, as he had planned. After spending two weeks in Vancouver, Canada, Perrault flew to Morocco, where he lived for the next twenty-five years.

In September 2017, a federal grand jury charged Perrault with six counts of aggravated sexual abuse, see 18 U.S.C. § 2241(c), and one count of abusive sexual contact, see 18 U.S.C. § 2244(a)(5). The charges, summarized in the chart below, all relate to Doe 1, whom Perrault abused in 1991 and 1992.

1 In addition to the eight victims who testified at trial, the government represented that it “is aware of dozens of additional victims” who have made similar allegations against Perrault. R. vol. 1 at 568 n.3.

Count Act Alleged Location Statute Violated

1 Digital Kirtland Air §§ 2241(c), penetration of Force Base 2246(2)(C) Doe 1’s anus

2 Digital Kirtland Air §§ 2241(c), penetration of Force Base 2246(2)(C) Doe 1’s anus, separate and apart from the act charged in Count 1 7 Digital Santa Fe National §§ 2241(c), penetration of Cemetery 2246(2)(C) Doe 1’s anus

3 Contact between Kirtland Air §§ 2241(c), Perrault’s mouth Force Base 2246(2)(B) and Doe 1’s penis 6 Contact between Santa Fe National §§ 2241(c), Perrault’s mouth Cemetery 2246(2)(B) and Doe 1’s penis 4 Touching Doe 1’s Kirtland Air §§ 2241(c), genitalia directly Force Base 2246(2)(D) 5 Touching Doe 1’s Kirtland Air §§ 2244(a)(5), genitalia directly Force Base 2246(3) and through the clothing

In 2018, FBI Agents traveled to Morocco to return Perrault to the United States to stand trial. Sometime earlier, the Moroccan government agreed to expel Perrault, and it held him in custody until the FBI came to retrieve him. While Perrault was detained in Morocco, a state department official visited him to ensure that the Moroccan government had been treating him properly. During this visit, Perrault *5 expressed frustration that the United States was continuing to pursue charges against him related to sexual abuse but admitted to “many transgressions” that the Catholic Church “had dealt with” in the 1980s and 1990s. R. vol. 3 at 1306:6–12.

On the flight from Morocco to the United States, Perrault spoke further with his FBI escort. He admitted that he had sexually abused Doe 8, including performing oral sex on him “and said he may have also used his hands.” Id. at 1437:14–1438:12. He also admitted having “touched” Doe 3 and writing a letter to Doe 3’s parents apologizing for what he had done. Id. at 1439:10–1440:3. The government introduced all these admissions at trial. II. Trial

A. Voir Dire

Though the district court ordered an expanded venire panel of over 110 citizens, the parties still were able to complete jury selection in just one day. In the morning, the district court told the prospective jurors that it would be questioning them about their personal experiences with sexual abuse, and it encouraged them to approach the bench to discuss those matters. One juror chose to discuss her experience in front of the entire venire. She related, “I have been sexually assaulted twice, and I thought I was okay, and clearly I’m not. But . . . I’m going to claim it because I was the victim. So I don’t care who hears it. I’m not embarrassed by it. But to say I’m biased is a little underrated.” R. vol. 4 at 99:1–5. After hearing her story, some members of the venire applauded by clapping. The court excused her for cause. *6 Perrault moved for a mistrial based on the venire members’ applause, but the district court denied the motion.

After lunch, counsel began questioning the jurors. As expected, many of the prospective jurors had already heard or seen news accounts about the case, including Perrault’s extradition from Morocco. Also as expected, many prospective jurors stated they had already formed an opinion about Perrault’s culpability. The court excused those jurors for cause.

When the questioning turned to the general subject of sexual abuse by Catholic priests, several members of the venire expressed disappointment. For instance, Juror 6 found the allegations “disheartening” and “upsetting.” Id. at 128:18–20. Juror 26 identified herself as a Catholic who had stopped going to Mass for reasons including her “disappoint[ment] in the church for all the cover-ups.” Id. at 232:5–6. Juror 38 likewise had heard of “cover-ups” in the church. Id. at 233:11–14. But in response to follow-up questions, each of the three stated that they could remain fair and impartial.

In the end, the parties agreed on twelve jurors and four alternates. Though the government exhausted each of its six peremptory challenges, Perrault used only six of his ten. Perrault’s trial counsel thanked the government’s counsel for how voir dire was handled, and the court complimented both parties on a “great job on voir dire.” Id. at 269:22.

B. Witness Testimony

As discussed below, Rules 413 and 414 of the Federal Rules of Evidence permit the government to introduce evidence that a defendant has committed other sexual assaults or child molestation not charged in the indictment. See infra Discussion Section II.B.1. Accordingly, in addition to Doe 1’s testimony, the government called seven Rule 414 witnesses2 who testified that Perrault had sexually abused them during his time in Albuquerque.3 Because Perrault’s primary challenge on appeal is the district court’s admitting those witnesses’ testimony, we summarize their testimony below.

1. Doe 1

In 1990, at age nine, Doe 1 became an altar server at St. Bernadette’s Church. Perrault was then the priest there. Doe 1 spent considerable time with Perrault while assisting him with Mass. Back then, Doe 1 described Perrault as his “best friend,” R. vol. 3 at 1211:4–5, and Perrault referred to Doe 1 as “[h]is boy,” id. at 16–17. At some point, Perrault gifted Doe 1 a “St. Michael’s medal.” Id. at 1225:14. Because he admired Perrault, Doe 1 hoped to become a priest.

2 Though the government moved to admit the sexual-assault testimony of its witnesses under both Rules 413 and 414, we refer to all of them as Rule 414 witnesses for simplicity. 3 The government moved to allow eight other victims to testify, but only seven testified at trial.

Doe 1 testified that Perrault began sexually abusing him when Doe 1 was age ten. Though Doe 1 failed to provide specific dates, he testified that Perrault had abused him at three geographical locations: the rectory4 at St. Bernadette’s, Kirtland Air Force Base (the “Base”), and Santa Fe National Cemetery. Addressing the abuse on the Base, Doe 1 testified that Perrault abused him both in a Base chapel and inside Perrault’s car at a Base parking lot. Sometimes Perrault would fondle Doe 1’s penis and testicles, both through Doe 1’s pants and skin-to-skin. Perrault also put his mouth directly on Doe 1’s penis on several occasions.5 Other times, Perrault used his fingers to penetrate Doe 1’s anus, sometimes causing bleeding. Including the times Perrault hugged and kissed him, Doe 1 testified that Perrault sexually abused him close to a hundred times. Perrault also told Doe 1 that the abuse “was our secret” and that if he spoke ill of a priest he would go to hell. Id. at 1234:19–25, 1235:10–14.

2. John Does 2–3, 5–96

Doe 9 first met Perrault in the fall of 1966 as a freshman in high school at St. Pius X. Perrault was his teacher. When Doe 9 was around thirteen or fourteen, Perrault began taking him to the movies. While there, Doe 9 testified that Perrault would “put his hands inside of my shirt and he would fondle my breasts, fondle my

4 A “rectory” is housing that a church provides for its clergy. 5 At trial, neither Doe 1 nor the government used the phrase “oral sex.” Doe 1 clarified that he viewed Perrault’s abuse “as a kiss versus [i.e., as opposed to] oral sex.” R. vol. 3 at 1258:3–18. 6 Because all the witnesses who testified at trial were male, we refer to each throughout as simply “Doe,” adopting the same numbering scheme as the parties.

nipples.” Id. at 924:20–21. At some point, Perrault gave Doe 9 gifts, including a tobacco pipe and a brass crucifix. Doe 9 also testified that Perrault regularly gave him inappropriate hugs in which Perrault “would embrace [him] and then [Perrault] would rub his thigh between [Doe 9’s] legs up into [his] crotch.” Id. at 927:15–16. Doe 9 estimated that Perrault hugged him that way anywhere from fifty to a hundred times.

The most jarring sexual encounter happened when Doe 9 was seventeen. Rather than using the front door, visitors to Doe 9’s home often used a side gate to go directly into the backyard. Doe 9’s room was accessible from the backyard by a sliding glass door. One Sunday morning, Perrault used the side gate and sliding door to enter Doe’s 9 room and climb into bed with him. Perrault began fondling Doe 9’s penis until Doe 9 ejaculated. After, Perrault told Doe 9 to keep the episode between the two of them.

Doe 8 became an altar boy at Our Lady of Guadalupe in 1970 at age twelve. He assisted Perrault with Mass, funerals, and weddings, and Perrault eventually made Doe 8 head altar boy. Perrault would take Doe 8 on excursions to the Albuquerque Little Theater and the Base. Perrault also gave Doe 8 a leather necklace with stained glass. The first time Perrault abused Doe 8, Doe 8 was a passenger in Perrault’s car. Perrault parked along the roadside, pulled Doe 8’s pants down, and masturbated him. Perrault told him that it was their “secret” and not to tell anyone. Id. at 896:24–897:1. Over a two-year period, Perrault abused Doe 8 “at least 100” times. Id. at 894:6–9. The abuse included oral sex, anal sex, masturbation, and fondling. Doe 8 also *10 testified about two specific incidents that occurred on the Base—in Perrault’s office and in the parking lot. Doe 8 tried to report the abuse to his mother, but she told him “to never talk about a Catholic priest like that.” Id. at 904:19–20. As noted above, the government admitted into evidence Perrault’s pre-trial statements acknowledging that he had sexually abused Doe 8.

Like Doe 8, Doe 3 met Perrault through his service as an altar boy at Our Lady of Guadalupe. Unlike many of the victims who testified, Perrault abused Doe 3 only once. On December 4, 1971, when Doe 3 was twelve, Perrault asked Doe 3 to accompany him to Perrault’s private room in the rectory of Our Lady of Guadalupe. Once inside the room, Perrault closed and locked the door. After telling Doe 3 to sit on the bed, Perrault began touching Doe 3’s legs and kissing his shoulder, chest, and neck. Perrault also started touching Doe 3’s genitals through the outside of his clothing. Eventually, Perrault unbuttoned Doe 3’s pants and touched Doe 3’s testicles and penis directly on the skin.

