Full opinion text
KANNE, Circuit Judge. In 2002, state and federal authorities began an intensive investigation into gang violence in Aurora, Illinois. Their investigation centered on the Insane Deuces street gang, as one member had agreed to serve as a confidential informant for local authorities. Mariano Morales, Arturo Barbosa, Miguel Rodriguez, Brian Hernandez, Lionel Lechuga, Harold Crowder, and Romel Handley (collectively, the “Defendants”) all participated in the gang’s activities. They were among sixteen individuals indicted by a federal grand jury on various racketeering-related charges in 2006. After the trial was bifurcated, the seven Defendants stood trial together and were each convicted of racketeering conspiracy; some were also convicted of related charges, including narcotics distribution and conspiracy, illegal firearm possession, and assault and murder in furtherance of their racketeering activities. Once proud, self-identifying Deuces banging on the street, each now tries to distance himself from the gang and claims to have been only tangentially involved in its activities, if at all. The Defendants present three joint and fourteen individual issues, alleging errors in both the guilt and sentencing phases of their trial. Finding that their arguments lack merit or that the errors alleged were harmless, we affirm their convictions and sentences. I. Background The Defendants were members or confederates of the Insane Deuces, a street gang with a heavy presence in northeastern Illinois in the previous decade. With the help of a gang member seeking to avoid prosecution following an arrest, local and federal authorities gathered information on the Insane Deuces and tried to dismantle the gang in Aurora, Illinois, where gang-related violence had soared in the early 2000s. Today we resolve two appeals arising from trials based upon those efforts: this case and its companion case, United States v. Benabe, 654 F.3d 753 (7th Cir.2011). We will briefly describe the background of the gang and its activities before discussing the proceedings leading to the current appeal. A. About the Gang The Insane Deuce Nation (“Insane Deuces,” “Deuces,” or the “Nation”) is an organized street gang affiliated with the Folks, a national network of various local gangs. Recently, the gang has been predominantly composed of Latino/Hispanic males and has operated primarily within northern Illinois, though factions exist throughout Illinois and various state and federal prisons. Cities in northeastern Illinois had individual chapters under the umbrella of the Nation and its centralized leadership. This case involves the Aurora Deuces, a chapter with a significant presence in Aurora, Illinois. A loose association between the Insane Deuces and the Latin Kings, a gang affiliated with the Peoples network, had evaporated before the events in this case occurred. By 2002, the Deuces and Kings were bitter rivals fighting for control of Aurora’s streets. The rivalry resulted in frequent and escalating violence, often leading to attacks against victims mistaken for rival gang members. This inter-gang violence precipitated the investigation leading to this case. The Deuces comprised at least three tiers of membership: Seniors, Juniors, and Shorties. Shorties were the gang’s youngest full members. They bore responsibility for executing most of the gang’s activities, including carrying out violent acts and selling narcotics to fund the gang. They were often juveniles, recruited in their youth to build the gang’s ranks while minimizing criminal liability. By participating in gang activities, young members could become Juniors. Juniors were responsible for daily Deuce operations. They directed the Shorties’ activities and planned narrow-scale operations, determining which Shorties would participate, assigning responsibilities and distributing firearms to them, and supervising their activities. The Seniors were longstanding members of the gang whose age and accomplishments made them the leaders, broad-scope planners, and advisors for the gang. They directed the Juniors but were often removed from the Deuces’ daily activities. The local chapter in Aurora was organized at the Junior level, likely due to the ongoing incarceration of the local Seniors. Each tier in the Deuces had its own leadership structure where three members in each tier bore a title and increased responsibilities. The highest rank in each tier was the “First Seat” or “Governor” who was responsible for the overall activities of that group. The “Second Seat” or “Lieutenant Governor” assisted in directing activities within the tier, often assigning missions. The third-ranking member, or “Enforcer,” was responsible for ensuring compliance with the gang’s disciplinary codes and for administering “violations” (punishments) to members who broke the gang’s “leyas” (laws) or failed to participate in its activities. The Deuces’ daily activities included dealing narcotics to benefit the gang, going on “missions” (planned attacks on rival gang members), protecting and supporting fellow gang members and their families, punishing gang members who violated the gang’s codes, and meeting to coordinate these efforts. To facilitate their activities, the Deuces maintained a “caja” system, a loose form of community property and money acquired for the benefit of gang members. For example, if a Deuce was arrested and held pending trial, other members would take money from the caja to post bail for him. The caja also included a quantity of narcotics from which a member could borrow to make sales or trades for the benefit of the Nation. Although the Deuces were allowed to buy and sell drugs in a “free enterprise” system — sales could be conducted from sources outside of the Nation — all sales needed to benefit the Nation in some way, typically by paying a portion of the proceeds into the caja. Alternatively, profits from sales could be used to purchase firearms to donate to the gang. These “Nation guns” would then be used in carrying out missions or protecting members; the guns would be returned to the caja after the mission for later use (unless disposed of to avoid ballistics evidence in the event of a murder). Participation was compulsory for Juniors and Shorties, including those members who were rolled back from Senior to Junior status in an effort to better train, lead, and supervise the Deuces’ swelling Shorty ranks. Every Junior and Shorty had to pay periodic dues into the caja to support the gang’s activities. Missions were specifically assigned to gang members and usually consisted of planned attacks directed against specific rival gang members in retaliation for violence or territorial encroachment. Nation guns would be distributed for the mission along with orders to empty the magazine or bear violations for every bullet not fired. In addition to missions, all members were expected to comply with a standing order to attack Latin Kings whenever the opportunity arose. Participation in missions, other acts of violence, and donations of money, contraband, or firearms to the caja through drug sales or theft led to advancement within the organization. In exchange for these efforts, Deuces and their families received financial support and physical protection; these benefits extended to both Deuces on the streets and those who were incarcerated in various facilities throughout Illinois. B. The Investigation Following a drug raid on Junior Enforcer