Full opinion text
Opinion PER CURIAM. Separate opinion dissenting in part filed by Circuit Judge STEPHEN F. WILLIAMS. PER CURIAM: This is a consolidated appeal arising from the second and third trials of members of a narcotics conspiracy led by Rayful Edmond III. See United States v. Edmond, 52 F.3d 1080 (D.C.Cir.1995). We remand the conspiracy conviction of Robert Hardy and the murder and weapons convictions of Columbus Daniels for further proceedings, and we affirm all remaining convictions. In addition, we remand the cases of all appellants except Ronald Morgan for resentencing. I. INTRODUCTION Appellants all stand convicted of participating in or conducting business with the Edmond narcotics conspiracy, an organized enterprise that sold massive quantities of cocaine in the District of Columbia in the late 1980s. The activities of the organization are detailed in Edmond, 52 F.3d at 1084-86. Twenty-nine people were originally indicted on a number of counts of conspiracy, nareot-ics-related activities, weapons offenses, and murder and other crimes of violence. In August 1989, the district court severed the counts of the indictment alleging weapons offenses and crimes of violence from those alleging conspiracy and drug-related activity. The court further divided the defendants indicted for drug crimes and conspiracy into two groups according to their roles in the enterprise, designating the leaders and principal members of the organization as Group I and the more peripheral actors — including all of the appellants here — as Group II. Three trials were held. The Group I defendants were tried in late 1989; their appeals are the subject of Edmond. The appellants in this case were tried in the Group II proceedings beginning in February 1990. The government’s evidence, which is described in detail in our analysis of the defendants’ attacks on its sufficiency, indicated that Willie Childress made at least one fifty-kilogram delivery of cocaine to the Edmond organization; that Columbus Daniels was a local courier for the group; that Raehelle Edmond stored large quantities of the group’s drugs and cash; that Robert Hardy ran errands for the enterprise; that Ronald Morgan purchased a kilogram of cocaine from the group; that Constance Perry counted and deposited narcotics proceeds; that Melvin Stewart ran errands and sold cocaine; and that Jeffrey and Raynice Thompson packaged large amounts of drugs for street sale. The jury convicted all nine of conspiracy to distribute and to possess with intent to distribute more than five kilograms of cocaine in violation of 21 U.S.C. §§ 841(a) and 846. It also convicted five defendants of additional crimes: Rachelle Edmond and the Thompsons were convicted of Travel Act violations on a conspiracy theory (18 U.S.C. § 1952(a)); Stewart was convicted of unlawful distribution of cocaine (21 U.S.C. §§ 841(a)(1) and (b)(1)(C)); and Morgan was convicted of possession with intent to distribute over 500 grams of cocaine (21 U.S.C. §§ 841(a)(1) and (b)(1)(B)(ii)(II)). The trial on the weapons and murder charges — the third trial — took place in June 1990. The government’s evidence suggested that appellant Columbus Daniels accompanied organization kingpin Rayful Edmond to a nightclub where the two met with a drug dealer who had purchased narcotics from Edmond. The three argued about money that the dealer still owed the organization. After the argument, Daniels shot the dealer seven times on Edmond’s signal, killing him. The jury convicted Daniels of second-degree murder while armed and of carrying a pistol without a license. In August 1990, the district court granted appellant Morgan’s motion for acquittal of the conspiracy charge on the grounds that his single corroborated purchase of cocaine from the Edmond organization and his uncorroborated confession of multiple drug dealings could not establish that he was a regular participant in the narcotics enterprise. The court denied all of appellants’ other post-trial motions, United States v. Childress, 746 F.Supp. 1122 (D.D.C.1990). Appellants were sentenced in September 1990. All appellants now challenge their convictions in the Group II and III trials and the denial of their joint motion for a new trial. Eight of the appellants — all but Morgan— also challenge the length of their sentences. We have considered all of the objections raised by appellants, but given the enormous number of issues involved, we address only those meriting separate discussion. We reject those challenges we do not discuss. We begin with appellants’ joint objections and then proceed to their individual ones. II. Jury AND Courtroom Prooedures Appellants jointly challenge the district court’s overall conduct of the Group II trial. They claim that they were denied a fair trial by the district court’s decisions to empanel an anonymous jury and keep it sequestered, to hold the trial in a secure courtroom, and to find the jury impartial notwithstanding its exposure to pretrial publicity. None of these decisions requires a new trial. A. Anonymous and Sequestered Jury Following the trial of the Group I defendants, the United States moved that the jury in the Group II trial be kept anonymous and sequestered. The district court granted the motion, United States v. Edmond, 730 F.Supp. 1144 (D.D.C.1990), ordering that the names, addresses, and workplaces of the Group II venire pool be withheld from all counsel and the media. To justify its order, the court cited numerous attempts by members of the Edmond group and their associates to disrupt the Group I trial and to intimidate