United States v. Charles Hall

United States Court of Appeals For the Eighth Circuit ___________________________ No. 14-2742 ___________________________ United States of America Plaintiff - Appellee v. Charles Michael Hall Defendant - Appellant ____________ Appeal from United States District Court for the Western District of Missouri - Springfield ____________ Submitted: January 15, 2019 Filed: December 19, 2019 ____________ Before LOKEN, GRASZ, and STRAS, Circuit Judges. ____________ STRAS, Circuit Judge. Charles Hall and Wesley Coonce killed Victor Castro-Rodriguez after brutally beating him. A jury found Hall guilty of first-degree murder, see 18 U.S.C. § 1111(b), and following the presentation of a number of aggravating and mitigating factors, returned a death sentence. Hall appeals only his sentence and challenges, among other issues, the aggravating and mitigating factors the jury considered, the evidence it heard, and the instructions it received. We affirm. I. The three men were serving federal sentences in a mental-health ward at a medical center for federal prisoners in Springfield, Missouri, when Hall and Coonce attacked Castro-Rodriguez. As Hall admitted, they had discussed the crime two or three days in advance. Hall has said that killing calms him and that if Castro- Rodriguez had not been available, he “would have randomly selected another inmate” to kill instead. According to the testimony of one inmate, Hall lured Castro-Rodriguez by promising him money if he would pretend to be a hostage. Hall allegedly told him that if he would go along, prison officials might agree to give them additional privileges, including cable television, in exchange for his release. After Castro- Rodriguez agreed, Hall and Coonce followed him into his cell, where they made him lie on his back, bound his hands and feet, stuffed a rag in his mouth, and blindfolded him. Once he could no longer move or call for help, Coonce repeatedly kicked him and stomped on his neck. Hall then stood on his throat. After several minutes, Hall stepped off, checked for a pulse, and punched his stomach “to see if he would react.” An autopsy revealed that Castro-Rodriguez died from suffocation caused by compression of his larynx, although he also had internal bleeding, scrapes, and bruises from the repeated blows to his head, neck, and chest. Hall and Coonce were tried together. The jury found them both guilty after deliberating for less than three hours. At a joint sentencing hearing, the jury heard evidence about various aggravating and mitigating factors. After deliberating again—this time into a second day—the jury unanimously recommended a death -2- sentence for both of them. The district court 1 accepted the jury’s recommendation and entered judgment. II. The first question is whether the district court abused its discretion by trying Hall and Coonce together. See United States v. Ortiz, 315 F.3d 873, 898 (8th Cir. 2002). Both initially objected to a joint trial, but after the jury found them guilty, Hall withdrew his objection to a joint sentencing hearing. His position now is that the refusal to sever the proceedings at the guilt phase allowed the jury to hear evidence that unfairly prejudiced him during the penalty phase. A joint trial is “often preferable when the joined defendants’ criminal conduct arises out of a single chain of events,” because it gives the jury a chance “to assign fairly the respective responsibilities of each defendant.” Kansas v. Carr, 136 S. Ct. 633, 645 (2016) (citation omitted). This case is a good example. Hall and Coonce agreed ahead of time to murder Castro-Rodriguez. Once they entered his cell together, Hall tied him up, Coonce took the lead in assaulting him, and then Hall dealt the fatal blow by standing on his throat. They acted side by side at every step, so it was logical for the district court to allow “their fates [to be] determined by a single jury.” Id. at 646. To be sure, holding a joint trial resulted in the same jury hearing both Coonce’s guilt-phase defense that Hall was the driving force behind the crime and the government’s case for the death penalty. This made the proceeding unfair, according to Hall, because Coonce’s theory “logically supported” and supplemented the government’s penalty-phase argument that Hall would likely be violent in the future. 