Orlando Rivera’s home in early 2002, Rivera absconded, and a warrant issued for his arrest. After two months, he was arrested pursuant to the warrant and was looking to avoid prosecution. He therefore began cooperating with local and federal authorities investigating the Deuces. He surreptitiously recorded gang meetings and conversations, he pretended to discard Nation guns only to hide them and later turn them over to authorities, he reported the gang’s activities and plans, and he received money to conduct controlled buys of narcotics and firearms from gang members. The intelligence gathered through Rivera produced evidence on four murders, eleven attempted murders, two solicitations to commit murder, other shootings, and narcotics distribution incidents — all of which the Deuces perpetrated in 2002 alone. In exchange for his services, Rivera received full immunity for his activities with the Deuces, and the government relocated and compensated him. Through Rivera’s cooperation and information arising from it, government authorities learned about each of the Defendants’ roles and general scope of participation in the Deuces’ criminal activities. We briefly relate some of that background information here. • Morales joined the gang in 1986 while in his early teens, and he carried out shootings and narcotics sales to move up in the Nation. He’d risen to a leadership position and helped craft the gang’s leyas, including the standing order to kill Latin Kings on sight. By 2002 he was a Senior, but was rolled back to Junior status during the Deuces’ reorganization. He became the Junior Enforcer, the third highest rank in the Aurora Deuces; he expressed support for the change, noting that it was “the best thing that happened.” • Barbosa was a long-standing Insane Deuce who immediately resumed his gang activities when released from prison in 2002. He personally approved of murdering his cousin, a rival gang member, and attempted to murder a Deuce the leadership erroneously believed to be cooperating with police. He encouraged the leadership to be more active in training the Shorties to be ready to carry out missions so the gang could gain control of the city. He also flatly stated his intention for the gang to kill rivals: “You see one, you do his ass. Plain and simple.” • Rodriguez was also a Senior before being rolled back to Junior status in 2002. Coming up in the Deuces, he attempted to kill rivals and distributed drugs for the gang’s benefit. He had assisted in developing the leyas and assigned missions to Shorties. After being rolled back, he advocated murdering members of the rival Ambrose street gang en masse and murdering whoever had been cooperating with the police from within the Deuces. • Hernandez joined the Deuces between the ages of 12 and 17, and he had risen to the Junior ranks by 2002. As a supervisor for the Shorties, he hosted meetings at his house, sent them out on missions, provided them with firearms for the missions, and handled violations if they failed to comply with his orders. He had been personally involved in at least three shootings that former gang members testified about at trial. He also was recorded requesting cocaine from the caja for himself and his cousin. • Lechuga was another older Deuce who had been rolled back to Junior status. The record shows conflicting accounts of his previous status: he asserts that he was retired and that he was pulled back in when Deuce leadership raised the retirement age to 35, but contrary evidence indicated that he was a Senior Deuce prior to the rollback. Regardless, he was upset with the manner of the rollback, but nevertheless did resume active participation in the gang’s affairs. He also noted that he’d remained willing to be involved while in retired or Senior status. He counseled Deuce leaders on how best to carry out inter-gang warfare within the Folks network without risking their ability to protect their members in prison. He also advocated changes in the caja system to increase efficiency in fronting drugs to gang members, perhaps as a result of his being unable to obtain from an enforcer the cocaine he’d requested for resale in July 2002. • Crowder held privileges in the Aurora Deuces, though he had not come up in the gang. He was a defector from the Gangster Disciples and had come to the Deuces in late 2001 to serve as a shooter on gang missions. He fulfilled — indeed, exceeded — that role, attempting to murder rival gang members and shooting individuals he erroneously believed to be Latin Kings. • Handley declared to an Aurora police officer that he’d been “a King killer and a Deuce all [his] life.” According to Rivera’s testimony, Handley was the Shorty Enforcer in Aurora at the time of the events in this case, assigning missions to Shorties and storing firearms for the missions at his home. He was involved in at least three of the gang’s attempted murders (either driving stolen cars during the missions or personally firing on rivals). At some point, however, Handley lagged in his participation, and the gang’s leadership ordered other members to beat Handley if they found him during the summer of 2002. The investigation likewise revealed information about the participation of other gang, members, including those whose activities are described in a companion case we decide today. See Benabe, 654 F.3d at 758. After months of fruitful investigation, the authorities decided to move forward with their attempt to dismantle the Insane Deuces in Aurora. C. Proceedings Below A federal grand jury indicted sixteen individuals alleged to be Deuces on charges of racketeering conspiracy; assault with a dangerous weapon, murder, attempted murder, and conspiracy to commit murder, all in aid of racketeering activities; illegal possession of firearms; and distribution and possession with intent to distribute narcotics (and conspiracy to do the same). The indictment described the Deuce organization as a racketeering enterprise and laid out some of the events comprising its pattern of racketeering activities. One of the indicted individuals pled guilty, and another remained at large. Nine of the remaining fourteen individuals moved for severance, requesting either individual trials or small group trials. Given the amount of briefing and argumentation, the challenges to juries inherent in lengthy and overtly complex trials, and the logistics of trying such a large group in a single courtroom, Judge Castillo severed the case into two trials. He grouped the Defendants together in the second of two near-simultaneous trials, describing them as “the less major players,” as compared to the “alleged leaders of the conspiracy” comprising the group in the first trial (the “Benabe trial”). Judge Leinenweber commenced the Defendants’ trial (after an earlier attempt that ended in a mistrial) in October 2008. On the government’s motion, the district court empaneled an anonymous jury. The Defendants had objected to the motion, and the district court heard arguments before ruling. It also had the luxury of considering Judge Castillo’s extensive reasoning and experiences in the Benabe trial before ruling. The district court determined that the circumstances warranted an anonymous jury and that any prejudice could be effectively mitigated through jury instructions and comments during voir dire. The venire members were told that their identifying information was withheld to protect them privacy and prevent any pai’ty