witnesses with threats and actual violence. In addition, it credited a prosecutor’s declaration that several potential witnesses refused to testify in the Group II trial out of fear and noted an FBI agent’s sworn declaration that, according to a reliable informant, a reward was available on the street to anyone who assassinated a key government witness. Id. at 1146-47. As a substitute for revealing the prospective jurors’ names, addresses, and places of employment or business, id. at 1149, the court gave the venire members a twenty-three-page questionnaire designed to solicit information about their demographies, general lines of work, and familiarity with the events and parties in the case, id. at 1159-65. It also sought to downplay the significance of the jury’s anonymity and sequestration. Throughout the voir dire, the court told the venire members that keeping the jury anonymous and sequestered was “a common practice followed in many cases in federal court” and “in no way unusual”; it was being done “to protect your privacy ... [and] so that no one can later say that the integrity of the process was tainted by any improper outside contact or conduct.” The court emphasized to the potential jurors that these precautions indicated nothing about the defendants’ guilt or innocence and it repeated its explanation for the safeguards in its charge to the jury at the close of the trial. Appellants now argue that the use of anonymous juries violates the Constitution because they create a prejudicial trial atmosphere and deny defendants meaningful opportunities to conduct voir dire and exercise peremptory challenges intelligently. In terms of their own trial, they claim that the government’s assertions of potential danger to the jury were vague and did not implicate any specific Group II defendants; they also argue that the dangerousness of the Group I defendants and their behavior at the first trial cannot justify empaneling an anonymous, sequestered jury in the second case. Finally, they say, the district court’s repeated anonymity instructions and statements that anonymity was routine (statements that in their view were obviously false), combined with the pressures of sequestration, subliminally communicated to the jurors that the defendants were especially dangerous and most likely guilty. We reject these arguments. No court — state or federal — has ever held that the use of anonymous juries is per se unconstitutional, and no federal court has ever overturned a conviction rendered by an anonymous jury for that reason alone. In our recent decision upholding the empaneling of an anonymous jury in the Group I trial, we joined the Second, Third, Seventh, and Eleventh Circuits in approving of their use in some cases. See Edmond, 52 F.Sd 1080; see also United States v. Wong, 40 F.3d 1347 (2d Cir.1994); United States v. Ross, 33 F.3d 1507 (11th Cir.1994); United States v. Crockett, 979 F.2d 1204 (7th Cir.1992); United States v. Scarfo, 850 F.2d 1015 (3d Cir.1988). We held that a district court may empanel an anonymous jury if it “conclud[es] that there is a strong reason to believe the jury needs protection [ ... and] tak[es] reasonable precautions to minimize any prejudicial effects on the defendant and to ensure that his fundamental rights are protected.” Edmond, 52 F.3d at 1090 (quoting United States v. Paccione, 949 F.2d 1183, 1192 (2d Cir.1991)). The Group I appellants assumed that a trial court’s decision to empanel and sequester an anonymous jury was reviewable only for abuse of discretion, an assumption implicitly followed by the Edmond court. See Edmond, 52 F.3d at 1091 (holding that the empaneling of an anonymous jury is “left to the district court’s discretion,” subject to the broad constraints noted above) (quoting Paccione, 949 F.2d at 1192). Appellants in this case, however, assert that these decisions should be reviewable de novo. We disagree. Decisions on sequestration and anonymity require a trial court to make a sensitive appraisal of the climate surrounding a trial and a prediction as to the potential security or publicity problems that may arise during the proceedings. With so many factors entering the calculus, each varying subtly, an appellate court’s de novo resolution of the issue would merely duplicate the trial judge’s efforts and yet yield almost nothing of precedential value. “Fact-intensive disputes, those whose resolution is unlikely to establish rules of future conduct, are reviewed under a deferential standard because the role of appellate courts in establishing and articulating rules of law is not at stake.” Mars Steel Corp. v. Continental Bank N. A., 880 F.2d 928, 933 (7th Cir.1989) (en banc) (discussing application of Rule ll’s “reasonable inquiry” requirement); see also id. at 936 (noting that “[fjact-bound resolutions cannot be made uniform through appellate review, de novo or otherwise” and that an appellate pronouncement in such a case “is unlikely to establish clear guidelines for lower courts; nor will it clarify the underlying principles of law.”); Cooter & Gell v. Hartman Corp., 496 U.S. 384, 405, 110 S.Ct. 2447, 2460, 110 L.Ed.2d 359 (1990) (citing and quoting Mars with approval). Furthermore, some of the relevant factors, such as the degree of menace presented by the defendants and the intensity of media interest, may be only incompletely captured in the written record, so that courts of appeal are particularly ill-equipped to second-guess these judgments. Cf. Scarfo, 850 F.2d at 1023 (noting that review of decision to empanel anonymous jury “must be particularly deferential to