Doe 3 managed to escape. He ran from the church to a friend’s house and called his mother. Doe 3 later told his father what had happened, and the two of them returned to the church to confront Perrault. Perrault told Doe 3’s father that “he was sorry, that he didn’t mean to do that.” Id. at 840:1–3. Perrault also wrote Doe 3’s mother a letter apologizing for what he had done, which the government entered into evidence. Additionally, Doe 3’s father had Doe 3 write down what had happened, which Doe 3 read into evidence. The Archbishop had assured Doe 3’s parents that “Father Perrault would be taken out of the church and would never be in the church *11 again around kids[,]” id. at 844:8–12, but in 1992 they were “shocked” to learn that Perrault was a pastor at St. Bernadette’s, id. at 846:17–20.

Does 5 and 6 are brothers. They met Perrault around 1974 through their church, Our Lady of Assumption, where they both served as altar boys. Doe 5 was then about age ten; Doe 6 was eleven. Their parents often invited Perrault over for dinner, and Perrault also took the boys to the Base one time. Does 5 and 6 testified that Perrault regularly abused them at their home. Doe 5 testified that he would sit on Perrault’s lap and that Perrault would fondle his penis through his pajama bottoms. Sometimes Perrault would tuck the boys in bed at night, and he would fondle Doe 5’s penis through his pajamas. Doe 5 estimated that Perrault had sexually touched him at least 50 times. Doe 6 testified that Perrault sexually abused him in his bedroom after Perrault had taken Doe 6’s confession. Perrault would fondle Doe 6’s penis through his underwear, and the last time the abuse occurred, Perrault also forced Doe 6 to fondle Perrault’s penis.

Doe 5 also described the last time Perrault abused him. Perrault came to the house when his parents weren’t home and brought a gift for Does 5 and 6 and their sister (“some kind of a drawing game”). Id. at 973:23–25. Perrault told Doe 5, “Now that I’ve given you a gift, I need you to give me something.” Id. at 973:25–974:1. He led Doe 5 to Doe 5’s bedroom and began fondling Doe 5’s penis through his underwear. Perrault also guided Doe 5’s hand down Perrault’s pants and had Doe 5 fondle Perrault’s penis through his boxers. During this episode, Doe 5 kept repeating, *12 “I don’t want to do this.” Id. at 975:21–22. After that incident, Does 5 and 6 told their parents about what had happened that day, but not about the other abuse.

Unlike the other Does, Doe 7 testified to being abused by two Catholic priests. When he was thirteen, Doe 7 met Perrault while in another priest’s private room in the rectory of Queen of Heaven Church. When Perrault arrived, Doe 7 was sitting on a couch watching TV, unclothed, with only a sheet “kind of on top” of him. Id. at 1026:20–1027:4. Perrault entered the room and pulled away the sheet, exposing Doe 7’s penis. Then Perrault sat down and started masturbating, and Doe 7 began masturbating as well. After that initial encounter, Doe 7 saw Perrault regularly. Doe 7 testified that Perrault sexually abused him about thirty times, including while at the Base. The abuse ranged from fondling to oral and anal sex.

In 1983, Doe 2 became an altar boy at St. Bernadette’s when he was about age twelve. Perrault cultivated a relationship with Doe 2 by taking him to a dinner on the Base with a few other altar boys and by giving him gifts. Doe 2 was fourteen the first time Perrault sexually abused him. Though Doe 2 testified that his “memories are blocked,” id. at 1174:12, he remembers Perrault kissing him on the lips in a bedroom in St. Bernadette’s rectory. Doe 2 also testified that he went to Perrault’s personal condominium two or three times. While there, Perrault would hug Doe 2, and Doe 2 “could feel [Perrault’s] erect penis up against” his thigh. Id. at 1176:22–1177:4. When Doe 2 talked to his mother about what had happened at Perrault’s condo, Doe 2 “was made to feel that by allowing Perrault to do what he did,” he “was doing a service.” Id. at 1177:18–20. *13 III. Verdict and Sentencing

After a six-day trial, the jury found Perrault guilty of all seven counts of sexual abuse. In September 2019, the district court sentenced Perrault to 365 months for the six counts of aggravated sexual abuse and to a concurrent term of 120 months for the abusive sexual contact count. The district court adopted the Presentence Investigation Report’s (“PSR”) recommended guidelines calculation in setting Perrault’s sentence. That calculation included a two-offense-level adjustment for an obstruction-of-justice enhancement, tied to Perrault’s fleeing to Morocco.

This appeal followed, and we have appellate jurisdiction under 28 U.S.C. § 1291.

DISCUSSION

Perrault raises five issues on appeal. First, he argues that he was denied his Sixth Amendment right to be tried by an impartial jury. Second, he argues that the district court erred by allowing testimony from seven Rule 414 witnesses about numerous incidents of uncharged sexual abuse. Third, he argues that the jury instructions lacked specificity, violating the requirement of jury unanimity and the prohibition against double jeopardy. Fourth, he argues that the district court improperly included an obstruction-of-justice enhancement in calculating his sentence. Fifth, he argues that the cumulative effect of multiple errors denied him a fair trial. We address each argument in turn, finding merit in none of them. *14 I. Perrault Was Tried by an Impartial Jury

Perrault argues that he was denied his Sixth Amendment right to an impartial jury. Specifically, he maintains that the venire was tainted by a mix of pretrial publicity and preconceived notions about sexual abuse by Catholic priests—both generally and within New Mexico. But because Perrault didn’t object to the jury panel, he can succeed only by proving the district court plainly erred in seating the jurors who rendered his guilty verdict. United States v. Portillo-Quezada, 469 F.3d 1345, 1349 (10th Cir. 2006). Under that standard, we reverse only if “(1) the district court committed error; (2) the error was plain—that is, it was obvious under current well-settled law; (3) the error affected the Defendant’s substantial rights; and (4) the error seriously affected the fairness, integrity, or public reputation of judicial proceedings.” United States v. Chavez-Morales, 894 F.3d 1206, 1214 (10th Cir. 2018).7 Perrault’s claim fails at step one.

The Sixth Amendment provides that “[i]n all criminal prosecutions, the accused shall enjoy the right to a speedy trial, by an impartial jury.” U.S. Const. amend. VI. “In essence, the right to jury trial guarantees to the criminally accused a fair trial by a panel of impartial, ‘indifferent’ jurors.” Irvin v. Dowd, 366 U.S. 717, 722 (1961). An impartial jury renders a verdict “based upon the evidence developed

7 The government argues that Perrault has waived this jury-partiality argument by failing to object in the trial court. But we conclude that Perrault’s failure to object came about by neglect, rather than an intentional waiver. So we review his claim for plain error. See United States v. Malone, 937 F.3d 1325, 1327 (10th Cir. 2019) (“While forfeiture comes about through neglect, waiver is accomplished by intent.” (citation omitted)).

at the trial”—not based on a previously formed opinion. Id. “[D]efendants bear the burden of establishing juror partiality.” Goss v. Nelson, 439 F.3d 621, 627 (10th Cir. 2006) (citing Irvin, 366 U.S. at 723).

Many of our decisions assessing claims of jury partiality involve defendants who sought (or opposed) transfers of venue. See Gardner v. Galetka, 568 F.3d 862, 887–90 (10th Cir. 2009); House v. Hatch, 527 F.3d 1010, 1023–25 (10th Cir. 2008); Goss, 439 F.3d at 627–28; Hale v. Gibson, 227 F.3d 1298, 1331–34 (10th Cir. 2000); Brecheen v. Reynolds, 41 F.3d 1343, 1350–51 (10th Cir. 1994). But despite now claiming that no impartial jury could have been drawn from the community in which he was convicted, Perrault never sought a venue transfer. See Fed. R. Crim. P. 21(a) (“Upon the defendant’s motion, the court must transfer the proceeding against that defendant to another district if the court is satisfied that so great a prejudice against the defendant exists in the transferring district that the defendant cannot obtain a fair and impartial trial there.”). Nevertheless, we find that line of cases instructive because they address the same question that we must answer here: was the defendant tried by an impartial jury as guaranteed by the Sixth Amendment?

In considering that question, we have distinguished between “presumed” and “actual” prejudice. House, 527 F.3d at 1023. Though Perrault doesn’t use the words “presumed” or “actual” prejudice, his arguments generally fit into those analytical categories. We address his arguments accordingly.

A. Presumed Prejudice

Claims of presumed prejudice are “rarely invoked and only in extreme circumstances.” Hale, 227 F.3d at 1332 (quoting Stafford v. Saffle, 34 F.3d 1557, 1566 (10th Cir. 1994)). This type of prejudice claim usually falls into one of two camps. Goss, 439 F.3d at 628. Sometimes defendants in high-profile cases assert that pretrial publicity was so widespread, so inflammatory, and so egregious that it’s obvious “we cannot expect to find an unbiased jury pool in the community.” Id. From our reading of Perrault’s argument, that’s not what he’s claiming. In the background section of his brief, he now asserts that he appeared “on the front page” four times in the six months leading up to trial. Appellant’s Opening Br. at 3. And he notes the following excerpt from the front page of a local newspaper the day before jury selection: “More than 26 years have elapsed since the Rev. Arthur Perrault vanished from New Mexico, leaving behind police, prosecutors and nearly 40 children he allegedly molested and who have had to bear in secret their shame and sorrow.” Id. But Perrault never refers back to these articles in the argument portion of his brief, nor asserts that this particular reporting biased the jury pool.

Instead, we understand Perrault’s argument as falling into the second category of presumed-prejudice claims, which considers whether voir dire demonstrated that the entire jury pool was tainted by “preconceived opinions about the crime.” Goss, 439 F.3d at 633. Perrault insists that voir dire revealed the jury pool’s collective biases:

The pervasive belief [among prospective jurors] that the Catholic Church routinely sheltered predatory priests, the knowledge that the Servants of Paraclete laundered pedophile priests into New Mexico, the unspoken mantra of “always believe the victim” . . . , the revelation [that] those in the venire who knew Mr. Perrault universally characterized their relationship as “unfortunate”, along with the belief that more allegations equated to guilt, combined so there was no chance despite any claim of impartiality that Mr. Perrault . . . could receive a fair trial.

Appellant’s Opening Br. at 29. According to Perrault, the district court excused for cause over half of the prospective jurors, which he asserts is strong evidence that no impartial jury could be drawn from the venire.