or trial participant from contacting them. During the course of the three-month trial, the govexmment presented witness testimony, recordings of Deuce meetings, and forensic evidence establishing the gang’s activities, ideology, and purpose. One of the witnesses, Rivera (the Deuce turned confidential infoimant), had surreptitiously recorded gang meetings and conversations, performed controlled buys, and turned over evidence that other Deuces had expected him to destroy or dispose of on the gang’s behalf. He testified over four days as to the inner workings of the gang, explaining what was said in meetings, identifying participants and speakers, and describing the gang’s organization and enforcement means. Two other former Deuces, Lorenzo Becerra and Akeem Horton, also testified regarding the gang’s activities and individuals’ scope of participation. The government also presented testimony from twenty-four police officers, eleven federal agents, and seven victims. The jury returned its initial verdicts a week after the Defendants concluded then-closing arguments, finding each of the Defendants guilty of racketeering conspiracy. It also found Barbosa guilty of conspiracy to commit murder, and it found Barbosa, Morales, Rodriguez, Hernandez, and Lechuga guilty of distribution of narcotics. By contrast, the jury acquitted Morales of conspiracy to commit murder and was unable to reach a verdict on the charges against Steven Perez; the district court declared a mistrial as to Perez. On the court’s instructions, the jury then returned special verdicts as to each defendant it had found to be guilty; the special verdicts were highly individualized, differentiating between the defendants on shared counts. Before the special verdict phase, however, the district court received a note from a juror. The juror warned that some of the jurors had been discussing the case amongst themselves prior to deliberations, in contravention of the court’s standing instructions. The Defendants moved for a mistrial, but the district court delayed ruling on the motion until written motions and arguments were submitted. It ultimately determined that a hearing was unnecessary, denied the mistrial motion, and proceeded to further jury deliberations. Several of the Defendants filed post-trial motions in March 2009 challenging various aspects of their convictions, including the sufficiency of the evidence presented at trial, and requesting judgments of acquittal. All of the Defendants filed post-trial motions requesting new trials on various grounds. In a consolidated ruling, the district court denied all of the post-trial motions in May 2009. The Defendants then proceeded to sentencing. Morales, Hernandez, and Rodriguez were imprisoned for life. Barbosa was sentenced to 480 months' imprisonment. Handley and Lechuga were sentenced to 240 months’ imprisonment. Finally, Crowder was sentenced to 177 months’ imprisonment. Each of the Defendants timely appealed. II. Analysis The Defendants appeal several aspects of their convictions and sentences. Taking full advantage of our leniency in allowing both joint and individual briefs, the Defendants filed a virtual cannonade of briefing, exceeding 280 pages in total. Their opening salvo is a joint brief presenting three issues. The Defendants collectively argue that the alleged errors apply to them all and that each error requires reversal of their convictions. Each of the Defendants (other than Barbosa) then fires an independent volley attacking his own conviction, sentence, or both. We will address the jointly presented issues first, then we will consider each appellant in sequence. A. Issues Presented by all Defendants The Defendants jointly contend the district court committed three errors that require us to reverse their convictions. First, the district court maintained juror anonymity during empanelment and throughout the trial. The Defendants argue both that the anonymous venire undermined the effectiveness of their voir dire and peremptory challenges and also that the anonymity caused the seated jury to be prejudiced against them. Second, the district court denied severance motions made by several of the Defendants. They argue that being required to stand trial together led to evidentiary spillover and undue lengthening of the trial, both of which prejudiced the jury against individual defendants. Third, the district court did not conduct a hearing to explore the extent of the jury’s alleged premature deliberations. The Defendants argue that the jurors’ discussions deprived them of a fair trial. We will analyze each of these issues in turn. 1. Anonymous Jury Empanelment An “anonymous jury” is selected from a venire whose members’ identifying information — such as names, occupations, addresses, exact places of employment, and other such facts — has been withheld from the parties in order to protect potential jurors and their families. United States v. Crockett, 979 F.2d 1204, 1215 n. 10 (7th Cir.1992). Empaneling an anonymous jury “raises the specter that the defendant is a dangerous person from whom the jurors must be protected” and potentially deprives defendants of information that could be used in making juror selections during voir dire. United States v. Mansoori, 304 F.3d 635, 650 (7th Cir. 2002) (quoting United States v. Ross, 33 F.3d 1507, 1519 (11th Cir.1994)). Accordingly, the use of anonymous juries is discouraged, and courts should be “highly circumspect in ordering the empanelment of anonymous juries.” Ross, 33 F.3d at 1522. But we have never held that it is presumptively inappropriate to rely on anonymity. If the district court concludes both that the circumstances strongly suggest that the jury needs protection and also that reasonable precautions — such as preliminary comments to the venire and instructions before deliberations — can mitigate any prejudice to the defendants, an anonymous jury may be empaneled. Crockett, 979 F.2d at 1216-17. Over the Defendants’ objection, the district court granted the government’s motion for an anonymous jury in this case, withholding the names, addresses, and other identifying information of the members of the venire. The district court did not enter its findings on the record, but instead ordered an anonymous jury from the bench: “I believe that is correct. So I am going to order an anonymous jury over the objection of all the defendants.” (Tr. 3/18/08 at 11.) The Defendants argue on appeal that the lack of explicit findings renders the empaneling inherently erroneous for two reasons. First, they contend that an on-the-record determination of why an anonymous jury is necessary is fundamental to protecting their rights to a fair trial and an impartial jury. Second, they contend that any reliance on Judge Castillo’s reasoning in the Benabe trial was inappropriate in this case because his reasons were not mentioned by the district court and are not part of the record on appeal. We review the decision to use an anonymous jury only for an abuse of discretion, Mansoori, 304 F.3d at 650, remaining particularly deferential to the district court’s substantial discretion in this area, Crockett, 979 F.2d at 1215-16. We agree that the district court should have stated its reasons for granting the government’s motion on the record. See Mansoori, 304 F.3d at 651. “Some statement of the district court’s reasoning is necessary for this court to be able to meaningfully review its decision,” United States v. Marion, 