the trial judge, familiar as he is with the local ambiance”). Finally, the factors counseling deference to the trial court’s decision to empanel an anonymous jury apply equally to its decision on sequestration. See United States v. Persico, 832 F.2d 705, 718 (2d Cir.1987) (“[sjequestration is a matter committed to the sound discretion of the trial court, and its decision will not constitute reversible error absent a showing of actual prejudice arising therefrom”); United States v. Greer, 806 F.2d 556, 557-58 (5th Cir.1986) (similar). The district court did not abuse its discretion in empaneling the Group II jury. The court reasonably found the serious potential for juror intimidation during and after the trial that would justify the extreme precautions of anonymity and sequestration. Though perhaps more peripheral than their Group I counterparts, the Group II defendants were still themselves alleged participants in an organized and extremely violent criminal conspiracy, and one of this group— appellant Daniels — stood accused of committing a brutal murder in furtherance of the conspiracy’s ends. The organization retained the capacity to threaten and harm jurors, even though its highest leaders were in jail. Furthermore, the district court reasonably thought that the Group I defendants or their associates would be inclined to interfere with the second trial, given that most of the Group II defendants were close friends or family members of people convicted in the first round; in fact, five of the nine Group II defendants were married or otherwise related to one or more of the Group I defendants. Contrary to appellants’ arguments, it was entirely appropriate in this context for the court to consider the dangers posed, not just by the Group II defendants themselves, but by other members of the conspiracy as well. Cf. Wong, 40 F.3d at 1377 (citing risks posed by non-defendant fellow gang members to justify anonymous jury); United States v. Vario, 943 F.2d 236, 240, 241 (2d Cir.1991) (attributing to defendant grand-jury tampering by co-conspirator and noting that “demonstrable history or likelihood of obstruction of justice on the part of the defendant or others acting on his behalf” justifies anonymity) (emphasis added). In fact, the district court’s experience with the Group I trial arguably provided it with more relevant evidence justifying an anonymous jury in this case than it possessed the first time around. Prior to the Group I trial, the court primarily knew just that the defendants allegedly belonged to a criminal enterprise that had used violence on the streets and that government informants had heard the lead defendant’s father say he would “take care of’ witnesses; the court’s conclusion that the defendants or their colleagues might tamper with the upcoming proceedings required an inferential leap, albeit a proper one. But that first trial proved the inference correct: The court now knew that associates of the Edmond organization were actually willing and able to interfere with the criminal trials of group members, and the only necessary inference was that they might continue such transgressions. During the Group I trial, the house of one witness’s mother was firebombed in the middle of her testimony, a potential witness was found shot after a letter from prosecutors was accidentally sent to a house she shared with a defendant, the court received bomb threats during the trial, and both the judge and courtroom personnel observed audience members communicating with the defendants by hand signals and glaring menacingly at witnesses and jurors. Furthermore, the declarations submitted by the government suggested that similar abuses might occur at the second trial. Such willingness to interfere with witnesses and trial proceedings indicates a real danger that defendants might threaten or otherwise tamper with jurors. See Edmond, 52 F.3d at 1092. Finally, the district judge knew from his experience with the first trial that the Group II proceedings would likely attract increasing media attention as they progressed, heightening the danger that information identifying the jurors could become public and potentially exposing them to intimidation or harassment. This consideration, too, was proper. See id. at 1091. We also find that the district court took appropriate precautions to minimize any prejudice to the defendants that might have resulted from the way the jury was empaneled. The court conducted a searching voir dire and gave jurors an extensive questionnaire, the scope of which appellants do not challenge. The judge’s repeated statements downplaying the significance of anonymity and sequestration and stressing their irrelevance to the defendants’ guilt or innocence were likewise appropriate. See Edmond, 52 F.3d at 1093 (approving identical instructions); Ross, 33 F.3d at 1521-22 n. 27 (approving of similar combination of downplaying safeguards and highlighting presumption of innocence); Crockett, 979 F.2d at 1216-17 (same); United States v. Tutino, 883 F.2d 1125, 1133 (2d Cir.1989) (same). Further, in the absence of some concrete reason to believe that jurors would discredit the judge’s suggestions that these procedures were commonplace, we will not indulge appellants’ assumption that the jurors suffered some sort of cognitive dissonance between what they knew of the conduct of trials and what they saw, much less that they would be led to infer that the defendants were guilty. In short, we find error in neither the district court’s initial decision to empanel and sequester an anonymous jury in the Group