1 The Honorable Gary A. Fenner, United States District Judge for the Western District of Missouri. -3- Even if these arguments overlapped to some degree, Hall cannot establish that Coonce’s guilt-phase evidence “so infected the sentencing proceeding with unfairness as to render the jury’s imposition of the death penalty” unconstitutional. Id. at 644–45 (citation omitted). In arguing that he was less blameworthy than Hall, Coonce never suggested that Hall, as the alleged mastermind, would do something similarly violent in the future. Nor did the government connect those dots. Its theory was simpler: Hall was dangerous because he repeatedly said he would commit crimes and had a lengthy history of threatening violence. See infra Part IV.B (addressing the admissibility of the government’s evidence). For us to accept Hall’s theory that the joint trial unfairly prejudiced him would require “an exercise in speculation, rather than reasoned judgment.” Romano v. Oklahoma, 512 U.S. 1, 14 (1994). And speculation is not enough. See Carr, 136 S. Ct. at 646. III. Hall’s next group of challenges focuses on the aggravating and mitigating factors presented to the jury. In determining whether a death sentence was “justif[ied],” the jury’s task during the penalty phase was to “consider” whether the aggravating factors “sufficiently outweigh[ed]” the mitigating factors. 18 U.S.C. § 3593(e); see also id. § 3592 (identifying potential aggravating and mitigating factors). A. The federal death-penalty statute divides aggravating factors into two categories: the listed ones that the jury “shall” consider and those “other aggravating factor[s]” that it “may” consider. Id. § 3592(c). Despite this open-ended language, Hall claims that one unlisted factor, grave indifference to human life, should never -4- have been submitted to the jury. 2 He points out that the jury already considered his mental state once in determining that he was eligible for the death penalty, so doing it again improperly counted the same factor twice. The premise is correct, but his conclusion is not. Nothing in the federal death-penalty statute says that the jury can only consider a defendant’s mental state once. To the contrary, the statute contemplates the possibility that the jury will do so at least twice, first when determining whether the death penalty is on the table at all and again when evaluating the aggravating factors presented by the government. Compare id. § 3591(a)(2) (listing the various mental states giving rise to death-sentence eligibility), with id. § 3592(c)(5) (requiring a “grave risk of death”); id. § 3592(c)(6) (committing the offense in an “especially heinous, cruel, or depraved manner”); id. § 3592(c)(9) (demonstrating “substantial planning and premeditation”); id. § 3592(c)(16) (having the intent to kill or attempting “to kill more than one person in a single criminal episode”). And some of the factors even seem to overlap with one another, further weakening Hall’s double-counting theory. See, e.g., id. § 3592(a)(1) (acting with a “significantly impaired” capacity to “appreciate the wrongfulness” of the conduct); id. § 3592(a)(6) (carrying out the offense “under severe mental or emotional disturbance”). Hall’s argument, in other words, is contrary to the statute, which constructs a two-stage process with overlapping inquiries, see id. §§ 3591(a)(2), 3592(a), (c), and is inclusive when it comes to the factors that a jury may consider, see id. § 3592(c). 2 Hall also argues that future dangerousness should not have been submitted as an aggravating factor because a prediction about events yet to occur cannot be proved beyond a reasonable doubt. See 18 U.S.C. § 3593(c) (requiring the government to prove the “existence of any aggravating factor” beyond a reasonable doubt). We recently considered and rejected this argument, see United States v. Coonce, 932 F.3d 623, 642–43 (8th Cir. 2019), and do so here too. -5- B. Hall challenges another aggravating factor for a different reason. He argues that the government did not provide enough evidence for the jury to find that he killed Castro-Rodriguez in an especially heinous, cruel, or depraved manner. See id. § 3592(c)(6). In explaining this factor, the district court told the jury that the killing was heinous if it was “extremely wicked or shockingly evil”; it was cruel if Hall “intended to inflict a high degree of pain”; and it was depraved if Hall “relished the killing or showed indifference to [Castro-Rodriguez’s] suffering.” The government also had to show “serious physical abuse,” which the instructions defined as “a significant or considerable amount of injury or damage to [Castro-Rodriguez’s] body.” See 18 U.S.C. § 3592(c)(6) (requiring “torture” or “serious physical abuse”); see also United States v. Montgomery, 635 F.3d 1074, 1095–96 (8th Cir. 