Indeed, a high rate of juror excusal for cause may indicate that no impartial jury can be drawn from a community. See Murphy v. Florida, 421 U.S. 794, 802–03 (1975) (“The length to which the trial court must go in order to select jurors who appear to be impartial is another factor relevant in evaluating those jurors’ assurances of impartiality.”). That’s because “[i]n a community where most veniremen will admit to a disqualifying prejudice, the reliability of the others’ protestations may be drawn into question; for it is then more probable that they are part of a community deeply hostile to the accused, and more likely that they may unwittingly have been influenced by it.” Id. at 803. In other words, if enough of the venire admits partiality, courts have some reason to be skeptical of prospective jurors who claim impartiality. For instance, in Irvin, the Supreme Court vacated a defendant’s conviction where the trial court excused just over 62% of the venire for cause (268 jurors excused for *18 cause/430-person venire). 366 U.S. at 727, 729.8 In contrast, in Murphy the Supreme Court concluded that an excusal rate of just over 25% “by no means suggests a community with sentiment so poisoned against petitioner as to impeach the indifference of jurors who displayed no animus of their own.” 421 U.S. at 803.

Our circuit has consistently rejected claims of jury partiality based on an allegedly high excusal rate. For example, in Goss, the trial court excused for cause 39% of the prospective jurors examined by counsel after they expressed “a predisposed opinion of Goss’s guilt.” 439 F.3d at 634. Though we acknowledged that 39% was “not an insignificant percentage,” we concluded it wasn’t enough to presume that the community as a whole was prejudiced against the defendant. Id. Likewise, in Brecheen we ruled, “the fact that slightly over one-quarter of the venire was excluded for cause does not indicate a pervasive community or courtroom hostility toward [Defendant].” 41 F.3d at 1351 (citing Murphy, 421 U.S. at 803).

Here, the parties dispute the true excusal rate: they disagree about both the numerator—how many jurors the district court excused for cause—and the denominator—how many potential jurors made up the venire. In claiming that the district court excused over half of the venire for cause, Perrault includes jurors excused for miscellaneous personal reasons (age, health, previously arranged travel plans, etc.). But the Supreme Court considers only those excused for having

8 In Irvin, apart from the high rate of excusal for cause, eight of the twelve impaneled jurors had admitted during voir dire that they “thought [the defendant] was guilty.” 366 U.S. at 727.

predetermined the defendant’s guilt when assessing whether the community is generally hostile to the defendant—after all, juror X’s previously planned trip to the Bahamas wouldn’t say much about community animus. See Murphy, 421 U.S. at 796, 803 (ignoring the thirty jurors excused for personal reasons). After accounting for jurors excused for personal reasons, we see forty-one jurors who said they couldn’t be impartial.9

As for the denominator, Perrault claims that the venire consisted of 126 potential jurors; the government asserts there were 149. Perrault bases his figure on the number of seats available in the courtroom. But the government provided a list of jurors excused for cause that shows numbered jurors up to 149. Because we have no reason to believe that the district court skipped numbers when assigning juror numbers, the government’s number appears more accurate.10

As for presumed prejudice, we conclude that the district court didn’t err in seating the jury that convicted Perrault. Using the government’s numbers, the district court excused 41 of 149 jurors for cause, an excusal rate of about 28%. As seen,

9 In his Reply, Perrault argues that we should include two jurors that the district court excused for cause because they had been sexually abused. We disagree. The reason those jurors couldn’t be impartial was because the issue was too personal to them, not because they had already made up their minds about Perrault’s guilt. So those two jurors’ inability to assess Perrault’s case impartially doesn’t bear on whether the community as a whole was hostile to Perrault. 10 The result would remain the same even using Perrault’s number. That would produce an excusal rate of 32.5% (41 jurors excused for cause divided by 126 prospective jurors), still appreciably below the kind of rates that would indicate overwhelming prejudice.

Supreme Court and Tenth Circuit precedents consider this rate insufficient to demonstrate that the community sentiment was “so poisoned against [the defendant] as to impeach the indifference of jurors who displayed no animus of their own.” Murphy, 421 U.S. at 803; see also Goss, 439 F.3d at 634. We thus find no reason to doubt the avowals of prospective jurors who affirmed that they could remain impartial. Perrault has failed to meet his burden to show presumed prejudice.

B. Actual Prejudice

To establish actual prejudice, Perrault “must demonstrate ‘the actual existence of [ ] an opinion in the mind of the juror’” that shows partiality. House, 527 F.3d at 1024 (brackets in original) (quoting Murphy, 421 U.S. at 800). In reviewing this, we examine “the totality of the circumstances” surrounding voir dire to assess whether “the accused receive[d] a fair trial by an impartial jury free from outside influences.” Hale, 227 F.3d at 1333 (first quoting Stafford, 34 F.3d at 1567; and then quoting United States v. Abello-Silva, 948 F.2d 1168, 1177 (10th Cir. 1991)).

Perrault argues that numerous prospective jurors’ statements during voir dire revealed their evident biases against him. Along this line, he asserts that the venire demonstrated “that it would religiously ‘believe the victim.’” Appellant’s Opening Br. at 25. Perrault draws this conclusion based on some prospective jurors’ statements and from the episode in which some venire members clapped for a juror who admitted for the first time out loud that she had been sexually abused. We’re unpersuaded.

Initially, we note that Perrault highlights the actions and statements of several prospective jurors who were never seated on the jury. But we limit our review of juror impartiality to the actual jurors. See Ross v. Oklahoma, 487 U.S. 81, 86 (1988) (“Any claim that the jury was not impartial, therefore, must focus not on [the challenged juror], but on the jurors who ultimately sat.”); United States v. McVeigh, 153 F.3d 1166, 1205 n.33 (10th Cir. 1998) (“We need only consider the voir dire of the twelve jurors who decided McVeigh’s case, because the bias of unseated jurors is irrelevant to whether McVeigh had an impartial jury.”). To the extent Perrault argues that statements by other venire members exemplify general bias or commonly-held views in the community, we have addressed that claim above. Even though some prospective jurors announced that they couldn’t remain impartial, that doesn’t mean the entire venire was poisoned.11

Perrault fails to show that the jurors actually seated harbored actual bias against him.

Impartiality does not mean jurors are totally ignorant of the case. Indeed, it is difficult to imagine how an intelligent venireman could be completely uninformed of significant events in his community. “It is sufficient if the

11 Even if a clapping venire member sat as a juror, we agree with the district court that the clapping was to show emotional support. That doesn’t demonstrate an “always believe the victim” mentality. See R. vol. 4 at 107 (“I think it was they clapped . . . when she said she wanted to own it, that she didn’t want to be ashamed of it . . . . [M]y interpretation of it was . . . they clapped[] when she was finding it difficult to say it in public, that she had been assaulted . . . . [T]he jurors clapped as if to encourage her, the fact that she did not want to be ashamed of it. So it wasn’t that they were applauding the fact that she had been sexually assaulted . . . . It was the fact that they were encouraging her taking a stance at finally not being ashamed of being a victim.”).

juror can lay aside his impression or opinion and render a verdict based on the evidence presented in court.”

Abello-Silva, 948 F.2d at 1178 (quoting Irvin, 366 U.S. at 723). In brief, that jurors are generally aware of allegations of sexual abuse by Catholic priests, support the #metoo movement, or may have heard about the allegations against Perrault doesn’t preclude them from serving on the jury.

Perrault attributes a colorable claim of actual bias to a single juror. He points to this exchange between his trial counsel and Juror 26:

MR. WINDER: Okay . . . . Now, how many individuals have read about the allegations against my client in the news media, Journal, seen it on the news stations, or read it on the Internet? All right. There’s a number of individuals . . . . What type of media have you seen with regard to my client?

JUROR [26]: I believe it was the news, a couple of different times. I don't think it was the newspaper, possibly the Internet. Basically it was the news.

MR. WINDER: Did you form any opinion based upon what you saw on the news?

JUROR [26]: Well, you know, knowing that he was indicted and extradited from another country, that affected me. I don’t think it said -- I don’t recall it saying how many allegations there were. All I remember was the indictment and then the extradition, and then when he finally came to Albuquerque.

MR. WINDER: So you saw that he had been extradited. Are you saying that you had an opinion with regard to whether he was innocent or guilty?

JUROR [26]: I don’t know if I had an opinion. I just -- you know, like I said before, I was saddened at the -- what has happened to the Catholic Church. And, to me, it was like, oh, gosh, here we go again.

MR. WINDER: All right. Thank you. *23 R. vol. 4 at 241:4–242:8. Perrault points to Juror 26 to show that the jurors’ promises to consider his case fairly couldn’t be trusted.

But Juror 26’s statements hardly rise to the level of actual prejudice. To start, though her statement, “oh, gosh, here we go again,” might suggest she harbored general suspicions about Catholic priests’ propensity for sexual abuse, she denied having an opinion about Perrault’s guilt or innocence. See id. at 242:4–7. The Supreme Court has found that more-concerning statements don’t necessarily prove juror partiality, especially when jurors represent that they can remain impartial. See Murphy, 421 U.S. at 801–02 (rejecting petitioner’s claim of jury impartiality despite one juror’s having “conceded that his prior impression of petitioner would dispose him to convict”). Indeed, “we give due deference to jurors’ declarations of impartiality and the trial court’s credibility determination that those declarations are sincere.” McVeigh, 153 F.3d at 1184 (collecting cases). And, though Juror 26 expressed disappointment in the Catholic Church’s alleged cover-ups of sexual abuse, she didn’t waver when Perrault’s trial counsel asked if she would hold that against Perrault: “No. Like I said . . . , I would be able to separate those issues from what is in front of us now.” R. vol. 4 at 232:16–18.

Moreover, we also consider what directions, if any, the district court provided to the venire to remind them of the importance of remaining impartial. See McVeigh, 153 F.3d at 1183 (noting that “the district court went to great lengths to admonish all potential jurors to ignore the publicity surrounding the issues of the case”). Though the district court didn’t instruct the jury specifically about ignoring publicity *24 surrounding the case, it urged the venire to take the process “extremely seriously” and to focus on the “ultimate question . . . . Can I be fair? . . . Can I make a decision in this case based on the evidence that is going to come from the witness stand . . . [?]” R. vol. 4 at 31:22–32:6 (emphasis added).