590 F.3d 475, 477 (7th Cir.2009), and the ruling we quoted from the district court’s status hearing did not contain much to indicate its reasons. Ac cordingly, the district court erred by granting the government’s motion for an anonymous jury without stating its reasons for doing so on the record. That error is not dispositive of the matter, however, as it may be harmless in light of the complete record of this case. See Fed.R.Crim.P. 52(a); Mansoori, 304 F.3d at 651-52. Indeed, the context of the district court’s statement quoted above does provide us with the information necessary to review the propriety of empaneling an anonymous jury. (Tr. 3/18/08 at 7-11.) The district court was holding a status hearing in which it considered several pre-trial motions in the Defendants’ case, including the government’s motion to empanel an anonymous jury. The government and defense counsel were arguing the merits of the jury’s anonymity, referring to written materials they submitted to the district court and to the proceedings and reasoning in the Benabe trial. The district court indicated its familiarity with the materials and arguments already submitted in the case, as well as the arguments before and decision by Judge Castillo in Benabe. The district court then correctly noted that “[t]he issue obviously is there is a certain prejudicial implication which has been raised and it was rejected by Judge Castillo on the basis that he believed that appropriate jury instructions could solve that.” (Tr. 3/18/08 at 8.) In response to defense counsel’s arguments regarding the inherent prejudice of an anonymous jury, the government indicated that in the Benabe trial the jury did not manifest prejudicial fear of the defendants because “they thought that their names were being withheld to keep them safe from the press as opposed to anybody else.” (Tr. 3/18/08 at 10.) The government also noted that its multiple filings regarding the danger posed by the Deuces showed that concerns of danger applied to the Defendants just as readily as their accused co-conspirators in the Benabe trial. The government concluded that any potential prejudice could be mitigated by the court’s handling of the jury panel. It was to these arguments, and to the latter statements in particular, that the district court responded, “I believe that is correct,” before ordering an anonymous jury. (Tr. 3/18/08 at 11.) Reviewing these arguments, the government’s written motion, and the entire record of the case, we conclude that the district court’s error in failing to articulate its reasons for empaneling an anonymous jury was harmless. See Mansoori, 304 F.3d at 651-52 (finding an erroneous anonymous jury empanelment harmless in a close case). The Defendants argue that the anonymity was harmful because it was not justified by the records and because it deprived them of a fair trial in two ways: it affected their ability to use peremptory challenges and it created an undue presumption that the defendants were dangerous. We find the Defendants’ arguments unconvincing in light of the circumstances of their case. We acknowledge that jury anonymity is warranted only where strong reasons lead the court to believe that the jury needs protection. See Crockett, 979 F.2d at 1215-16. In its motion, the materials referenced therein, and its arguments at the status hearing, the government fully articulated its view of the reasons below, discussing in detail the multiple factors that the Defendants advocate on appeal. See Mansoori, 304 F.3d at 650-51. As the indictments indicated, there was good reason to believe that the Defendants were involved in organized, violent crime. Several of the Aurora Deuces remained unindicted, and at least one indictee was still a fugitive at the time of trial, raising the clear inference that the gang remained capable of carrying on its violent activities despite the incarceration of the Defendants. The government indicated that it would present evidence, as it did in the Benabe trial, that the Deuces conspired and attempted to kill two of the gang’s members the Deuces suspected of cooperating with law enforcement. The government also planned to present evidence that the Deuces attempted to murder Rivera in 2003 in retaliation for his known cooperation. Clearly there were substantiated allegations that the Defendants had acted with violence to interfere with the judicial process; it is not unreasonable, given the gang’s history and capacities, to infer that it might use violence again to influence or harass the jury. The Defendants uniformly faced profound sentences, each having the potential of a life sentence, except for Handley’s twenty-year potential. While long sentences are not uncommon in narcotics cases, life sentences clearly are among the most severe that can be imposed. The arrests of the Aurora Deuces — and the trial of seven of them in the Benabe trial — garnered news media and internet coverage. The district court had these arguments and supporting case law before it when considering the government’s motion. On appeal, the Defendants do not convincingly undermine the validity of these points. They argue that the government did not mention any situations where any Deuce had attempted to interfere with jurors and that there was no evidence that juror intimidation was likely. They do not, however, address the allegation that Deuces attempted to murder Rivera after several of them were arrested and learned that he was a cooperating witness. Nor do they acknowledge the resulting, reasonable inference that jurors may be as susceptible as cooperators and witnesses to their gang’s violent acts. See United States v. Edmond, 52 F.3d 1080, 1092 (D.C.Cir. 1995) (per curiam) (“[A] general willingness to obstruct justice on the part of a defendant or his associates! ] is more than adequate to suggest a real possibility that a defendant will threaten or otherwise tamper with jurors.”). While the ability and incentive to threaten jurors alone are not enough to warrant an anonymous jury, Mansoori, 304 F.3d at 651, we conclude that an evidenced history of interference with the administration of justice — added to the gang’s ability and incentive — would bring a case within the district court’s discretion to empanel an anonymous jury, see Crockett, 979 F.2d at 1216; Edmond, 52 F.3d at 1091-92; United States v. Vario, 943 F.2d 236, 241 (2d Cir.1991). The anonymous jury jurisprudence of this and other circuits does not require proof of impending harm to jurors; rather, it only requires reason to believe that jury protection is necessary. See, e.g., Mansoori, 304 F.3d at 651 (evidence of accused’s unusually profound pattern of violence could cause jurors to fear for their safety); United States v. DiDomenico, 78 F.3d 294, 301-02 (7th Cir.1996) (protection through anonymity appropriate to prevent tampering by organized criminals with history of bribery); Edmond, 52 F.3d at 1091 (history of jury tampering sufficient, but not necessary, to ascertain a threat to jurors from the charges in the indictment). We also note that the district court did not stop with a decision to empanel an anonymous jury; rather, it took efforts to mitigate any potential prejudice arising from the jurors’ anonymity. Unlike in Mansoori, see 304 F.3d at 652, here the district court instructed the jury venire that their names were being withheld to prevent out-of-court contact, not out of concern for juror safety: “[T]he reason why we