II trial nor its manner of doing so. B. Heightened Courtroom Security The Group II trial took place in the “secure courtroom” of the United States Courthouse in Washington. That courtroom had a twelve-foot-high plexiglass partition separating the spectator section from the well of the court (i.e., the counsel tables, jury box, and bench). Two videocameras were located in the top rear corners of the courtroom. The record indicates that there were more courthouse security personnel than usual on hand throughout the proceedings, although we have no record as to how many officers were actually present at trial, where in the courtroom they were stationed, whether they were uniformed, and the like. It is also uncertain whether there was any additional security equipment (metal detectors, for example) in or outside the courtroom. Appellants jointly petitioned to move their trial out of the secure courtroom and to reduce the number of security personnel present, claiming that these measures “im-permissibly ereate[d] the impression of the assemblage of a group of wild desperadoes.” The district court denied the requests, citing the security and manageability concerns presented by a trial with such a large number of defendants and involving “an organization that purchased its place in the community through the spilling of blood.” During the Group II jury selection, however, the court did instruct the venire members to ignore the extra security precautions: The judge noted that many different civil and criminal trials were held in that same courtroom, that the partitioned courtroom design was common in courthouses across the country, and that uniformed security personnel were always present at criminal trials “in direct proportion to the number of defendants on trial.” The court repeatedly made clear that none of these measures had anything to do with the guilt or innocence of the defendants, and it confirmed several times that the veni-re members understood the instructions. Appellants now claim that these extra security measures — especially in combination with the jury’s sequestration and anonymity-denied them a fair trial by signaling to the jurors that the defendants were dangerous and most likely guilty. They correctly point out that the constitutional presumption of innocence may be undermined by the physical indicia of guilt; criminal defendants do have a right to be free of court-imposed physical appearances that are unfairly suggestive of their guilt. See, e.g., Estelle v. Williams, 425 U.S. 501, 96 S.Ct. 1691, 48 L.Ed.2d 126 (1976) (holding unconstitutional a requirement that defendant appear in prison garb at trial). On the other hand, this “does not mean ... that every practice tending to single out the accused from everyone else in the courtroom must be struck down,” Holbrook v. Flynn, 475 U.S. 560, 567, 106 S.Ct. 1340, 1345, 89 L.Ed.2d 525 (1986), especially when the proceedings present legitimate security concerns to which the presiding judge must respond. Like the decision to empanel an anonymous jury, the trial court’s choice of courtroom security procedures requires a subtle reading of the immediate atmosphere and a prediction of potential risks — judgments nearly impossible for appellate courts to second-guess after the fact. For that reason, the balancing of the competing concerns for the presumption of innocence and for the integrity of the courtroom and its proceedings is best left to the sound discretion of the trial judge. See, e.g., Scarfo, 850 F.2d at 1024; United States v. Nicholson, 846 F.2d 277, 279 (5th Cir.1988). In light of the security concerns noted above and the large number of defendants at the proceedings, we cannot say that the trial judge abused his discretion. Appellants do not point to any evidence of actual prejudice resulting from the security measures taken in their trial. Nor will we presume prejudice: We agree with another district judge who has held trials in the secure courtroom that the plexiglass partition and the video-cameras (meticulously described in his opinion) are minimally intrusive, do not come between the jury box and the defendants or witnesses, and are far less stigmatizing than many other security measures—such as the shackling of unruly, flight-prone, or dangerous defendants—used in other courtrooms and upheld by other circuits. See United States v. Whitehorn, 710 F.Supp. 803, 835-41 (D.D.C.), rev’d on unrelated grounds sub nom. United States v. Rosenberg, 888 F.2d 1406 (D.C.Cir.1989). Appellants have likewise failed to demonstrate that the number of security officers present during the proceedings was disproportionate to the number of defendants being tried or that the officers were stationed in the courtroom in a way that might particularly influence the jury. Finally, the district court’s lengthy instructions to the venire members seem more than adequate to alleviate any incidental prejudice that may have resulted. C. ' Pretrial Publicity Prior to the severance of the trials, appellants moved jointly for a change of venue, claiming that the media attention focused on the proceedings would make a fair trial in the jurisdiction impossible. The motion stated that the ease had attracted lots of publicity because of the large quantities of drugs the enterprise was alleged to have brought into the District, the number of homicides potentially connected to the defendants, the fact that several defendants associated with members of the Georgetown University basketball team, and the District’s recent anointment as the murder capital of the nation. Thirty-eight