Perrault’s claims of presumed and actual jury prejudice fail. The district court wasn’t forced to excuse enough jurors for cause to support an inference of widespread community prejudice. Nor has Perrault presented compelling evidence that the jurors who convicted him harbored biases that rendered them unfit to serve. We thus see no error in the district court’s accepting the jurors’ avowals that they could objectively evaluate the charges against Perrault.12

12 Perrault has filed a Rule 28(j) letter directing us to United States v. Tsarnaev, 968 F.3d 24, 37 (1st Cir. 2020), cert. granted, No. 20-443, 2021 WL 1072279 (U.S. Mar. 22, 2021). That case involved the trial and sentence of Dzhokhar Tsarnaev, one the so-called “Boston Bombers,” whose terrorist acts garnered national and international news coverage. See id. at 48, 113 n.96. There, the First Circuit reversed after determining that the district court had failed to adequately question prospective jurors about the kinds of media they had been exposed to related to the case. See id. at 57 (noting that voir dire should elicit “the kind and degree of prospective juror[s’] exposure to the case or the parties” (internal quotation marks omitted)). Perrault argues that we too must reverse, on grounds that the district court failed to independently examine the venire about the details of its exposure to media concerning the case. We disagree. To our knowledge, our precedent has never required district courts to press prospective jurors about what specific media they may have been exposed to about a case before allowing them to sit on the jury—even in high-profile cases. But even if we decided to adopt the First Circuit’s heightened requirement, that wouldn’t affect the result. We’re reviewing the district court’s ruling for plain error. Questioning jurors about the particular media they had seen about Perrault’s case wouldn’t have been obvious to the district court when our precedents don’t require it and when the standard Perrault urges us to adopt was established by the First Circuit over a year after the district court oversaw voir dire.

II. The District Court Properly Allowed the Government’s Rule 414 Witnesses to Testify

A. Standard of Review

We review the district court’s evidentiary rulings for an abuse of discretion. United States v. Sturm, 673 F.3d 1274, 1285 (10th Cir. 2012). Whether “to exclude evidence under Rule 403 is within the sound discretion of the trial court, and will be reversed only upon a showing of a clear abuse of that discretion.” United States v. Guardia, 135 F.3d 1326, 1331 (10th Cir. 1998) (citation omitted). A court abuses its discretion when it “renders a judgment that is arbitrary, capricious, whimsical, or manifestly unreasonable.” United States v. Munoz-Nava, 524 F.3d 1137, 1146 (10th Cir. 2008) (internal quotation marks and citation omitted).

B. The District Court Didn’t Abuse Its Discretion by Allowing the Government’s Seven Rule 414 Witnesses to Testify

1. Legal Framework

Federal Rule of Evidence 413 provides: “In a criminal case in which a defendant is accused of a sexual assault, the court may admit evidence that the defendant committed any other sexual assault. The evidence may be considered on any matter to which it is relevant.” Fed. R. Evid. 413(a). Federal Rule of Evidence 414 mirrors Rule 413, substituting “child molestation” for “sexual assault.” Fed. R. Evid. 414(a).13 These rules constitute exceptions to the general rule of 404(a) prohibiting admission of evidence to show a defendant’s propensity to commit bad

13 “Child” means a person under the age of fourteen. Fed. R. Evid. 414(d)(1).

acts. See Guardia, 135 F.3d at 1331 (“[T]hese rules carve out exceptions to Rule 404(a) and reflect a legislative judgment that certain types of propensity evidence should be admitted.”). Congress enacted these rules in part to address a frequent problem in sexual-assault cases—trials presenting he-said, she-said stalemates, what Congress described as “unresolvable swearing matches.” United States v. Enjady, 134 F.3d 1427, 1431 (10th Cir. 1998) (quoting 140 Cong. Rec. S129901–01, S12990 (R. Dole, Sept. 20, 1994)).

Before the government may admit Rule 413 or Rule 414 evidence, it must overcome “several hurdles.” Guardia, 135 F.3d at 1332. First, the evidence must meet three threshold requirements. A district court must determine (1) “that the defendant is accused of an offense of sexual assault”; (2) that the evidence proffered is “evidence of the defendant’s commission of another offense of sexual assault”; and (3) that the evidence is relevant. Id. at 1328 (ellipsis and citations omitted). Perrault doesn’t challenge the district court’s ruling that the evidence met these threshold requirements, so we don’t discuss them further.

Second, the district court must conduct a Rule 403 balancing test to assess whether the evidence’s “probative value is substantially outweighed by the danger of unfair prejudice.” United States v. Benally, 500 F.3d 1085, 1090 (10th Cir. 2007) (quoting Guardia, 135 F.3d at 1329; Fed. R. Evid. 403); see also Charles Alan Wright & Arthur R. Miller 23 Fed. Prac. & Proc. Evid. § 5387 (2d ed.) (“The starting point of that analysis is the proposition that evidence offered under Rule 413 [or 414] is not unfairly prejudicial simply because it is offered to show a propensity to commit *27 sexual assault. While the evidence is prejudicial when offered for that purpose, that prejudice is not unfair since the rule makes the evidence admissible for that very purpose.”).

The balancing test proceeds in two phases. First, the district court must consider what have become known as the four “Enjady” factors:

(1) how clearly the prior act has been proved;14

(2) how probative the evidence is of the material fact it is admitted to prove;

(3) how seriously disputed the material fact is;15 and

(4) whether the government can avail itself of any less prejudicial evidence. Enjady, 134 F.3d at 1433 (citation omitted). In considering the above factors, “no single factor is dispositive.” United States v. Mann, 193 F.3d 1172, 1175 (10th Cir. 1999).

Second, district courts must weigh the Enjady factors against “the probative dangers,” by assessing three additional factors:

(1) how likely is it such evidence will contribute to an improperly-based jury verdict;

(2) the extent to which such evidence will distract the jury from the central issues of the trial; and

14 Under this factor, the district court must conclude by a preponderance of evidence that a jury could reasonably find that the “other act” occurred. Enjady, 134 F.3d at 1433 (citation omitted). 15 “The third factor requires the district court to ‘evaluate the seriousness of the dispute over the material fact the Rule 414 evidence is admitted to prove.’” Sturm, 673 F.3d at 1286 (emphasis added) (quoting Mann, 193 F.3d at 1174). Perhaps counterintuitively, “[t]he more seriously disputed the material fact, the more heavily this factor weighs in favor of admissibility.” Id. (citation omitted).

(3) how time consuming it will be to prove the prior conduct. Enjady, 134 F.3d at 1433 (citation omitted). In conducting this balancing, our precedents caution that “the exclusion of relevant evidence under Rule 403 should be used infrequently, reflecting Congress’ legislative judgment that the [Rule 413 and 414] evidence normally should be admitted.” Id. (internal quotation marks and citation omitted); see also Benally, 500 F.3d at 1090 (“Consistent with congressional intent regarding the admission of evidence tending to show the defendant’s propensity to commit sexual assault or child molestation, ‘courts are to “liberally” admit evidence of prior uncharged sex offenses.’” (quoting United States v. Meacham, 115 F.3d 1488, 1492 (10th Cir. 1997))).

2. Perrault’s Arguments on Appeal

In two thorough orders, the district court carefully weighed the four Enjady factors against the probative dangers, concluding that the Rule 414 witnesses’ testimony was both relevant and probative. Perrault doesn’t dispute the district court’s conclusions regarding three of the four Enjady factors: he concedes that the government proved by a preponderance of the evidence that he had committed the other sexual abuse; that the material fact (Perrault’s abuse of Doe 1) was seriously disputed; and that the government couldn’t avail itself of any less-prejudicial evidence. Instead, Perrault raises two challenges to the district court’s Rule 403 ruling related to the evidence’s probativeness (the second Enjady factor) and two challenges related to unfair prejudice. We address each in turn.

a. Probative Value of the Evidence

When considering the probative value of evidence, we have enumerated five considerations to guide our analysis:

(1) the similarity of the prior acts and the charged acts;

(2) the time lapse between the other acts and the charged acts;

(3) the frequency of the prior acts;

(4) the occurrence of intervening events; and

(5) the need for evidence beyond the defendant’s and alleged victim’s testimony.

Benally, 500 F.3d at 1090–91 (citing Guardia, 135 F.3d at 1331). The district court concluded that each of these factors favored admitting the other victims’ testimony. We agree.

Perrault disputes the probative value of the uncharged-acts evidence on two grounds, neither of which is persuasive. First, Perrault argues that evidence of his giving alleged victims gifts and taking them on excursions is probative “only . . . with a prior belief in Mr. Perrault’s guilt—otherwise the significance of such a detail as a priest giving a parishioner a religious medallion is indeterminate.” Appellant’s Opening Br. at 35. True, a priest can give a congregant something without raising any suspicion of ulterior motive. But Perrault ignores that gift-giving is a common grooming technique. See R. vol. 3 at 1374 (expert testimony introduced *30 at trial describing ways perpetrators groom victims).16 And we permit courts to admit prior-act evidence to show a defendant’s “pattern of grooming.” United States v. Batton, 602 F.3d 1191, 1198 (10th Cir. 2010).

Second, Perrault argues that the district court should have excluded the testimony of Does 2, 7, and 9 as being too dissimilar from Doe 1. Because Doe 1 “was prepubescent” (age 10–11) when Perrault allegedly assaulted him, Perrault maintains that Does 2, 7, and 9 weren’t similar to Doe 1 because they were all adolescents when Perrault abused them.17 Appellant’s Opening Br. at 35. In support, Perrault cites the Sentencing Guidelines, which recommend harsher sentences for abuse of minors under age twelve compared to abuse of minors between twelve and sixteen. Id. (citing U.S.S.G. § 2G2.1(b)(1)). He also cites the American Psychiatric Association’s 2013 Diagnostic and Statistical Manual of Mental Disorders (5th ed.),

16 “[T]he term ‘grooming’ is often used in the child sexual abuse research field to look at how some perpetrators will gradually involve a child in sexual abuse. So it might start with favoring the child, praising the child, patting the child, hugging, and gradually more and more gets into more sexual acts. At the same time, you see the offenders then saying, you know, ‘Don’t tell. Keep it a secret.’ And that grooming process, you’re building up a relationship with the child, sometimes with the parents as well. And at the same time, you’re gradually increasing the sexual part of the activity.” R. vol. 3 at 1374 (expert testimony introduced at trial). 17 Doe 2 testified that he was fourteen when Perrault began abusing him. Doe 7 was thirteen when Perrault first abused him, but the record suggests that the abuse continued into Doe 7’s later teen years. See R. vol. 3 at 1035:19–20 (Judge Vazquez cautioned the government, “I don’t want you to get into anything more past his age of 16.”). Doe 9 was between the ages of thirteen and seventeen when Perrault abused him.

which defines pedophilia as attraction to prepubescent children (typically under age thirteen).