are selecting the jury in this particular way without disclosing names or specific addresses is because we have an ironclad rule that no participant can contact a juror for any reason and the jurors are not to contact any participant. So in order to make sure that no one contacts you we are not having your name disclosed.” (Tr. 10/06/08 at 218.) In its pretrial hearing, the district court agreed with the government that this kind of admonition helped limit any prejudicial impact of anonymity. See Crockett, 979 F.2d at 1216-17. The district court also engaged in a thorough voir dire of the venire, a procedure we have held may adequately protect the fundamental right to an impartial jury. Id. at 1217; Mansoori, 304 F.3d at 652. The district court questioned the panel members at length, and the members did provide general information about their professions and residence neighborhoods. At the beginning and close of trial, the district court also reminded the empaneled jurors of the presumption of innocence, in part to dispel any prejudice that anonymity might have engendered. See Mansoori, 304 F.3d at 652. The Defendants argue that the juror anonymity prevented their effective use of peremptory challenges, but they neither argue that the voir dire or instructions were inadequate nor address the effectiveness of the district court’s mitigating efforts. In summary, we find that the district court erred by ordering the use of an anonymous jury without articulating its reasons for doing so. Nevertheless, Judge Leinenweber’s decision was well informed by argumentation and briefing, as well as by Judge Castillo’s thorough treatment of the issue in — and his experiences during— the antecedent Benabe trial. The record demonstrates that the circumstances of this case fulfill each of the factors we have identified as informing the anonymity decision. Further, the district court undertook several measures to mitigate any potential prejudice arising from the jurors’ anonymity. As “we can discern no conCrete way in which the anonymous jury deprived the [Defendants of a fair trial,” id., we conclude that the district court’s error was harmless. 2. Denial of Severance Motions The district court initially severed the Defendants from their alleged co-conspirators in the Benabe trial, reducing the number of defendants in each case from fourteen to seven. This initial severance was a Solomonic compromise between the numerous and mutually exclusive severance proposals advanced by various defendants. Naturally, it was neither a perfect solution nor a pleasing one for the Defendants who sought individual trials. After a mistrial was declared as to Crowder due to a hung jury in the Benabe trial, the district court granted the government’s motion to add him as a defendant in this case. Following this addition, and again after some defense counsels’ opening statements, and then throughout trial, several of the Defendants renewed their severance motions; the district court denied each motion in turn. On appeal, Handley, Lechuga, Rodriguez, and Crowder argue that the erroneous denials of their individual severance motions unduly prejudiced them. They request new trials where they would each be tried individually. The Defendants do not challenge the propriety of the initial joinder of each Aurora Deuce indicted in 2006 in a single trial. See Fed.R.Crim.P. 8(b). Rather, they contend that the joinder of so many defendants — fourteen at the outset and even eight in their resulting trial — was inherently prejudicial and warranted discretionary relief from the district court. The Federal Rules of Criminal Procedure allow a district court to separate defendants or counts in light of possible prejudice: “If the joinder of offenses or defendants in an indictment, an information, or a consolidation for trial appears to prejudice a defendant or the government, the court may order separate trials of counts, sever the defendants’ trials, or provide any other relief that justice requires.” Fed. R.Crim.P. 14(a). As the permissive language of the rule indicates, the decision whether to grant a defendant’s severance motion under Rule 14 is discretionary, and we review the district court’s denial of such a motion for an abuse of discretion. United States v. Alviar, 573 F.3d 526, 539 (7th Cir.2009). We will reverse the conviction on severance grounds only if actual prejudice resulted from the denial of the severance motion, United States v. Lopez, 6 F.3d 1281, 1285 (7th Cir.1993), and such prejudice requires that the defendant have been deprived of a fair trial, not merely a better chance at acquittal than an individual trial may have afforded, United States v. McAnderson, 914 F.2d 934, 948 (7th Cir. 1990). When alleged co-conspirators are indicted together, as they were here, there is a strong preference that they be tried together. Zafiro v. United States, 506 U.S. 534, 537-38, 113 S.Ct. 933,122 L.Ed.2d 317 (1993); Alviar, 573 F.3d at 539. Considerations of judicial economy, consistency of verdicts, and systematic efficiency inform this preference, Zafiro, 506 U.S. at 539, 113 S.Ct. 933, as “[o]ur system of criminal justice would crumple beneath the weight of individual trials if every defendant who demanded severance was provided one,” McAnderson, 914 F.2d at 949. But we acknowledge that defendants may face some prejudicial spillover in large complex cases, and we therefore expect district courts to balance the risk to defendants who move for severance against the benefits gained by joint trials. Id. District courts should continue to evaluate the risk of undue prejudice resulting from joint trials throughout the proceedings. United States v. Harris, 761 F.2d 394, 400 (7th Cir.1985). In the end, however, “Rule 14 does not require severance even if prejudice is shown; rather, it leaves the tailoring of the relief to be granted, if any, to the district court’s sound discretion.” Zafiro, 506 U.S. at 538-39, 113 S.Ct. 933. Alternative mitigative measures, including limiting instructions, may suffice to offset prejudice where severance would be too drastic of a remedy. Id. at 539, 113 S.Ct. 933. On appeal, four of the Defendants allege undue prejudice, in that the joint trial raised the likelihood that they were convicted based on mere association with their co-defendants rather than on their own culpability. In general, the four argue that the logic behind Judge Castillo’s initial severance of the fourteen-defendant trial into two separate trials applied with equal force to their resulting eight-defendant grouping. The trial lasted longer than two months, the indictment covered a period of three years, and the acts of exceedingly violent men were discussed at length. Specifically, the four contend that the district court abused its discretion in denying their severance motions despite the risk of evidence pertaining only to other defendants spilling over to them in the eyes of the jury — a risk that is heightened in large multi-defendant trials. See Zafiro, 506 U.S. at 539, 113 S.Ct. 933. We will consider each of their arguments independently, mindful that “a district court should grant a severance under Rule 14 only if there is a serious risk that a joint trial would compromise a specific trial right of one of the defendants, or prevent the jury from making a reliable judgment about guilt or innocence.” Id. Our touchstone is' the jury’s capacity to follow its instructions and consider the evidence as to each