articles from local newspapers were attached to the motion, many dealing with the defendants’ proceedings, but others simply discussing drug violence, narcotics arrests, and gangs generally. The district court denied the motion for a change of venue; however, it included in the Group II juror questionnaire several questions asking whether venire members had been exposed to the media coverage of this or related eases, it individually questioned prospective jurors who had indicated on their questionnaires that they had been so exposed, and it asked the venire pool repeatedly whether they could put aside any opinions formed from this publicity. After their conviction, appellants moved for an acquittal or a new trial based upon the denial of the change of venue. The district court denied the motion, holding that the media coverage of the case, though extensive, had been dispassionate and factual, that the court had screened extensively for potential bias, and that appellants’ counsel had leftover peremptory challenges with which they could have stricken any juror they thought potentially prejudiced. See Childress, 746 F.Supp. at 1138-40. Appellants now do not specifically challenge the denial of their motion for a change in venue, nor do they challenge the scope of the district court’s questionnaire or voir dire. Instead, they argue that the media attention surrounding the case was so inflammatory that the jury could not have rendered a fair verdict. Although they cannot point to any indications that a single juror was actually prejudiced by the pretrial publicity, appellants urge us to infer prejudice from the fact that twenty-seven members of the venire pool indicated that they had been exposed to the media coverage. We reject this claim. The mere existence of intense pretrial publicity is not enough to make a trial unfair, nor is the fact that potential jurors have been exposed to this publicity. Although the Constitution is understood to require that defendants be judged by “a panel of indifferent jurors,” those jury members “need not be totally ignorant of the facts and issues involved,” Irvin v. Dowd, 366 U.S. 717, 722, 81 S.Ct. 1639, 1642, 6 L.Ed.2d 751 (1961). Rather, “it is sufficient if the juror can lay aside his impression or opinion and render a verdict based on the evidence presented in court.” Id. at 723, 81 S.Ct. at 1643. We review the trial court’s finding of juror impartiality only for “manifest error,” Mu’Min v. Virginia, 500 U.S. 415, 428-29, 111 S.Ct. 1899, 1906-09, 114 L.Ed.2d 493 (1991); Patton v. Yount, 467 U.S. 1025, 1031, 104 S.Ct. 2885, 2888, 81 L.Ed.2d 847 (1984); and absent a showing that individual jurors were actually prejudiced and unable to meet this standard, we will infer prejudice only in those “rare” eases, Nebraska Press Ass’n v. Stuart, 427 U.S. 539, 554, 96 S.Ct. 2791, 2800, 49 L.Ed.2d 683 (1976), in which the community has been saturated with particularly damning publicity. The Supreme Court presumed jury prejudice, for example, where a defendant’s videotaped murder confession was broadcast on three consecutive nights to audiences of 24,000, 53,000, and 29,000 people in a rural parish of 150,000 residents. Rideau v. Louisiana, 373 U.S. 723, 83 S.Ct. 1417, 10 L.Ed.2d 663 (1963). But such a presumption is reserved for only the most egregious cases. The Court, for example, would not presume prejudice even where a defendant’s crime had provoked “substantial” unfavorable publicity — more than would have attended most capital murders because the defendant was an inmate charged with committing murder while on work release and the crime had occurred during a presidential campaign highlighting similar crimes by furloughed inmates. Mu’Min, 500 U.S. at 428-30, 111 S.Ct. at 1906-08. The Mu’Min Court stressed the importance of the community context, which as here was part of the “metropolitan Washington statistical area, which has a population of over 3 million and in which, unfortunately, hundreds of murders are committed each year.” Id. at 429, 111 S.Ct. at 1907. Simply put, the standard is high, and appellants do not meet it. Appellants merely rehash the newspaper articles they submitted with their original motion for a change in venue before the Group I trial; the latest of these stories ran seven months before their own trial started. In fact, counsel for several appellants argued to the district court that media coverage just before the Group II trial had been so thin that jury sequestration and anonymity were unnecessary. Twenty-seven members of the venire pool may have been exposed to whatever publicity there was, but only two of these people made it onto the jury and then only after the court questioned them individually and appellants’ counsel did not object. In short, we find no “manifest error” in the district court’s finding that the jurors were impartial, Childress, 746 F.Supp. at 1139-40. As the Edmond court found in reviewing the Group I trial, “there is no reason for concluding that the population of Washington, D.C. was so aroused against appellants and so unlikely to be able objectively to judge their guilt or innocence on the basis of the evidence presented at trial” that their right to a fair trial was violated. Edmond, 52 F.3d at 1099 (internal quotes omitted). III. CONSPIRACY Intent InstRuction Appellants challenge the trial court’s refusal to instruct the jury that specific intent is an element of conspiracy. We conclude that the district court erred in instructing that a conspiracy to possess and distribute drugs under 21 U.S.C. § 846 (1988) is a general intent