But Perrault fails to explain why these sources bear on the probativeness of the evidence. The relevant question here is whether Perrault’s prior acts of abusing young men ages thirteen to seventeen are similar to his accused conduct of abusing Doe 1. They are. That Does 2, 7, and 9 were a few years older than Doe 1 doesn’t mean the prior abuse isn’t probative of Perrault’s propensity to abuse young boys. Nor does our precedent establish such a rigid rule. In Benally, the defendant was charged with sexually abusing his twelve-year-old granddaughter. 500 F.3d at 1086. At trial, the district court allowed the government to introduce four Rule 414 witnesses, each of whom testified that the defendant had sexually assaulted them. Id. at 1088. Those witnesses’ ages ranged from ten to twenty years old when the defendant sexually assaulted them. Id. Yet even though the defendant had assaulted one of the witnesses when she was an adult, we nonetheless upheld the district court’s ruling that the Rule 414 witness testimony was probative in part because “each [assault] involved a young woman whom [Defendant] previously knew.” Id. at 1092. Though the other victims were not the same ages as the defendant’s granddaughter, we still concluded that the similar ages and circumstances supported admitting the evidence. See id. at 1091. This holds with even greater force here when all the Rule 414 witnesses were adolescents.

In sum, the substantial, detailed Rule 414 evidence was highly probative of the material fact that it was admitted to prove—that Perrault had a propensity to sexually abuse boys.

b. Prejudice

Perrault argues that the prejudice of the Rule 414 witnesses’ testimony substantially outweighed its probative value. Perrault maintains that this testimony (1) was cumulative and confused the jury, and (2) resulted in an improperly-based jury verdict. We find neither argument persuasive.

“Evidence is cumulative if its probative effect is already achieved by other evidence in the record; that is, ‘if the small increment of probability it adds may not warrant the time spent in introducing it.’” Jewell v. Life Ins. Co. of N. Am., 508 F.3d 1303, 1314 (10th Cir. 2007) (quoting United States v. Davis, 40 F.3d 1069, 1076 (10th Cir. 1994)). This stands in contrast to corroborative evidence, “which buttresses weak or assailable evidence, often by establishing data which refute possible discrediting circumstances.” Id. (internal quotation marks and citation omitted).

We have never directly assessed a challenge like the one Perrault asserts here—that the sheer number of Rule 414 witnesses who testified unduly prejudiced the jury. Most of our decisions addressing Rule 414 witnesses have concerned challenges involving three witnesses or fewer. See, e.g., Batton, 602 F.3d at 1195, 1202 (finding no error in district court’s admitting evidence of the defendant’s prior conviction for sexual abuse); United States v. McHorse, 179 F.3d 889, 894 (10th Cir. 1999) (upholding district court that allowed testimony by two other alleged victims); *33 Meacham, 115 F.3d at 1491 (two other victims); United States v. Mercer, 653 F. App’x 622, 624–25 (10th Cir. 2016) (unpublished) (three other victims). But we have affirmed district-court rulings that have allowed more Rule 414 witnesses. For instance, in Benally, we found no error in the district court’s admitting four Rule 414 witnesses, 500 F.3d at 1088–89, nor did we question the admission of six such witnesses in United States v. Magnan, 756 F. App’x 807, 820 (10th Cir. 2018) (unpublished).18

At least one of our sister circuits has discouraged district courts from admitting more than five Rule 414 witnesses: “We are troubled by the district court’s admission of six Rule 413 and 414 witnesses . . . . [I]n this case, six such witnesses injects cumulative evidence with little additional probative value.” United States v. Never Misses A Shot, 781 F.3d 1017, 1028 (8th Cir. 2015). Though the court clarified that it “state[d] no inflexible rule that provides a maximum limit of Rule 413 or 414 witnesses that can testify,” id. at 1028 n.6, it reasoned that, “at some point . . . enough is enough,” id. at 1028 (citation omitted).

Like the Eighth Circuit, we decline to set a maximum number of Rule 414 witnesses that district courts may allow. Instead, when a defendant challenges Rule 414 testimony as cumulative, the district court should consider the diminishing

18 But Magnan differs from our case. There, the allegations of five of the six women provided the basis for the twelve counts of sexual abuse brought against the defendant. Magnan, 756 F. App’x at 811. That is, only one of the women testified solely to show the defendant’s propensity to sexually abuse young women. Here, all seven of the Rule 414 witnesses testified solely to demonstrate Perrault’s propensity to sexually abuse young men.

marginal return on each additional witness’s testimony. Cf. United States v. Cunningham, 694 F.3d 372, 389–90 (3d. Cir. 2012) (reversing the district court for permitting the government to show numerous video clips of child pornography because, “[e]ven though the . . . videos were probative, . . . the law of diminishing marginal returns still operates”). Particularly in cases when the government seeks to introduce a large number of Rule 414 witnesses at trial (more than four), this may require district courts to consider the proffered testimony seriatim to determine whether each additional witness’s testimony provides enough probative value in view of the testimony already permitted.

This isn’t to say that each Rule 414 witness’s testimony must be uniquely different from other Rule 414 witnesses’ testimony. In Benally, for example, four Rule 413 and 414 witnesses testified that the defendant had raped them. 500 F.3d at 1088. Though the specific facts of each witness’s experience varied, each testified for one purpose: to show the defendant’s propensity to commit sexual assault. See id. at 1088–89. That didn’t render the testimony cumulative. Juries may find one Rule 414 witness’s testimony credible, while dismissing the next. For that reason, the government may need multiple Rule 414 witnesses to effectively bolster the credibility of the victim tied to the indictment, particularly where the government’s case would otherwise rely on that victim’s testimony alone. And we have repeatedly affirmed that, consistent with Congress’s intent in enacting Rules 413 and 414, “courts are to liberally admit evidence of prior uncharged sex offenses.” Id. at 1090 (internal quotation marks and citation omitted).

But, like the Eighth Circuit, we recognize that at some point enough is enough, and the probative value of the fifth, sixth, or seventh Rule 414 witness is probably less than the first three or four. See Never Misses A Shot, 781 F.3d at 1028. Thus, in those rarer cases involving numerous Rule 414 witnesses, district courts may assure themselves that each witness’s testimony remains corroborative by identifying specifically how it “buttresses weak or assailable evidence.” Jewell, 508 F.3d at 1314.

Here, the district court allowed testimony from seven Rule 414 witnesses, many of whom testified about dozens of instances of Perrault’s sexual abuse. Though the number of witnesses and allegations gives us pause, we conclude that the district court acted within its discretion by admitting their testimony. Contrary to Perrault’s contention, none of the seven witnesses merely parroted identical accusations. Each witness’s testimony “buttresse[d] weak or assailable evidence.” Id. Because the government lacked physical evidence supporting Doe 1’s claims of sexual assault all those years ago, the testimony of other alleged victims bolstered Doe 1’s credibility, thus reducing the likelihood that the trial would reduce to an “unresolvable swearing match[].” Enjady, 134 F.3d at 1431 (citation omitted). For example, the testimony of Does 3 and 8 was especially probative and important to the government’s case because the government introduced Perrault’s own admissions that he had abused those witnesses. Doe 7’s testimony differed from most of the others because, apart from Doe 8, he was the only other witness to testify that Perrault had abused him at the Base—important to corroborate Doe 1’s allegations that he too had been abused *36 there. The testimony of Does 5, 6, and 9 collectively helped establish Perrault’s pattern of gift-giving and grooming methods. Finally, Doe 2 served a specific purpose in refuting Perrault’s defense. At trial, Perrault attempted to discredit Doe 1 by highlighting the $300,000 civil settlement that Doe 1 had obtained from the Catholic Church, presumably to suggest Doe 1 had a financial incentive to fabricate his allegations. But Doe 2 testified that, despite being sexually abused by Perrault, he had never pursued a civil action against Perrault or the Catholic Church. Thus, Doe 2’s testimony served to rebut Perrault’s suggestion that many supposed victims had come forward to accuse him solely in search of a payout from the Catholic Church. In short, because we conclude that each of the Rule 414 witnesses retained independent probative value, we reject Perrault’s argument that their testimony amounted to “needless[] . . . cumulative evidence.” Fed. R. Evid. 403.

In a similar vein, Perrault insists that the Rule 414 witness testimony unfairly prejudiced him because it led the jury to convict him based on the emotional impact of those witnesses’ testimony, rather than because he had sexually abused Doe 1. In effect, Perrault argues that the government convinced the jury that he was a monster and that it convicted him for that reason. But “Rule 414 evidence will almost always have a profound impact on the jury and cause it to feel disgust toward the defendant.” Mercer, 653 F. App’x at 629. Against the advice of the Federal Judicial Conference, Congress adopted Rules 413 and 414 anyway. See id. at 630 n.4. So although Rule 414 witnesses undoubtedly produce a powerful emotional impact on jurors, that alone isn’t sufficient to justify excluding their testimony. Indeed, if we accepted Perrault’s *37 argument, courts could rarely, if ever, admit Rule 414 witnesses for fear of the emotional impact their testimony would have on the jury.

Moreover, we have recognized that district courts can mitigate any potential jury bias from Rule 414 witnesses by giving limiting instructions. Benally, 500 F.3d at 1093 (“[A]ny jury bias that may have resulted from the district court’s decision to admit [the Rule 413 or 414] evidence would have been mitigated by the court’s two separate instructions to the jury regarding the jury’s permissible and impermissible uses of the evidence.”). At least five times during the trial, including as part of the final instructions, the district court gave the jury a limiting instruction along these lines:

Ladies and Gentlemen, you are instructed that the evidence of conduct by the Defendant on a previous occasion with other young boys has been offered by the Government for its bearing on any matter to which it is relevant, including the Defendant’s disposition or propensity to commit the offense that is charged in the Indictment and the improbability that the Defendant has been falsely or mistakenly accused of these crimes. It is entirely up to the jury to determine what weight, if any, such other conduct evidence deserves . . . . However, you are cautioned that the Defendant is not on trial here for any acts or crimes not alleged in the Indictment. The Defendant may not be convicted of the crimes charged in the Indictment if you are to find only that he committed other crimes at such other time.