2011) (defining “serious physical abuse” to include “inflict[ing] suffering . . . above and beyond that necessary to cause death” (internal quotation marks and citation omitted)). Contrary to Hall’s suggestion, the government’s evidence was sufficient to find that the killing was heinous, cruel, or depraved, and that it involved the infliction of serious physical abuse. The primary source of the government’s evidence was Hall’s initial statement to investigators shortly after the murder. In it, he recounted how he had bound Castro-Rodriguez’s hands and feet, stuffed a rag in his mouth, blindfolded him, watched Coonce assault him, and then stood on his throat until he stopped breathing. He also acknowledged the senselessness of the crime, explaining that if Castro-Rodriguez had not been available, he “would have randomly selected another inmate and killed him” instead. The medical examiner, after describing the injuries and confirming that they were consistent with Hall’s account, estimated that it likely took “more than three minutes, maybe [even] five minutes” for Castro- -6- Rodriguez to die and that it was possible that he remained conscious nearly the entire time. From this evidence, a reasonable juror could conclude that Hall intended to “inflict a high degree of pain” and was “indifferen[t] to [Castro-Rodriguez’s] suffering.” The fact that Castro-Rodriguez had been tricked and rendered helpless only increased his distress, and the prolonged beating inflicted gratuitous pain. And even as the assault was coming to a close, Hall extended Castro-Rodriguez’s suffering by suffocating him slowly, even as he likely remained conscious. It is true, as Hall argues, that other death-penalty cases have involved greater cruelty and more extreme violence. See, e.g., id. at 1079–80. But those cases do not set the floor, and the fact that some murders are even more heinous, cruel, or depraved does not mean that this one was not. C. Turning to the mitigating factors, the district court allowed the jury to consider, as relevant here, Hall’s good behavior in prison after the murder and his difficulties with a chronic gastrointestinal disease, neither of which was disputed factually. Even so, some jurors found that the factors did not “exist[],” which Hall attributes to a failure to follow the court’s instructions to make factual findings on each of the mitigating factors first, before deciding whether to assign them weight. We agree that the most likely explanation is that some jurors misunderstood the instructions on the the verdict form. The form required the foreperson to record “the number of jurors who . . . found the existence of [each] mitigating factor to be proven by a preponderance of the evidence,” yet she wrote “6” in the space for Hall’s medical condition and “0” in the other. Following their discharge, some jurors informed the district court that the foreperson may have recorded the number who voted to give these factors mitigating weight, rather than the number who found that -7- they “had been proved.” See Fed. R. Evid. 606(b)(2)(C) (providing an exception for juror testimony when it involves “a mistake . . . in entering the verdict on the verdict form”). Given that the government did not dispute either factor, however, there was no reason for the foreperson to write anything other than “12” for both. Unlike Hall, however, we are convinced beyond a reasonable doubt that any mistake the jury may have made was harmless. See 18 U.S.C. § 3595(c)(2). As a baseline principle, the jurors were not required to give any mitigating weight to Hall’s post-murder good behavior or his medical condition, see United States v. Paul, 217 F.3d 989, 999 (8th Cir. 2000), even if he had a right to present them to the jury. Only if the jurors assigned no weight to them because they believed, “as a matter of law,” that they were “unable even to consider the evidence” would a constitutional problem arise. Eddings v. Oklahoma, 455 U.S. 104, 113 (1982). Nothing in the record suggests that this was the case. To the contrary, the fact that six jurors assigned some weight to Hall’s medical condition suggests that they understood that it was their call to make. See Paul, 217 F.3d at 1000 (explaining that the test is whether “there exists a reasonable likelihood that the jurors believed themselves precluded from considering relevant mitigating evidence”). There is, in short, no reason to believe that marking the form with one piece of information rather than another affected the jury’s decision-making or changed its overall conclusion. IV. Hall also