crime, but this error was harmless because its instructions correctly apprised the jurors of the elements of intent they had to find in order to convict the appellants of conspiracy. The district court instructed the jury that there are two elements to a conspiracy charge under § 846: first, that an agreement existed between two or more persons to distribute or possess with intent to distribute the requisite amount of cocaine or cocaine base, and second, that the defendant “knowingly and voluntarily joined the conspiracy.” XIV Joint Appendix (“J.A.”), 3/30/90 Tr. at 61. The court also instructed that the conspiracy charged “require[d] only a general intent.... Where this is so, and it is shown that a person has knowingly committed an act which the law makes a crime, intent may be inferred from doing the act.” XIV J.A., 3/30/90 Tr. at 65. In elaboration on the requisite elements, the court explained that (1) the government was required to prove that the members of the conspiracy “came to a mutual understanding to accomplish an unlawful purpose or a lawful purpose by unlawful means,” XIV J.A., 3/30/90 Tr. at 62, (2) that “the fact that the acts of a defendant, without knowledge, merely happen to further the purposes or objectives of the conspiracy, does not make the defendant a member of the conspiracy,” id. at 63, and (3) “if a defendant, with an understanding of the unlawful character of the conspiracy, knowingly encourages, advises or assists in furthering the purpose of the conspiracy, that defendant thereby becomes a knowing and voluntary participant and member of the conspiracy.” Id. Appellants objected to these instructions at the close of the evidence, arguing that conspiracy is a specific intent crime. Rachelle Edmond, for instance, argued that under conspiracy law a defendant “not only ha[s] to have knowledge, but ... ha[s] to intentionally become a member [of the conspiracy] and ... become a member with the specific intent of furthering the objects of the conspiracy.” XIII J.A., 3/28/90 Tr. at 73. Raynice Thompson proposed an instruction that the jury must find that the defendants participated in the conspiracy “with the intent to see that the object and purposes of the conspiracy were achieved.” Proposed Modification of Jury Instructions (Mar. 30, 1990) at 2, reprinted in II J.A. at 158. The district court rejected these requests, adhering to its position that conspiracy under § 846 is a general intent crime. In United States v. Bailey, 444 U.S. 394, 100 S.Ct. 624, 62 L.Ed.2d 575 (1980), the Supreme Court discussed the distinction between general and specific intent. Noting that the “venerable distinction ... has been the source of a good deal of confusion,” the Court observed that, “[i]n a general sense, ‘purpose’ corresponds loosely with the common-law concept of specific intent, while ‘knowledge’ corresponds loosely with the concept of general intent.” 444 U.S. at 403, 405, 100 S.Ct. at 631, 632. As to the difference between knowledge and purpose, the Court explained that a person who causes a particular result is said to act purposefully if he consciously desires that result, whatever the likelihood of that result happening from his conduct, while he is said to act knowingly if he is aware that that result is practically certain to follow from his conduct, whatever his desire may be as to that result. 444 U.S. at 404, 100 S.Ct. at 631-32 (internal quotations omitted). The Court further explained that while proof of knowing action is generally adequate to support criminal conviction, certain classes of crimes merit “special attention” to “heightened culpability.” Id. at 405, 100 S.Ct. at 632. Among these crimes, the Court identified conspiracy: “Another such example is the law of inchoate offenses such as attempt and conspiracy, where a heightened mental state separates criminality itself from otherwise innocuous behavior.” Id. With this understanding in mind, it is clear that conspiracy is a “specific intent” crime. The common law definition of conspiracy is “ ‘a combination of two or more persons ... to accomplish some criminal or unlawful purpose, or to accomplish some purpose, not in itself criminal or unlawful, by criminal or unlawful means.' ” 2 LaFave & Scott, Substantive Criminal Law § 6.5 at 86 (1986) (quoting Commonwealth v. Hunt, 45 Mass. (4 Met.) 111 (1842)). Thus, purposeful intentor “conscious desire” to achieve a “result,” Bailey, 444 U.S. at 404, 100 S.Ct. at 631 — is the essence of conspiracy. Accordingly, as we have explained in the past, proof of conspiracy requires proof of specific intent to further the conspiracy’s objective: “A single conspiracy is proven if the evidence establishes that each conspirator had the specific intent to further the common unlawful objective.” United States v. Tarantino, 846 F.2d 1384, 1392 (D.C.Cir.1988); see also United States v. Clarke, 24 F.3d 257, 264-65 (D.C.Cir.1994) (to convict defendants of conspiracy to possess drugs with intent to distribute, “the government had to establish ... that the defendants purposefully agreed to act in partnership”) (emphasis in original); United States v. Haldeman, 559 F.2d 31, 112 (D.C.Cir.1976) (“[T]he specific intent required for the crime of conspiracy is ... the intent to advance or further the unlawful object of the conspiracy”). Indeed, at oral argument the government conceded that the charged conspiracy is a specific intent crime. We briefly address ourselves to the district court’s reasons for concluding otherwise. Recognizing that conspiracies are, in general, specific intent crimes, the district court