R. vol. 3 at 1022 (emphasis added); see also id. at 915, 954–55, 1166–67, 1766–67. We’re confident that after hearing this instruction five times, the jury knew it couldn’t convict Perrault of abusing Doe 1 solely on account of a belief that he had abused other young boys at other times.

c. Balancing the Probative Value Against Possible Prejudice

Despite Perrault’s argument on appeal, we conclude that the district court properly weighed the Enjady factors against the probative dangers. At bottom, the Rule 414 witness testimony here was both highly probative and highly prejudicial. But to exclude this kind of relevant evidence under Rule 403, the prejudice must substantially outweigh the probative value. Because the district court acted within its discretion in weighing the value and danger of the evidence, we affirm the admission of this evidence.

III. The District Court Didn’t Plainly Err in Adopting the Challenged Jury Instructions

A. Background

Before trial, the parties jointly submitted proposed jury instructions. Besides one minor dispute not relevant here, the parties agreed on the language for each instruction. They also agreed on consolidating into a single jury instruction multiple counts charging the same kind of sexual abuse. So, for example, counts 1, 2, and 7 charging Perrault with digitally penetrating Doe 1’s anus on three separate occasions were grouped into a single instruction (Instruction 16):

The defendant, Arthur Perrault, is charged in counts 1, 2, and 7 of the indictment with Aggravated Sexual Abuse . . . . To find Mr. Perrault guilty of the crime of Aggravated Sexual Abuse as charged in counts 1, 2, and 7 of the indictment, you must be convinced that the government has proved each of the following beyond a reasonable doubt:

First: The Defendant knowingly engaged in a sexual act with John Doe;

Second: The sexual act was committed in the special maritime and territorial jurisdiction of the United States; and

Third: At the time of the sexual act, John Doe had not attained the age of 12 years.

For purposes of counts 1, 2, and 7 of the indictment, “sexual act” is defined as the penetration, however slight, of the anal or genital opening of another by a hand or finger or by any object, with an intent to abuse, humiliate, harass, degrade, or arouse or gratify the sexual desire of any person.

***

You are instructed, as a matter of law, that Kirtland Air Force Base and the Santa Fe National Cemetery are within the special maritime and territorial jurisdiction of the United States. Accordingly, if you find beyond a reasonable doubt that such offense occurred in the location described in the indictment in Counts 1, 2, and 7, then you are instructed that the alleged aggravated sexual abuse occurred within the special maritime and territorial jurisdiction of the United States.

R. vol. 1 at 508 (emphasis added). Similarly, counts 3 and 6 charging Perrault with “contact between the mouth and the penis” were grouped into Instruction 17. Id. at 509.

The parties also jointly proposed that the district court incorporate the Indictment into Instruction 1. In reading this instruction to the jury at the start of the trial, the court read aloud the Indictment. But the parties’ proposed Instruction 1 didn’t make it into the final jury instructions for the jury to have during deliberations. At the jury-instruction conference held at the end of the fourth day of trial, the district court addressed the already given Instruction 1, stating: “So we start with Jury Instruction No. 1 . . . , and it [includes] the Indictment. That’s a preliminary *40 instruction, so that one’s already been given.” R. vol. 3 at 1516:25–1517:2. The court then turned to Jury Instruction 2, and no one addressed Instruction 1 further.

The final jury instructions that the court distributed to the jurors skipped the parties’ proposed Instruction 1, instead beginning with their proposed Instruction 2. We’re unsure whether the parties intended for the district court to remove that instruction from the jury packet. But Perrault didn’t object before, during, or after the court read the jury its instructions.

On appeal, Perrault argues that the jury needed the Indictment to understand the instructions referring to multiple counts of associated conduct. Absent giving the jury a physical copy of the Indictment, he says, the instructions were vague, causing juror unanimity and double-jeopardy problems. To illustrate this argument, we turn to Jury Instruction 16. Because the court followed the parties’ lead and included counts 1, 2, and 7 in this instruction, Perrault now maintains that the jury may have believed that it could convict him of all three counts based on a single incident of sexual abuse. He phrases it this way: “The instructions did not inform the jury they all had to agree that three distinct acts in three different locations occurred for three convictions.” Appellant’s Opening Br. at 46. The Indictment, Perrault says, would have cleared up any confusion because it “clearly separated out each count with each act in each place needed to secure a conviction.” Reply Br. at 17.

The government urges us not to reach the merits of Perrault’s argument, asserting that the invited-error doctrine applies. Alternatively, the government argues that Perrault’s argument can’t withstand plain-error review. Though we conclude that *41 the invited-error doctrine doesn’t apply, we agree that Perrault has failed to show that the district court plainly erred in adopting the challenged jury instructions.

B. The Invited-Error Doctrine Doesn’t Apply

The government argues that the invited-error doctrine bars Perrault’s claim because Perrault and the government jointly proposed the now-challenged jury instructions. According to the government, this issue is open-and-shut—after all, we regularly apply the doctrine when defendants challenge jury instructions on appeal that they proffered in the trial court. See, e.g., United States v. Jereb, 882 F.3d 1325, 1335–41 (10th Cir. 2018); Sturm, 673 F.3d at 1280–81. But this case isn’t so straightforward.

As we have previously recognized, the invited-error doctrine is rooted in reliance interests: it “prevents a party who induces an erroneous ruling from being able to have it set aside on appeal.” Jereb, 882 F.3d at 1338 (quoting United States v. Morrison, 771 F.3d 687, 694 (10th Cir. 2014)). “[T]his Court will not engage in appellate review when a defendant has waived his right to challenge a jury instruction by affirmatively approving it at trial.” Id. at 1335 (quoting United States v. Cornelius, 696 F.3d 1307, 1319 (10th Cir. 2012)). So the question here is whether Perrault affirmatively approved the final jury instructions even though the district court removed the parties’ proposed Instruction 1. We conclude that he didn’t.

In Jereb, we considered a similar argument by the government that the invited-error doctrine precluded the defendant’s appeal based on erroneous jury instructions. Id. at 1335. There, the defendant had been charged with violating 18 U.S.C. § 111, *42 which subjects to liability anyone who “forcibly assaults, resists, opposes, impedes, intimidates, or interferes with” various federal officials. Id. at 1333. The defendant proffered a jury instruction to clarify that, though the government needn’t prove all the different acts, the jury must unanimously agree that he committed at least one of those acts. Id. at 1336. The jury did just that, unanimously finding that the defendant “forcibly opposed” a federal officer. Id. at 1334. But the jury didn’t find that the defendant had assaulted the officer. Id. On appeal, the defendant argued that was a problem because this Circuit’s precedents establish that “assault is an element of every conviction under 18 U.S.C. § 111(a)(1).” Id. at 1335 (quoting United States v. Wolfname, 835 F.3d 1214, 1216 (10th Cir. 2016)). Because the jury didn’t find an assault, the defendant maintained that his conviction couldn’t stand. Id. at 1334.

We concluded that the invited-error doctrine barred the defendant’s challenge to the jury instructions because his argument on appeal “directly contradict[ed]” what he argued in the trial court. Id. at 1341. The defendant urged the district court to adopt a jury instruction that allowed the jury to convict even if it didn’t find an assault; on appeal, he argued that the jury’s verdict had to be reversed because it didn’t find an assault. Id. In ruling that the defendant had induced the complained-of error, we emphasized the defendant’s active involvement in crafting the jury instruction he challenged on appeal. See id. at 1339 (“The record further reflects that [the defendant] meaningfully participated in crafting the jury instruction actually given at trial, which reflected the language [the defendant] sought.”); id. at 1340 (noting that the defendant “requested (and received) jury instructions construing a *43 statute that contradict the construction he now prefers on appeal”). In contrast, we reaffirmed that mere passivity on the defendant’s part doesn’t trigger the doctrine:

A defendant’s failure to object to a district court’s proposed jury instruction, or even the affirmative statement, “No, Your Honor,” in response to the court’s query “Any objection?”, is not the same as a defendant who proffers his or her own instruction, persuades the court to adopt it, and then later seeks to attack the sufficiency of that instruction.

Id. at 1341 (quoting United States v. Harris, 695 F.3d 1125, 1130 n.4 (10th Cir. 2012)).

If Perrault had not proffered a jury instruction reciting the Indictment, Jereb would squarely foreclose his challenge—he couldn’t dismiss the Indictment’s importance below only to challenge its exclusion on appeal. That’s not what happened here. It’s true that Perrault jointly proposed the instructions that he now challenges on appeal. But the district court didn’t include in the jury packet all the instructions that the parties proposed; it omitted Instruction 1, which incorporated the Indictment. That omission, Perrault argues, rendered the otherwise-acceptable jury instructions “a morass.” Reply Br. at 18. Perrault maintains that the jury couldn’t properly interpret the instructions without the Indictment too. See id. at 17 (“Had the indictment gone back with the jury instructions, there could be no claim of confusion.”). In other words, Perrault contemplated jury instructions in which the Indictment was part and parcel; once the district court removed that piece, he no longer found them adequate.

And, unlike the defendant in Jereb, Perrault never affirmatively approved the instructions without the Indictment. The closest he came to doing so was at the jury-*44 instruction conference when the district court arguably suggested that it wasn’t going to include the parties’ proposed Instruction 1 in the jury’s instruction packet. But the district court didn’t say it was going to remove proposed Instruction 1 from the jury’s instruction packet. And even had the court done so, Perrault’s failing to object wouldn’t automatically trigger the invited-error doctrine. See Jereb, 882 F.3d at 1341. Perrault’s proffered jury instructions included the full text of the Indictment, and he never induced the court to remove it. The court did that on its own.

Still, if Perrault believed, as he argues now, that the final jury instructions were deficient without the parties’ proposed Instruction 1, he should have objected. He didn’t. Nor did he object to the district court’s failing to send the Indictment back with the jury. So we review his challenge to the jury instructions only for plain error. Id. at 1335.