defendant independently. United States v. Hedman, 630 F.2d 1184, 1200 (7th Cir.1980). Handley sought an individual trial (or a narrow grouping with Lechuga and Rodriguez) because he (and the other two) were not accused of any murder or attempted murder. He argues that the district court acknowledged the possibility of prejudice but denied his severance motion in deference to judicial economy — despite the fact that no economy was actually realized through the joint trial. He further contends, without reference to any precedent, that it is obviously prejudicial to subject a defendant to a trial almost two months longer than an independent trial would have been. He touts his minimal involvement in the gang, as signified by a general lack of testimony about him particularly, the lack of drug charges against him, and the standing punishment order against him for his poor participation in the gang’s activities. He concludes that he was highly prejudiced by spillover evidence, given the great disparity of the evidence against him compared to that against his co-defendants. We have repeatedly rejected similar arguments, noting that where the evidence against the appellant sufficed to convict him of the charges against him specifically, jury instructions and other mitigative measures sufficed to limit prejudicial spillover and rendered the severance denial proper. See, e.g., Alviar, 573 F.3d at 539 (“[T]he fact that the government has greater evidence against one co-defendant does not automatically give the other defendant grounds for severance.”); United States v. Serpico, 320 F.3d 691, 696 (7th Cir.2003); McAnderson, 914 F.2d at 949; Redman, 630 F.2d at 1200. Here, Handley stood charged only with racketeering conspiracy, and he denied only his participation in the conspiracy while conceding his knowledge of its existence. Recordings and testimony at trial implicated him directly in the conspiracy; although this evidence might have been susceptible to multiple interpretations, it was sufficient to sustain Handley’s conviction. See United States v. Caliendo, 910 F.2d 429, 438 (7th Cir.1990); United States v. Hall, 212 F.3d 1016, 1023 (7th Cir.2000) (“Merely ceasing participation in the conspiracy, even for extended periods, is not enough.”) 0quoting United States v. Bafia, 949 F.2d 1465, 1477 (7th Cir.1991)). Further, his claim that he would not have been convicted in the absence of evidence admitted in the trial that would have been irrelevant and inadmissible in a separate trial is undermined by the jury’s inability to reach a verdict regarding his co-defendant Perez. Like Handley, Perez was not charged with the murder and narcotics conspiracies, and the jury demonstrated its capacity to act on its limiting evidentiary and deliberation instructions by giving individual consideration to the defendants and the evidence against them. See Alviar, 573 F.3d at 539 (“There was no actual prejudice to [the defendant] on account of ‘spillover’ evidence because the jury distinguished between him and his co-defendants.”). The bottom line is that Handley’s simple disparity-in-the-evidence argument does not render the district court’s disposition of his severance motion erroneous. The evidence presented against him was sufficient to support his conviction, the more robust evidence against his codefendants notwithstanding. And the insufficiency of his disparity argument applies to both the quantity of the evidence against his co-defendants as well as its nature (describing violent acts). The jury was instructed to consider each defendant’s case independently, and we assume that it heeded those instructions even though Handley argues that he was the smallest player in this drama. See McAnderson, 914 F.2d at 949. Indeed, the record strongly suggests that the jury did follow its instructions, given the individualized verdicts it returned. Lechuga also argues that the denial of his severance motion exposed him to undue prejudice. Like Handley, he argues that he was not indicted on any acts of violence and that the evidence of his co-defendants’ acts of murder and assault tainted his case. He also contends, however, that he had withdrawn from the conspiracy in the mid-1990s. He therefore argues that, by refusing to sever him from the Defendants’ trial, the district court prejudiced the jury against him in three ways. First, the evidence of his co-defendants’ violent acts could not have been introduced against him in a separate trial, where he would have stipulated to many of the facts the government sought to prove in this trial. Second, he claims that there would have been no reason to empanel an anonymous jury in a separate case because he was not accused of violent acts. Third, a post-arrest statement from his co-defendant Morales, stating that Morales and other Senior Deuces were rolled back to Junior status, was inconsistent with his theory of defense and would not have been admissible in a separate trial against him. Lechuga’s defense — unique among the Defendants — does not change the outcome of our severance analysis. The acts of violence forming the backbone of the gang’s racketeering activities would have been admissible against Lechuga in a separate trial, even if he were not accused of perpetrating those acts himself. Further, he cites no authority suggesting that his withdrawal defense would preclude the introduction of this competent evidence, see Zafiro, 506 U.S. at 540, 113 S.Ct. 933 (“[A] fair trial does not include the right to exclude relevant and competent evidence.”), and even his alleged willingness to stipulate would not necessarily bar the government from proving its case as it chose, see United States v. Phillippi, 442 F.3d 1061, 1064 (7th Cir.2006); United States v. Conner, 583 F.3d 1011, 1022 (7th Cir.2009). Further, Lechuga’s withdrawal defense was quite uncertain; his recorded statements, far from showing an affirmative act to withdraw, indicated that he was willing to participate in the Deuces’ activities during his purported withdrawal: “I’ve told some of you in the room before that, hey, if there’s somethin’ I can do ... I’ll help you out.” While he claims to have been inactive for some time, “[h]is not taking actions in furtherance of the conspiracy is not the same as taking affirmative action to withdraw.” United States v. Wren, 363 F.3d 654, 665 (7th Cir.2004), vacated on Booker grounds sub nom. Yarbor v. United States, 543 U.S. 1101, 125 S.Ct. 1021, 160 L.Ed.2d 1005 (2005); see also Hall, 212 F.3d at 1024. Accordingly, his planned withdrawal defense did not distinguish his case from Handley’s, so Lechuga cannot show prejudice requiring severance as a result of the evidence of his co-conspirators’ violence. See United States v. Handlin, 366 F.3d 584, 591 (7th Cir.2004). Lechuga’s second argument is likewise without merit. The violent tendencies of the individual on trial are not the sole determinant for jury anonymity. See Mansoori, 304 F.3d at 650-51. While Lechuga himself was not accused of any discrete acts of violence in the indictment, he tacitly admits that the organization engaged in violence, referring to “numerous senseless acts of violence committed by other younger, active members of the Insane Deuces” and describing his co-defendants as “those who terrorized the streets of Aurora.” As discussed above, some of the Aurora Deuces remained on the street, and one of the gang’s purposes was to protect its members after they are arrested and if