challenges several evidentiary rulings. Under the federal death- penalty statute, the standard for admission of evidence at penalty-phase proceedings is “relaxed”: parties may present ordinarily inadmissible evidence without worrying about the Federal Rules of Evidence. United States v. Purkey, 428 F.3d 738, 758 (8th Cir. 2005); see 18 U.S.C. § 3593(c). The idea is that the more information the jury has when choosing between life and death, the better. See United States v. Lee, 274 F.3d 485, 494 (8th Cir. 2001). Yet the statute also vests the district court with -8- broad discretion to exclude evidence if it is irrelevant or “its probative value is outweighed by the danger of creating unfair prejudice, confusing the issues, or misleading the jury.” 18 U.S.C. § 3593(c). Compare id. (requiring only that the probative value be “outweighed” by the “danger[s]” created by the evidence), with Fed. R. Evid. 403 (adopting a similar standard, but requiring the probative value to be “substantially outweighed” by these risks (emphasis added)). The court has more gatekeeping power, in other words, even though a greater range of evidence can potentially pass through the door. See Purkey, 428 F.3d at 756. A. The first evidentiary ruling that we have been asked to review was the decision to limit the scope of Dr. Park Dietz’s cross-examination. The government called him to rebut the opinions of several mental-health experts who testified that Hall’s bipolar disorder caused him to lose self-control. While explaining his qualifications, Dr. Dietz stated that he had been asked to testify in over one thousand cases, including some high-profile prosecutions. He admitted that in one of them, he had made false statements that eventually led to an overturned conviction. See Yates v. State, 171 S.W.3d 215, 221–22 (Tex. App. 2005). The district court allowed Hall’s attorney to thoroughly question Dr. Dietz about what happened in Yates. But it would not allow him to question Dr. Dietz about a judge’s refusal to appoint him as a court-appointed expert in another case, Gates v. Chappell, No. C 88-2779 WHA (N.D. Cal.). Because the judge in Gates had refused to appoint Dr. Dietz based solely on concerns about his impartiality after Yates, the court reasoned, there was little to be gained from the additional -9- questioning. In Hall’s view, cutting off this line of inquiry was an abuse of discretion. 3 See 18 U.S.C. § 3595(c)(2). We disagree. The additional questioning had low probative value even if it was arguably relevant. The first possibility was that it could have shed light on Dr. Dietz’s credibility. But the opinion of a single district judge on his fitness for a different type of role in an unrelated case only marginally assisted the jurors in answering the critical question they faced: should they believe his testimony? The second possibility was that it could have rebutted Dr. Dietz’s claim that he had testified in over one thousand cases. See id. § 3593(c) (allowing both the government and the defendant to “rebut any information received at the hearing”). This is an impressive-sounding number, and as Hall points out, Gates was one case in which Dr. Dietz was not allowed to testify. Still, the evidence had limited probative value, especially after Dr. Dietz admitted that the figure was not as impressive as it sounded. Earlier in his career, he explained, he worked in a hospital and participated in about a dozen commitment hearings per week. More recently, by contrast, he had not appeared in court nearly as often. Not only was the probative value low, it carried a high risk of confusing or misleading the jury. It is unlikely that the jurors could have understood the differences between the roles of a court-appointed expert and a government witness, at least without additional factual development. In fact, some may well have concluded that a federal judge’s decision not to appoint Dr. Dietz superseded their own opinions about his testimony, no matter what he said about Hall. Cf. Rosales- 3 Hall also claims that the decision violated his rights under the Confrontation Clause of the Sixth Amendment. As we recently explained, however, the Confrontation Clause does not apply to sentencing proceedings, even those involving a potential death sentence. See Coonce, 932 F.3d at 640–41; see also Williams v. New York, 337 U.S. 241, 251–52 (1949). -10- Lopez v. United States, 451 U.S. 182, 188 (1981) (stating that jurors “must reach conclusions as to impartiality and credibility by relying on their own evaluations of demeanor evidence and of responses