nevertheless concluded that a § 846 conspiracy is different because (1) Congress legislated a distinct definition of conspiracy in § 846, see United States v. Childress, 746 F.Supp. at 1128 n. 8, and (2) several other circuits have construed the elements of a § 846 conspiracy to require that “ ‘the government must prove only that the defendant knew of it, and that, with knowledge, the defendant voluntarily became a part of the conspiracy,’ ” id. at 1127 (quoting United States v. Terzado-Madruga, 897 F.2d 1099, 1121 (11th Cir.1990)). Neither ground is persuasive. Although a § 846 conspiracy is different from a general federal conspiracy in certain other respects, they do not differ in the intent required. As the Supreme Court held in United States v. Shabani, — U.S. -, -, 115 S.Ct. 382, 383, 130 L.Ed.2d 225 (1994), a § 846 conspiracy, unlike a conspiracy under the general federal conspiracy statute, 18 U.S.C. § 871 (1988), requires no overt act. But this distinction is in the text of the respective statutes: § 371 requires proof that “one or more of such persons do any act to effect the object of the conspiracy,” while § 846 contains no such requirement. In the absence of a statutory requirement of an overt act, the common law controls, and the common law required no overt act. See — U.S. at -, 115 S.Ct. at 384. With respect to the intent requirement, by contrast, there is no textual basis for a distinction between § 371 and § 846; they are equally silent on the issue. Section 846 provides that “[a]ny person who attempts or conspires to commit any offense defined in this subchapter [Control and Enforcement of Drug Abuse] shall be subject to the same penalties as those prescribed for the offense, the commission of which was the object of the attempt or conspiracy.” Section 371 provides: “If two or more persons conspire ... to commit any offense against the United States, ... and one or more of such persons do any act to effect the object of the conspiracy,” each shall be guilty of conspiracy. As neither establishes an explicit intent requirement, each is governed by the general law of conspiracy, and the case law in this circuit and others is clear that conspiracy is a specific intent crime. See, e.g., Tarantino, supra; United States v. Rivera, 6 F.3d 431, 443 (7th Cir.1993) (“[B]ecause drug conspiracy is a specific intent crime, the government must prove intent as an element of the offense.”) (internal citation omitted); United States v. Rengifo, 858 F.2d 800, 808 (1st Cir.1988) (“While association with conspirators is evidence of participation in the conspiracy, something more is needed to show beyond a reasonable doubt the deliberate, knowing, and specific intent of the defendant to join the conspiracy.”) (internal quotation omitted). In further support of its conclusion that a § 846 conspiracy is distinct from the general conspiracy statute, with no specific intent requirement, the court pointed to several circuit court opinions holding that “ ‘[i]n order to convict a defendant of a Section 846 conspiracy, the government must prove only that the defendant knew of it, and that, with knowledge, the defendant voluntarily became a part of the conspiracy.’ ” 746 F.Supp. at 1127 (quoting Terzado-Madruga, 897 F.2d at 1121 & collecting cases). These cases do not support the district court’s conclusion that there is no specific intent requirement in conspiracy. Although there is a “loose[]” “eorrespond[ence]” between “ ‘knowledge’ [and] ... the concept of general intent,” Bailey, 444 U.S. at 405, 100 S.Ct. at 632, a characterization of a crime as requiring knowledge does not of its own force preclude a specific intent requirement. To the contrary, the Eleventh Circuit, one of the district court’s principal authorities, requires proof of specific intent in order to prove “knowing and voluntary” participation: “[t]o prove [the defendant’s] knowing and voluntary participation, the Government must prove beyond a reasonable doubt that [he] had a deliberate, knowing, and specific intent to join the conspiracy.” United States v. Harris, 20 F.3d 445, 452 (11th Cir.) (internal quotation omitted) (emphasis added), cert. denied, — U.S. -, 115 S.Ct. 434, 130 L.Ed.2d 346, cert. denied, — U.S. -, 115 S.Ct. 611, 130 L.Ed.2d 521, cert. denied, — U.S. -, 115 S.Ct. 612, 130 L.Ed.2d 521 (1994). Thus, although the “knowing” and “purposeful” terms are not always used with perfect consistency, there is, so far as we can see, no fundamental dispute that a § 846 conspiracy is a specific intent crime. Though the district court thus erred in instructing the jury that a § 846 conspiracy is a “general intent” crime, this error was harmless because the district court’s instructions gave the jury adequate guidance on the intent required. Under the district court’s instructions, the jury had to find that each defendant entered the conspiratorial agreement with the purpose of furthering the ends of the conspiracy — -that he (1) “came to a mutual understanding to accomplish an unlawful purpose” and (2) “with an understanding of the unlawful character of the conspiracy, knowingly encourage[d], advise[d] or assisted] in furthering the purpose of the conspiracy.” In so characterizing the elements of the crime, the district court clearly instructed the jury that a purposeful state of mind was required — the jury could not find someone to be a “knowing and voluntary” participant in the conspiracy without finding that he or she “came to a mutual understanding to accomplish an unlawful purpose.” We are thus satisfied that the jury could not convict under