they are incarcerated. The sufficient reasons for empaneling an anonymous jury in this case may have applied, even if not as strongly, in an independent trial for Lechuga or for any smaller subset of the Defendants. The admission of Morales’s statement— that he and other Senior Deuces had been rolled back to Junior status — also did not require severance at Lechuga’s request. To succeed on this point, Lechuga must “rebut the dual presumptions that a jury will (1) capably sort through the evidence and (2) follow instructions from the court to consider each defendant separately.” Lopez, 6 F.3d at 1286. His only argument was that the statement was inconsistent with his withdrawal defense and was therefore prejudicial. But neither Morales’s statement nor the testimony introducing it mentioned Lechuga by name, and recordings of Lechuga’s own words demonstrated his reinvolvement in the gang’s activities in 2002. Even if Lechuga’s withdrawal defense had some merit, any minimal prejudice arising from the introduction of Morales’s statement was adequately tempered by the district court’s repeated instruction that Morales’s statement was to be considered only against Morales. See United States v. Marshall, 75 F.3d 1097, 1105 (7th Cir.1996). We next consider Rodriguez’s allegation of substantial prejudice from the severance denial. He proposed three separate trial groupings of the indictees before ultimately seeking an independent bench trial of his own. Like Handley and Lechuga, he argues that the evidence against his co-defendants prejudiced him and that an anonymous jury would not have been necessary in his various proposed separate trials. As these arguments are indistinguishable from his co-defendants’, we reject them based on our earlier analysis. Rodriguez’s only new argument is that, five times during the trial, witnesses and government counsel referred to him (Miguel Rodriguez) as Miguel Martinez. Rodriguez argues that these misstatements prejudiced the jury against him because the jurors could have confused his culpability with Martinez’s greater culpability. We can see neither how an individual trial would have prevented such unintentional misstatements nor how the jury could have confused Rodriguez with a fugitive who was not before the court. Of the five misstatements, one was immediately corrected by the witness, another was promptly corrected on counsel’s objection, and the remaining three were not brought to the court’s attention. There is no indication that these understandable and fleeting misnomers were intentional or related to the presence of the several Defendants in the trial. Further, Martinez was charged with nine murders or attempted murders while Rodriguez was charged with no violent acts, and Martinez was not tried in the same proceeding. Despite Rodriguez’s assertion to the contrary, the jury did not have an “impossible” or “insurmountable” task in distinguishing Rodriguez from Martinez during its deliberations. Accordingly, the district court did not abuse its discretion by denying Rodriguez’s severance motions. Crowder is the last appellant alleging actual prejudice from the district court’s denial of his severance motion. He was grouped with the Defendants after the jury was unable to reach a verdict on the charges against him in the preceding Benabe trial. He argues that he was not an official Insane Deuce, while “each defendant in the second trial was an Insane Deuce gang member whose responsibility was to engage in all gang activities including selling drugs and shooting rival gang members.” (Appellants’ Joint Br. at 56.) According to Crowder, his inclusion with the Defendants subjected the jury to evidence of his myriad violent acts that were extraneous to the Insane Deuces’ activities and to evidence of his codefendants’ acts that did not involve him. This evidentiary overlap, he concludes, left the jury unable to distinguish between his non-gang-related violent acts and his participation in the racketeering conspiracy. As his counsel summarized during oral argument, Crowder was “one of the most violent individuals in this case.... [T]he absence of violence of his co-defendants made him appear much more heinous and increased the likelihood that the jury would take that into consideration and not appropriately examine his intent or agreement to join the conspiracy.” Crowder’s argument, despite being creative, is unconvincing. While he might not have formally been an Insane Deuce, the relevant question was whether he had joined the racketeering conspiracy. The government’s theory was that the shootings Crowder committed were intended to benefit the Deuces, were part of the racketeering conspiracy, and were examples of Crowder merely mistaking the victims for Latin Kings. Crowder’s argument that these were unrelated acts calls into question the weight of the evidence presented by the government, not its admissibility or prejudicial impact. It was the province of the jury to determine whether the evidence against Crowder indicated that he was acting as a member of the racketeering conspiracy. Further, others among the Defendants stood accused of violence in their own actions, including a plot to kill an Aurora Deuce suspected of cooperating with authorities and two murders in retaliation against Latin Kings. Even if we were inclined to hold that a violent actor could be prejudiced by juxtaposition with his (comparatively) benign co-conspirators in a joint trial, Crowder’s circumstances simply do not fit that mold. Accordingly, the district court did not abuse its discretion in denying Crowder’s motion to sever. 3. Juror Misconduct Allegations After the jury returned its verdicts on the Defendants (and reported that it could not reach a verdict as to Perez), the district court received a note from one of the jurors. Juror 107 alleged that, for about three weeks during the government’s casein-chief, some of the jurors had violated the district court’s instruction not to discuss the case among themselves prior to deliberations: “I spoke to some of the loud & boisterous jurors ... about making remarks about witnesses, attorney’s [sic] and discussing the ease.... Jokes and other inferences about the ease were made.” The Defendants moved for an immediate mistrial, and the district court denied the motion, recommending that defense counsel file a written motion and that the district court would then decide whether a hearing would be necessary. After considering written submissions from the government and defense counsel, and the Defendants’ accompanying motion for a new trial based on jury misconduct, the district court declined to hold a hearing and denied the Defendants’ motion. The district court reasoned that there was no allegation of external influence on the jury and that no circumstance in the case justified departure from the general rule that post-verdict interrogation of jurors is inappropriate when only internal pressures or premature deliberations are involved. The Defendants appeal the district court’s decision to not hold a hearing, claiming that the jury’s alleged premature discussions denied them a fair trial. While the Defendants acknowledge that the district court could not inquire into the jury’s or a juror’s thought process, see Fed. R.Evid. 606(b), they argue that a hearing to discover whether premature deliberations occurred and, if so, the extent of those deliberations would not invade any territory protected by Rule 606. See United States v. Resko, 3 F.3d 684, 692 (3d Cir.1993) (“Although the district court expressed concern that if it engaged in any colloquy with the jurors it might invade their deliberative process, the court could have easily tailored a colloquy to elicit information about the jurors’ impartiality without so intruding.”). We review a district court’s handling of allegations of premature deliberations and juror bias for an abuse of discretion. United States v. Moore, 641 F.3d 812, 830 (7th Cir.2011); United States v. Vasquez-Ruiz, 502 F.3d 700, 704 (7th Cir.2007). The Defendants compare their case to two federal cases, suggesting they compel a reversal here. First, they liken their case to Oswald v. Bertrand, a habeas corpus case in which we held that the trial court’s failure to investigate pervasive allegations of bias among venire-persons was erroneous. 