to questions”). Given these risks, we cannot say that the court abused its discretion in cutting off this line of questioning. See id. B. Another contested evidentiary ruling was allowing the government to prove Hall’s future dangerousness through, among other things, evidence of threats Hall had made against various officials, his pre-murder misbehavior in prison, and his numerous prior crimes. According to Hall, this evidence was minimally probative and highly prejudicial and never should have been admitted. During the penalty-phase proceedings, the jury heard that Hall had threatened to kill a federal judge and prosecutors; to bomb various sensitive locations, including an airport and President George H.W. Bush’s home; and to orchestrate a mass poisoning. This evidence was presented along with various statements that he had made to psychiatrists, psychologists, nurses, and other staff about wanting to harm or kill people in prison. As the government pointed out, he even acted on those desires once before murdering Castro-Rodriguez, when he assaulted another inmate. Although the victim suffered only minor injuries, Hall later stated in a letter to a psychiatrist that his intent “wasn’t . . . just to punch him. . . . [It] was to choke him to death.” The letter went on to say that if he “was on the open unit today, [killing someone] would” have been his “intent and goal.” The jury also heard that Hall had difficulty following the rules, both inside and outside of prison. The government introduced evidence that he had been convicted of a number of crimes over a 25-year period, including burglary, forgery, theft, false public alarm, receiving stolen goods, and assault. His behavior was problematic in prison too, where he fought, gambled, and possessed contraband prior to the murder. -11- All of this evidence was relevant. See 18 U.S.C. § 3593(c). A reasonable juror could view Hall’s threats as reflecting a propensity for violence. It is true, of course, that Hall could have been bluffing, as he now claims, knowing full well that he could not possibly carry out some of his threats. But that was a credibility determination for the jury to make. See id.; see also Lee, 274 F.3d at 494 (discussing the “very low barriers to the admission of evidence”). To be sure, the evidence was of mixed value. For example, Hall’s decades- old convictions, some for nonviolent crimes, arguably shed less light on his future dangerousness. But collectively, the evidence tended to show that Hall wanted to hurt others and might do so if given an opportunity, regardless of where he was or what constraints were placed on him. On the other side of the scale, the evidence was only prejudicial in the sense that it made it hard for Hall to argue that he was not dangerous. But there was nothing unfair about its admission. Cf. United States v. Looking Cloud, 419 F.3d 781, 785 (8th Cir. 2005) (“Evidence is not unfairly prejudicial because it tends to prove guilt, but because it tends to encourage the jury to find guilt from improper reasoning.”). It did not cause confusion or lead the jury to consider a factor it should not have in making its sentencing recommendation. Without showing that one of those things may have occurred, Hall has given us little reason to question the district court’s discretionary decision to admit it. See Lee, 274 F.3d at 494 (“[D]etermining whether there is a threat of unfair prejudice [under 18 U.S.C. § 3593(c)] is a fact specific inquiry . . . .”). C. Hall also questions the district court’s decision to allow the government to introduce two letters that he sent to the prosecutor before trial. In the first, he expressed his desire to be executed and offered to “enter a ple[a] of guilty” if the -12- prosecutor “could guarantee” that he received a death sentence. He also explained that “[t]he only thing that [would] stop [him] from killing again [was] to put [him] to death.” In the second, which was delivered both to the prosecutor and to his own attorney, he explained in detail why he thought he was eligible for the death penalty, directed his attorney to work toward a death sentence, and requested a guilty plea “as long as” the prosecutor would recommend one. The government did not respond to either letter. The district court admitted both letters during the guilt phase of Hall’s trial, but they played a significant role at sentencing too. See 18 U.S.C. § 3593(c) (allowing the jury to consider information from the guilt phase when evaluating aggravating and mitigating factors). Because Hall is appealing only his sentence, however, we