C. The District Court Didn’t Plainly Err in Adopting the Challenged Jury Instructions

Perrault argues that vagueness in the jury instructions violated two of his constitutional rights—his Sixth Amendment right to a unanimous jury verdict and his Fifth Amendment right against double jeopardy. The Sixth Amendment guarantees that “[a] jury must reach a unanimous verdict in order to convict.” Ramos v. Louisiana, 140 S.Ct. 1390, 1395 (2020); id. at 1396 (“This Court has, repeatedly and over many years, recognized that the Sixth Amendment requires unanimity.”); see also United States v. McElhiney, 275 F.3d 928, 935 (10th Cir. 2001) (“The Sixth Amendment guarantees a federal criminal defendant the right to a unanimous *45 verdict.” (citation omitted)). This generally requires a jury “to be unanimous as to the elements of [each] offense” charged in the indictment. United States v. Allen, 603 F.3d 1202, 1214 (10th Cir. 2010). And the Double Jeopardy Clause of the Fifth Amendment “protects a defendant from being punished multiple times for the same offense.” United States v. Worku, 800 F.3d 1195, 1198 (10th Cir. 2015) (citation omitted).

As noted, Perrault challenges jury Instructions 16 and 17 for addressing multiple counts in a single instruction without clarifying that the jury had to unanimously agree on what separate acts supported a conviction on each count.19 Based on Instruction 16, Perrault asserts that the jury may have convicted him on counts 1, 2, and 7 even if it concluded that he had penetrated Doe 1’s anus only once. And because Instruction 17 similarly groups counts 3 and 6 together, Perrault argues the jury may have convicted him of both counts even if it found only a single incident with Doe 1 of “contact between the mouth and penis.”

We conclude that Perrault has failed to establish that the district court plainly erred and, even if it had, Perrault has failed to show that the error affected his substantial rights. See Chavez-Morales, 894 F.3d at 1214. Perrault derives his argument from our unpublished decision, United States v. McGill, 359 F. App’x 56 (10th Cir. 2010). There, we considered the defendant’s appeal challenging a jury

19 Perrault advances a similar claim about counts 4 and 5, reflected in Instructions 18 and 19, respectively. But his challenges concerning jury unanimity and double jeopardy fail for the same reasons explained below.

conviction of six counts of aggravated sexual abuse. Id. at 57–58. We cautioned that “[t]he use of generic, facially indistinguishable counts with broad overlapping time frames could potentially raise . . . double-jeopardy and jury-unanimity concerns.” Id. at 60. “As for double jeopardy, the jury could be misled into convicting the defendant on more than one count for the same conduct; as for jury-unanimity, different jurors might vote to convict on the same count on the basis of different conduct.” Id. But we concluded that the jury instructions there “obviated” those concerns because they “linked specific counts with particular incidents identified by unique factual circumstances.” Id. We also noted that the government in its closing argument helped dispel possible confusion “by highlighting the specific incidents linked to each count.” Id. at 61 (citation omitted).

Perrault also relies on another decision in which the defendant raised a challenge like his, United States v. Davis, 55 F.3d 517 (10th Cir. 1995). In Davis, police found two firearms in the defendant’s car after conducting a lawful search. Id. at 519. The indictment charged the defendant with “using or carrying a .25 caliber firearm in connection with possession of powder cocaine (Count 4)” and “using or carrying a .22 caliber firearm in connection with possession of crack cocaine (Count 5).” Id. But the district court combined the counts into a single jury instruction. Id. at 520. Relevant here, that instruction required the jury to find that “the defendant used the specific firearm alleged in the Indictment.” Id. We rejected the defendant’s argument that the jury instruction “afforded the jury the opportunity to convict [the *47 defendant] of two § 924(c) firearm counts based on one underlying offense.” Id. The key was the indictment:

Reading Jury Instruction No. 23 in light of the Indictment, we are able to determine that the jury linked each alleged § 924(c) conviction to a separate underlying offense, because each count of the Indictment clearly pairs one of the firearms to one of the drug charges. We find it significant, therefore, that the Indictment was included in the jury instructions . . . .

Id. (emphasis added) (citations omitted).

These decisions clarify that, when reviewing challenges to allegedly vague jury instructions, the inquiry turns on “if the reviewing court can determine whether the jury linked each alleged [criminal] violation to a separate underlying offense.” Id. (internal quotation marks and citation omitted). Together, Davis and McGill suggest several ways courts can verify that the jury did so: whether the instructions included unique identifiers that clearly relate to only one count (for example, by specifying the location where a particular offense occurred); whether the instructions incorporated the indictment; and whether the government clarified during the trial which specific conduct related to each count. See id.; McGill, 359 F. App’x at 60–61. And we must be mindful that “[w]hen reviewing claims of error in regard to jury instructions, we review the instructions as a whole . . . .” United States v. Allen, 603 F.3d 1202, 1213 (10th Cir. 2010). Thus, other instructions can alert the jury that they must link each count to separate conduct.

Considering the record and instructions as a whole, including the government’s clarifying remarks during its closing argument, we are satisfied that the jury linked each count to a separate underlying offense. First, though the district *48 court didn’t include the Indictment in the jury’s instruction packet, it did read the Indictment aloud to the jurors at the start of jury selection and again at the start of the trial. And, as Perrault concedes, the Indictment meaningfully separated the counts. For example, the Indictment informed the jury that counts 1 and 2 charged Perrault with twice sexually abusing Doe 1 at the Base, and count 7 charged Perrault with sexually abusing Doe 1 at Santa Fe National Cemetery. The jurors twice heard the Indictment’s language clarifying that each count referred to separate acts of abuse at different locations.

Second, Instruction 14 further clarified that each count related to separate conduct:

A separate crime is charged against Mr. Perrault in each count of the Indictment. You must separately consider the evidence against Mr. Perrault as to each count and return a separate verdict for each count. Your verdict as to any one count . . . should not influence your verdict as to any other counts.

R. vol. 1 at 506 (emphasis added).

Third, though Instructions 16 and 17 referenced multiple counts in single instructions, the verdict form separated each of the counts. This would have again suggested to the jury that it had to identify separate conduct to find Perrault guilty of each count.

Fourth, in its closing argument, the government reviewed each of the counts, linking the specific evidence that supported each count. For example, in discussing counts 1, 2, and 7, the government directed the jury to Doe 1’s testimony: “He said *49 that the Defendant put his fingers in his rectum several times, at the base, at the cemetery. While at the base, different times, in the chapel, in the car, in the parking lot field area. So you’ve got enough evidence on Counts 1, 2 and 7.” R. vol. 3 at 1803. This highlighted separate acts of sexual abuse in distinct locations. With regard to counts 3 and 6, the government again referenced Doe 1’s testimony and distinguished the counts based on separate conduct at separate locations—one count for the sexual abuse at the Base and one count for the abuse at Santa Fe National Cemetery. Based on the above, we’re confident the jury linked each count to separate acts of sexual abuse.

Even so, we agree with Perrault that the jointly submitted instructions weren’t ideal. Perhaps the simplest, most effective revision would have been to include a separate jury instruction for each count. Or the district court could have stuck with the parties’ proposed Instruction 1, which included the Indictment. The Indictment would have reminded the jury that several of the counts took place in different locations and that each count related to a distinct charge of sexual abuse.

But even if we concluded that the district court erred by failing to make the instructions clearer, we couldn’t conclude that the error was plain—“obvious under current well-settled law”—as required when conducting plain-error review. Chavez-Morales, 894 F.3d at 1214. Perrault directs us to no authority from either the Supreme Court or our Circuit holding that a district court erred by referencing multiple counts in a single instruction or by failing to include the indictment as part of the jury instructions. Though this case demonstrates the perils of that course, our *50 precedents compel us to uphold the instructions when, as here, we can see how the jury linked individual counts to separate conduct. See McGill, 359 F. App’x at 60–61; Davis, 55 F.3d at 520.

Finally, even if Perrault managed to clear the first two hurdles of plain error review, he couldn’t surmount the third, which requires a showing that the error affected his substantial rights. United States v. Mendoza, 698 F.3d 1303, 1310 (10th Cir. 2012). To prove that the district court’s error affected his substantial rights, Perrault has the burden to show “a reasonable probability that, but for the error claimed, the result of the proceeding would have been different.” Id. (quoting United States v. Weiss, 630 F.3d 1263, 1274 (10th Cir. 2010)). “A reasonable probability is a probability sufficient to undermine confidence in the outcome.” United States v. Rosales-Miranda, 755 F.3d 1253, 1258 (10th Cir. 2014) (citation omitted). “We therefore must ask whether we are in doubt that,” had the jury received clearer jury instructions, it would not have convicted Perrault of all seven counts. See id.

We are confident that, even if the district court had given the jury instructions revised as Perrault suggests, the jury would have returned the same guilty verdict. Doe 1’s testimony—bolstered by seven Rule 414 witnesses’ testimony—provided ample evidence to support the jury’s guilty verdict. Doe 1 testified that Perrault sexually abused him on dozens of occasions, including at least ten times at both the Base and the Santa Fe National Cemetery. The jury obviously found Doe 1 credible. Perrault fails to explain why the jury would believe Doe 1’s allegations about some but not all of the incidents of sexual abuse. And even beyond Doe 1’s testimony, the *51 government submitted compelling evidence that Perrault had admitted abusing other boys. In the face of this considerable evidence, Perrault fails to persuade us of a reasonable probability the outcome would have been different with better instructions.20

IV. The District Court Didn’t Plainly Err by Enhancing Perrault’s Sentence for Obstructing Justice

Perrault argues that the district court erroneously included an obstruction-of-justice enhancement in calculating his advisory guideline range. Perrault concedes that he failed to object to the enhancement, but he maintains that his challenge meets plain-error review. So he must demonstrate plain error that affected his substantial rights and seriously affected the fairness and integrity of the proceeding. See Chavez-Morales, 894 F.3d at 1214. Though we agree with Perrault that the district court erred by applying the obstruction-of-justice enhancement for his flight to Morocco, we affirm on a different ground.21

20 Perrault also argues that, because “counts 1, 2 and 7 for sexual abuse by digital penetration [18 U.S.C. § 2241(c)] subsume counts 4 and 5 for sexual contact [18 U.S.C. § 2244(a)(5),] . . . conviction on both results in double jeopardy.” Appellant’s Opening Br. at 51 (citation omitted). But Perrault concedes that “[t]his court has not determined whether digital sexual abuse subsumes digital sexual contact as a lesser included offense.” Id. at 52 (citation omitted). Thus, because we’re reviewing for plain error, we needn’t take up this argument—any error wouldn’t have been “plain.” 21 The government argues that Perrault’s argument “is so conclusory that this Court should consider the argument waived.” Appellee’s Answer Br. at 42 (citation omitted). But we view Perrault’s argument as concise, not conclusory, and deserving of our review.