374 F.3d 475, 483-84 (7th Cir. 2004). The Defendants argue that the trial court’s refusal to inquire further of one panel member and others in Oswald parallels the district court’s refusal to question jurors in their case. Second, they liken their case to Resko, in which the Third Circuit reversed the appellant’s conviction because the district court did not investigate the nature and extent of jurors’ preverdict discussions of the case. Resko, 3 F.3d at 686. There, a juror alerted a court officer on the seventh day of a nine-day trial that members of the jury had been discussing the case. Id. The reviewing court noted six (somewhat overlapping) policies behind prohibiting premature deliberations in criminal cases, including both potential bias toward the prosecution due to the presentation of the government’s case first and also the risk of jurors crediting one side’s evidence over the other due to the natural human tendency to remain committed to a view already expressed to others. Id. at 689-90. The district court addressed the jury as a whole and distributed questionnaires meant to gauge whether each juror had discussed the case and whether each had formed an opinion as to guilt or innocence of either defendant. Id. at 688. The Third Circuit ultimately determined that, “given the discovery that the jurors had all engaged in premature discussions of the case, ... this method was inadequate to enable the [district] court to fulfill its responsibility of providing an appropriate cautionary instruction and of determining whether prejudice resulted from the jury misconduct.” Id. at 691. The Defendants argue that the district court’s failure to voir dire jurors in this case after receiving the note compels a new trial, just as circumstances in Oswald and Resko did. We disagree. The Defendants overlook two fundamental distinctions between their case and those they cite. First, the misconduct allegation here arose after the general verdicts had been returned, whereas every case they cited dealt with pre-verdict allegations and concomitant opportunities to investigate the circumstances before the juries retired to deliberate. Second, the majority of their cited cases involved far clearer indications of juror bias than were present here, where any potential bias is a matter of sheer speculation. As a preliminary matter, the government and the Defendants agree that Juror 107’s note neither alleged nor implied any external influence on the jury’s deliberations. Accordingly, no presumption of prejudice arises, see Remmer v. United States, 347 U.S. 227, 229, 74 S.Ct. 450, 98 L.Ed. 654 (1954) (external communication, contact, or tampering with the jury rebutably presumed to be prejudicial to the defendants), and no hearing or investigation in which jurors are questioned was absolutely required, see United States v. McClinton, 135 F.3d 1178, 1186 (7th Cir.1998) (“If outside contacts may have affected the jury, due process requires some form of hearing.”). We evaluate this case as a matter of alleged intra-jury influence or misconduct. See Vasquez-Ruiz, 502 F.3d at 705 (distinguishing the scope of inquiry necessary based on lack of clarity whether source of message was external or internal to the jury); United States v. Kimberlin, 805 F.2d 210, 243-44 (7th Cir. 1986) (no per se abuse of discretion found in district court’s decision not to conduct a hearing where only intra-jury communications were alleged). The allegations of intra-jury misconduct in this case arose only after the jury returned its general verdicts as to each of the Defendants. The Supreme Court has warned against pervasive post-verdict inquiries into juror misconduct: There is little doubt that postverdict investigation into juror misconduct would in some instances lead to the invalidation of verdicts reached after irresponsible or improper juror behavior. It is not at all clear, however, that the jury system could survive such efforts to perfect it.... [F]ull and frank discussion in the jury room, jurors’ willingness to return an unpopular verdict, and the community’s trust in a system that relies on the decisions of laypeople would all be undermined by a barrage of postverdict scrutiny of juror conduct. Tanner v. United States, 483 U.S. 107, 120-21, 107 S.Ct. 2739, 97 L.Ed.2d 90 (1987). Federal Rule of Evidence 606(b) flatly prevents many inquiries of jurors intended to impeach the jury’s verdict. The Defendants acknowledge that none of the three areas of permissible testimony specified by the rule apply in their case, but they cite Kimberlin for the proposition that the district court nevertheless could have asked jurors about the alleged premature deliberations without violating Rule 606(b). See Kimberlin, 805 F.2d at 243-14 (“We recognize that these communications between jurors were allegedly made during the course of trial. Hence, they are not literally included in the prohibition of Rule 606(b) against testimony by a juror as to a statement during the course of the jury’s deliberations.”). The Defendants may be correct that the district court could have inquired about whether the discussions occurred and whether they constituted premature deliberations. But they ignore the remainder of Rule 606(b) and our holding in Kimberlin when they state that “the issue was whether there were premature discussions of the evidence and then whether those discussions prejudiced the defendants.” (Appellants’ Joint Br. at 64 (emphasis added).) Any inquiry as to bias arising from the alleged premature deliberations would run afoul of the Rule’s clear proscription: after the verdict is entered, “a juror may not testify as to ... the effect of anything upon that or any other juror’s mind or emotions as influencing the juror to assent to or dissent from the verdict ... or concerning the juror’s mental processes in connection therewith.” Fed.R.Evid. 606(b). As we held in Kimberlin, any “hearing would be fruitless unless these statements, if made, would be presumed to be prejudicial.” 805 F.2d at 244. For reasons we discuss below, we decline to presume prejudice under the circumstances of this case. Our determination comports with the holdings'of this and other courts to have considered district courts’ obligations when allegations of jury misconduct arise after the verdicts have been entered. In United States v. Stafford we determined that a post-verdict motion for a hearing regarding jury bias was too late, “as the defendants would have been seeking testimony from a juror designed to impeach the jury’s verdict without any basis for supposing that the jury had been