review their admissibility under the “more lenient” evidentiary standard applicable at the penalty phase, Purkey, 428 F.3d at 756, leaving aside whether the jury should have seen them earlier when it was evaluating Hall’s guilt. Coonce, 932 F.3d at 637. Under this standard, the district court did not abuse its discretion when it admitted the letters. They contained expressions of guilt and threats to kill again, both of which were highly probative of several aggravating factors, including his future dangerousness, lack of remorse, and indifference to human life. It does not matter that admitting the letters left him with an uphill battle during the penalty-phase proceedings, because there was nothing unfair about their admission, even if, as Hall claims, his “notions about the case changed dramatically after [he] benefitted from different legal advice offered by new counsel.” Cf. Looking Cloud, 419 F.3d at 785 (explaining that “unfair prejudice” refers to encouraging “the jury to find guilt from improper reasoning”). Just because he changed his mind does not mean that it was unfair to use his earlier admissions -13- against him. 4 Cf. United States v. Muhlenbruch, 634 F.3d 987, 1001 (8th Cir. 2011) (holding that a district court did not abuse its discretion under Rule 403 when it permitted the government to play a defendant’s videotaped confession). His argument to the contrary is really just an attempt to import Federal Rule of Evidence 410(a)(4)—which prohibits using statements made during plea negotiations against a defendant at trial—into penalty-phase proceedings. But this theory gets him nowhere because, according to the federal death-penalty statute, “[i]nformation is admissible” at the sentencing hearing “regardless of its admissibility under the rules governing admission of evidence at criminal trials.” 18 U.S.C. § 3593(c). So it makes no difference whether, as Hall insists, the statements would otherwise count as having been “made during plea negotiations.” They were admissible either way. D. The final evidentiary issue was the exclusion of Hall’s “comparative proportionality” evidence, which summarized the aggravating and mitigating circumstances in other death-penalty cases. Hall’s theory was that the jury might be less likely to think that he deserved the death penalty if it could evaluate how he compared to others who committed similar crimes. The evidence consisted of short 4 Hall vaguely asserts that allowing the jury to see that he had previously asked for the death penalty was unfairly prejudicial. But this is a different argument than he raised before the district court or in his briefs, which took the position that the letters were completely inadmissible because they contained statements made during plea negotiations. See Fed. R. Evid. 410(a)(4). In any event, Hall never sought a limiting instruction or argued that the court should have redacted the letters, so we can hardly fault it for failing to take one of these actions on its own. Cf. United States v. Melton, 870 F.3d 830, 837–38 (8th Cir. 2017) (holding that admitting a document containing potentially prejudicial statements was not an abuse of discretion because, although the defendant requested a limiting instruction, he “did not pursue the government’s offer to redact” them). -14- summaries of sixteen cases—all involving murders committed in federal prison— along with a list of the aggravating and mitigating factors in each case. Also included were three spreadsheets that repackaged this information in various ways. As the district court recognized, the evidence had limited probative value. The most obvious shortcoming was that no one had any way of knowing whether the selected cases were representative. According to Hall’s attorney, they shared three characteristics: each went to trial, a death sentence was on the table, and information was available from public sources or the attorneys who tried them. This approach made sense from a practical, information-gathering perspective, but relevant information could have been missing for a variety of reasons. Indeed, Hall has trimmed his list down even more on appeal. The explanation, according to his brief, is that “the detailed facts . . . necessary for a proper comparative analysis could not be assembled” for five of his original sixteen cases. It is unclear what, if anything, changed, but this development highlights the problems posed by the evidence. The presentation of the cases posed other potential problems. The nuances were arguably lost in the long lists of aggravating and mitigating