The PSR states: “After Perrault discovered an investigation against him was initiated, he obstructed justice by taking a sabbatical and leaving the country.” R. vol. 2 at 223. The PSR then recommends adding two offense levels to each of the seven counts when calculating Perrault’s sentence. The district court adopted the PSR’s guidelines calculation, placing Perrault at an offense level of 40. Because his criminal-history category was I, the advisory guideline range was 292 to 365 months. On counts 1-4, 6, and 7 the district court sentenced Perrault to a 365-month prison term. On count 5, it sentenced him to a concurrent term of 120 months.

But we conclude that the PSR improperly recommended applying the obstruction enhancement for Perrault’s flight to Morocco. Under the applicable guideline, sentencing courts impose a two-level adjustment for obstructing justice when “the defendant willfully obstructed or impeded, or attempted to obstruct or impede, the administration of justice during the investigation, prosecution, or sentencing of the instant offense.” U.S. Sentencing Guidelines Manual § 3C1.1 (U.S. Sentencing Comm’n 1991). The application notes provide that the adjustment is not warranted based on a defendant’s “avoiding or fleeing from arrest.” Id. cmt. n.4(d).

As mentioned, § 3C1.1 applies to obstructive conduct “during the investigation, prosecution, or sentencing of the instant offense.” United States v. Gacnik, 50 F.3d 848, 851 (10th Cir. 1995); see also United States v. Stolba, 357 F.3d 850, 852 (8th Cir. 2004) (“We conclude that an obstruction adjustment was unavailable . . . because no official investigation relating to Mr. Stolba’s offenses was underway when he directed Ms. Morgan to delete the computer files. Section *53 3C1.1 unambiguously requires obstructive conduct to have occurred ‘during’ investigation, prosecution, or sentencing, and at the time that Mr. Stolba directed Ms. Morgan to delete the files, no governmental entity had started investigating or even become aware of Mr. Stolba’s fraudulent conduct.”).

Here, the government never established that law enforcement had even begun investigating Perrault when he fled the country. On appeal, it merely asserts that “Perrault knew he was under investigation when he fled to Morocco.” Appellee’s Answer Br. at 43. But like the PSR, the government omits who was investigating. It was a local reporter, not any governmental entity, who had begun investigating claims about Perrault’s misconduct. Though the reporter’s investigation likely precipitated Perrault’s flight, that situation doesn’t meet § 3C1.1’s plain language.

The government’s arguments fail to persuade us otherwise. It cites two cases to establish that defendants who have fled arrest warrants receive the obstruction enhancement. But in both cases, the defendant fled after he had previously been arrested and had been released pending trial. See United States v. Glenn, 166 F.3d 1222 (10th Cir. 1999) (unpublished table decision); United States v. Cisneros, 846 F.3d 972, 974–75 (7th Cir. 2017). Because law enforcement had begun criminal investigations at the time the defendants fled arrest, those decisions comport with § 3C1.1’s plain language requiring that the obstruction of justice occur “during” an investigation, prosecution, or sentencing. See Gacnik, 50 F.3d at 852.

But we affirm on a different ground. See Spring Creek Expl. & Prod. Co., LLC v. Hess Bakken Inv., II, LLC, 887 F.3d 1003, 1032–33 (10th Cir. 2018) (“[W]e may *54 affirm the district court for any reason supported by the record.” (quoting Amro v. Boeing Co., 232 F.3d 790, 796 (10th Cir. 2000))). Though the government focused its argument on Perrault’s leaving the country (before law enforcement had launched an investigation), we’re more concerned with his failure to return voluntarily to stand trial. Other circuits have concluded that putting the government to the expense and hassle of retrieving a defendant from a foreign country constitutes obstruction of justice. See United States v. Nduribe, 703 F.3d 1049, 1050–51 (7th Cir. 2013) (affirming the district court’s applying an obstruction enhancement when the defendant had “delayed his apprehension by five years and during that period put the government to the expense of searching for him on three continents before finally procuring his arrest and extradition, which undoubtedly involved our government’s active participation in the extradition proceeding”); see also United States v. Carty, 264 F.3d 191, 195 (2d Cir. 2001) (upholding application of obstruction enhancement in part because the defendant “did not return to the United States after being ordered to do so by a D.E.A. agent”). We agree with those decisions that this kind of conduct ordinarily triggers the obstruction enhancement.

Here, even after he was detained by the Moroccan government at the United States’ request, Perrault fought his removal from Morocco. See R. vol. 3 at 1425:24– 25 (“Morocco agreed to expel him from the country, so we [i.e., FBI agents] had to go over there to recover him.”). He even wrote a letter to Morocco’s King, pleading for permission to stay in Morocco. By then, Perrault was well-aware of the charges against him. But he still put the government to the expense and trouble of retrieving *55 him from abroad. This supports the district court’s imposing the obstruction-of-justice enhancement, and we affirm on that basis. V. Perrault Fails to Demonstrate That Cumulative Error Tainted His Trial

A. Standard of Review

A cumulative-error analysis “is an extension of the harmless-error rule.” United States v. Rivera, 900 F.2d 1462, 1469 (10th Cir. 1990) (en banc). This doctrine recognizes that “[t]he cumulative effect of two or more individually harmless errors has the potential to prejudice a defendant to the same extent as a single reversible error.” Id. To assess that possibility, we “aggregate[] all errors found to be harmless and analyze[] whether their cumulative effect on the outcome of the trial is such that collectively they can no longer be determined to be harmless.” Grant v. Royal, 886 F.3d 874, 954 (10th Cir. 2018) (quoting Hooks v. Workman, 689 F.3d 1148, 1194 (10th Cir. 2012)). Unless the court identifies at least two harmless errors, we will decline to undertake a cumulative-error analysis. Hooks, 689 F.3d at 1195.

B. There Was No Cumulative Error

Besides the alleged errors already discussed,22 Perrault argues that two additional errors tainted his trial. He asserts that the district court erred by

22 In arguing for cumulative error, Perrault repackages his arguments that the Rule 414 witnesses were overly prejudicial and that the jury instructions lacked clarity. Because we have already addressed those arguments above, we don’t separately consider them here.

(1) denying his motion for a continuance and (2) denying his motion for a mistrial. Neither argument has merit.

We review for an abuse of discretion the district court’s decision to deny Perrault’s motion for a continuance, United States v. Toombs, 574 F.3d 1262, 1268 (10th Cir. 2009), and his motion for a mistrial, United States v. Wells, 739 F.3d 511, 533 (10th Cir. 2014). We will reverse only if we have “a definite and firm conviction that the lower court has made a clear error of judgment or exceeded the bounds of permissible choice in the circumstances.” United States v. Hargrove, 911 F.3d 1306, 1316 (10th Cir. 2019) (citation omitted).

1. The District Court Acted Within Its Discretion in Denying Perrault’s Motion for a Continuance

Perrault asserts that “[n]o attorney could have prepared adequately for this trial in six-months.” Reply Br. at 25. According to Perrault, the district court’s denying his third motion to continue prejudiced his trial preparation. We disagree.

The district court granted (at least in part) two of Perrault’s three motions for a continuance. The court originally scheduled Perrault’s trial for November 26, 2018. But the district court granted Perrault’s first unopposed motion to postpone the trial by three months to February 25, 2019. In January 2019, Perrault filed a second motion to continue, asking for another ninety-day extension. The district court granted an extension of just over thirty days and rescheduled the trial for April 1, 2019.

But when Perrault moved for a third time to continue the trial an additional forty-five days, the district court denied the motion. It determined that Perrault had failed to meet the Toombs requirements because he had failed to explain “what further investigation is necessary or what possible motions might be filed.” R. vol. 1 at 366. Under Toombs, “short, conclusory statements lacking in detail” are “insufficient” to justify a continuance under the ends-of justice exception to the Speedy Trial Act. 574 F.3d at 1271–72.

Perrault offered only conclusory statements to the district court, and that’s all he offers now. He continues to rest on his assertion that it was “obvious” that his trial counsel couldn’t adequately prepare for his trial in six months. Reply Br. at 25. Yet he never explains what he would have done with the extra time or how the denial of his request for a continuance prejudiced his trial preparation. Given Perrault’s inability to articulate why he needed additional time to prepare, the district court acted within its discretion in denying his third motion for a continuance.

2. The District Court Didn’t Err in Denying Perrault’s Motion for a Mistrial

When attempting to impeach Perrault’s character witness, a friend of his, the government queried, “When [Perrault] returned to the United States in 2018, you visited him in jail; isn’t that right?” R. vol. 3 at 1633. Perrault immediately moved for a mistrial, arguing that the question violated Perrault’s presumption of innocence. Though the district court acknowledged that the question “should not have been asked,” it denied Perrault’s motion for a mistrial. Id. at 1649, 1651. Considering the *58 question in the context of all the evidence, the district court concluded that a curative instruction would protect Perrault’s right to a fair trial.

The district court didn’t abuse its discretion in denying Perrault’s motion for a mistrial. A “district court has discretion to grant a mistrial only when a defendant’s right to a fair and impartial trial has been impaired.” United States v. Meridyth, 364 F.3d 1181, 1183 (10th Cir. 2004) (citing United States v. Kravchuk, 335 F.3d 1147, 1155 (10th Cir. 2003)). In assessing whether that right was impaired, we consider “the prejudicial impact of an error or errors when viewed in the context of an entire case.” Id. (quoting United States v. Gabaldon, 91 F.3d 91, 93–94 (10th Cir. 1996)). Here, Perrault received a fair and impartial trial notwithstanding the government’s improper question. It was no secret that Perrault had been detained before trial; the jury had already heard evidence that the Moroccan government had detained Perrault and that FBI agents had traveled overseas to retrieve him. The district court’s curative instruction redressed whatever minimal prejudicial impact the prosecutor’s question had on the trial. In short, the district court acted within its discretion in denying the motion.

CONCLUSION

For the foregoing reasons, we affirm.