factors that he provided. For example, two cases might involve the same future-dangerousness aggravating factor, but a jury might find a death sentence appropriate in one case and not the other, leaving the jurors in this case to wonder why. Compare United States v. Houston, 648 F.3d 806, 811–12 (9th Cir. 2011) (involving a life sentence for a racially-motivated-stabbing death), with United States v. Snarr, 704 F.3d 368, 377 (5th Cir. 2013) (imposing a death sentence for a stabbing attack that killed a fellow inmate and left two prison guards wounded). Admitting the evidence could have opened the door to a debate about whether Hall’s lists and spreadsheets adequately captured their details, not to mention whether they were similar enough to Hall’s case. If so, the potential for confusion and distraction was clear. Another more serious risk was also looming. Using lists and spreadsheets potentially invited the jurors to view their assignment in Hall’s case in similar terms: -15- simply tally up the aggravating and mitigating factors and see where he fell on the list, rather than undertake an “individualized determination” based on the facts and circumstances of his case. Tuilaepa v. California, 512 U.S. 967, 972 (1994) (citation omitted); see Purkey, 428 F.3d at 762 (explaining that the federal death-penalty statute “avoids arbitrary death sentences by requiring juries to weigh [the] aggravating and mitigating factors rather than [just] tally the factors on each side and declare a winner based on sheer numbers”). It is possible, of course, that the district court could have found a way to manage these difficulties. But this was a discretionary call, and the court was entitled to conclude that admitting the evidence would have caused more trouble than it was worth. 5 See 18 U.S.C. § 3593(c) (giving the district court broad authority as gatekeeper). V. Hall’s remaining challenge is to the district court’s response to two notes from the jury foreperson during penalty-phase deliberations. One of them stated that the jury could not reach a unanimous decision on Hall’s sentence, but the court directed the jury to keep trying. Hall’s position is that it felt pressured after that point to recommend a death sentence. In evaluating Hall’s argument, context matters. See United States v. Walrath, 324 F.3d 966, 970 (8th Cir. 2003) (listing factors that help identify impermissible jury coercion). The jury began its deliberations on Friday afternoon, had the weekend off, and then resumed on Monday morning. Shortly after lunch, the foreperson sent a note asking if “the lawyers have the option to poll the jury if [it] 5 In light of this conclusion, we leave for another day the broader question of whether comparative-proportionality evidence is ever admissible as a mitigating factor. See 18 U.S.C. § 3592(a) (providing a nonexclusive list of mitigating factors). -16- can’t reach a unanimous decision[?]” The court replied that the jury “w[ould] not be polled if a nonunanimous verdict [was] accepted.” The foreperson then sent a second note, this time announcing that the jury had reached a decision on Coonce, but “with 100% certainty” could not reach one on Hall. Hall’s attorney urged the court to declare that the jury was “hung,” which would have resulted in a life sentence. See Jones v. United States, 527 U.S. 373, 380–81 (1999). Instead, the court told the jury to “continue [its] deliberations, and [to] try to reach unanimous verdicts.” After deliberating for approximately another hour, the jury returned a recommendation of death. The district court was faced with a choice: declare the jury deadlocked or direct it to continue its deliberations. Given that the jury had been deliberating for less than one full day, it sent the jury back so that it could keep trying. See United States v. Hagan, 412 F.3d 887, 890 (8th Cir. 2005) (reviewing a coercive-instruction claim for an abuse of discretion). In making this decision, the court did not have to automatically acquiesce in the jury’s assessment. Lowenfield v. Phelps, 484 U.S. 231, 238 (1988). Rather, as the Supreme Court has recognized, courts “incontestably” have the “authority to insist that [jurors] deliberate further,” at least under certain circumstances. Id. This is one of those circumstances. The jury had previously been instructed that its task was to choose between life in prison and a death sentence. Telling the jurors to deliberate further in an effort to reach unanimity did not “coerce” them into picking one alternative over the other. Nor did the court tell the jurors to reconsider their positions or that they must be unanimous in the end, which are the types of statements that have necessitated prophylactic instructions in the past. See United States v. Robinson, 953 F.2d 433, 437 (8th Cir. 1992). Here, the court was simply exercising its judgment that it was sensible under the circumstances to ask the jury to deliberate longer before giving up. Lowenfield, 484 U.S. at 238; see also United States v. Reed, 686 F.2d 651, 652 n.1 (8th Cir. 1982) (“A supplemental instruction merely to continue deliberating after impasse . . . is not properly characterized as a -17- traditional Allen charge.” (citing Allen v. United States, 164 U.S. 492, 501–02 (1896))). The jury’s response showed that it did not feel pressured to choose either alternative. Although it continued to deliberate for only about an hour more, the jury requested three exhibits connected to the lack-of-remorse and future-dangerousness aggravating factors. Cf. United States v. Warfield, 97 F.3d 1014, 1022 (8th Cir. 1996) (“Although the jury’s return with a verdict approximately one hour after receiving the Allen charge is somewhat expeditious, we do not believe the postinstruction deliberation time . . . raises an inference of coercion.”). This timeline suggests that the jurors continued to debate whether the death penalty was justified; discussed and reexamined the evidence that they found most compelling; and eventually agreed on a recommendation. If they felt no choice but to impose the death penalty, as Hall now claims, there would have been little reason to request the exhibits. The bottom line is that nothing here raises a red flag that there may have been coercion. The design of the verdict form does not change our conclusion. Hall argues that by instructing the jurors to “proceed to” consideration of a life sentence if it could not unanimously agree on a death sentence, the district court should have known that the second note reported that the jury had effectively ruled out death.6 There are two flaws in this argument. First, the second note said only that the jury 6 It is true that at some point the foreperson marked “NO” in the space for recommending a death sentence, before scratching it out and marking “YES” instead. But we have no way of knowing when it happened or why. Cf. Blueford v. Arkansas, 566 U.S. 599, 607 (2012) (explaining that until deliberations are over and the jury announces its final decision, jurors are free to “rethink” their votes). Neither did the district court, because even if the foreperson marked “NO” before sending the second note, the court never saw the partially completed form. So this possibility cannot have a bearing on our assessment of its decision to send the jury back for further deliberations. -18- was not unanimous, not that death was off the table. Second, the argument ignores the fact that the jury had a binary choice to make: life in prison or a death sentence. Nothing prevented the jury from moving back and forth between these two alternatives, and the second note simply reflected the fact that, up to that point, there was a lack of agreement on which of the two to choose. Under these circumstances, it was reasonable for the court to conclude that the jury was not trying to return a final verdict and that further deliberations would eventually allow it to do so. VI. Finally, we note that Hall has raised three issues for potential en banc or Supreme Court review: it was unconstitutional to use future dangerousness as an aggravating factor, the district court misstated the law when it instructed the jury that Hall “must be sentenced to death” if it made certain findings, and it was unconstitutional to treat his lack of remorse as both proof of future dangerousness and as a standalone aggravating factor. Controlling precedent squarely forecloses Hall’s positions on these three issues.7 See Barefoot v. Estelle, 463 U.S. 880, 896 (1983) (future dangerousness); Montgomery, 635 F.3d at 1099 (the jury instruction); Purkey, 428 F.3d at 762 (overlapping aggravating factors). VII. We accordingly affirm the judgment of the district court. ______________________________ 7 Moreover, to the extent that Hall argues that he is entitled to a new trial because of cumulative error, this argument fails on its own terms. See supra Part III.C (identifying as the only potential error that the jury may have misunderstood one part of the verdict form); see also United States v. Darden, 70 F.3d 1507, 1549 (8th Cir. 1995) (rejecting a cumulative-error argument because the court had only identified one error). -19-