these instructions without finding what amounts to “the specific intent to further the common unlawful objective.” Tarantino, 846 F.2d at 1392. In light of these instructions, the district court’s incorrect identification of the crime as one of “general intent” was harmless. IV. Single v. Multiple CONSPIRACIES Appellants argue that the evidence at trial varied from the indictment because it established multiple conspiracies rather than a single one and that they were prejudiced by this variance. In a related argument, several of the appellants argue that there was insufficient evidence to convict them of the charged conspiracy. Like the individual suf-fieieney-of-the-evidence arguments, “[t]he existence of a single conspiracy or multiple conspiracies is primarily a question of fact for the jury.” Tarantino, 846 F.2d at 1391. Our review of both claims is therefore limited to whether there is sufficient evidence— when viewed in the light most favorable to the government — to support a jury finding of a single conspiracy agreed to by the individual appellants. We address the claims together. A. Variance in Conspiracy Trials A defendant can establish a variance by showing that the evidence established multiple conspiracies rather than the single conspiracy charged in the indictment. Even if the defendant establishes such a variance, it is grounds for reversal only if the defendant also shows that he was substantially prejudiced by the variance — by, for instance, the spillover effect of evidence from other conspiracies to which he was not a party. See Tarantino, 846 F.2d at 1391. B. Single v. Multiple Conspiracies As is common in drug distribution cases, the government’s theory in this case is that the evidence establishes a single “chain” conspiracy. Courts have long recognized that participants in a continuous drug distribution enterprise can be parties to a single conspiracy even if they do not all know one another, so long as each knows that his own role in the distribution of drugs and the benefits he derives from his participation depend on the activities of the others. As the Second Circuit explained in concluding that separate groups of smugglers, middlemen, and retailers could form a single conspiracy to import drugs, “the smugglers knew that the middlemen must sell to retailers, and the retailers knew that the middlemen must buy of importers of one sort or another. Thus the conspirators at one end of the chain knew that the unlawful business would not, and could not, stop with their buyers.... That being true, a jury might have found that all the accused were embarked upon a venture, in all parts of which each was a participant.” United States v. Bruno, 105 F.2d 921, 922 (2d Cir.1939). Thus, while proof of conspiracy requires proof of an agreement, the jury may infer this agreement from the defendant’s knowing participation in a distribution network organized along division of labor principles — in which his own role necessarily depends on the cooperation of other parties to that network: “The existence of such a vertically integrated, loose-knit combination may raise the inference that each conspirator has agreed with the others (some whose specific identity may be unknown) to further a common unlawful objective, e.g., the distribution of narcotics.” Tarantino, 846 F.2d at 1393 (internal quotation omitted). The chain metaphor, of course, is not without limits in its ability to establish a single conspiracy. Even if, for instance, there exists a core, single chain conspiracy, “certain players may have performed activities wholly unrelated to the aims of the conspiracy.” Tarantino, 846 F.2d at 1393. In addition, some courts have been reluctant to conclude that the chain conspiracy construct can automatically bind all participants in a drug distribution enterprise into a single agreement when certain participants are involved in the enterprise during radically different time periods, see United States v. Borelli, 336 F.2d 376, 383-85 (2d Cir.1964), or when there are no indications of interdependence between the various participants, see United States v. Anderson, 39 F.3d 331, 337, 347 (D.C.Cir.1994) (evidence of “an extensive cocaine distribution network organized and managed by” a single kingpin who “purchased cocaine from several suppliers ... and resold it to both wholesale distributors and street level dealers” “likely ... manifested] ... several conspiracies rather than a single overarching one”), en banc rehearing ordered on unrelated grounds, 1995 WL 79398 (D.C. Feb. 14, 1995) (en banc); United States v. Townsend, 924 F.2d 1385, 1391, 1395-1402 (7th Cir.1991) (finding three separate conspiracies between three different suppliers and a common purchaser; “the liability of members of the distribution chain is predicated upon the notion that participants at different levels in the chain know that the success of those at each level hinges upon the success of the others and therefore cooperate for their mutual benefit”). In this case however, we do not run up against such limits. As detailed below, we find that the evidence against each appellant was sufficient both to tie him to the core of the drug distribution operations and to support the jury’s inference that he knew of, depended on, and agreed to the related activities carried out by the co-conspirators. Overview of the Evidence. In the course of the trial, the government presented evidence of a single coordinated drug distribution network, orchestrated by Rayful Edmond III and geared toward retail trade on “the Strip,” a residential area in the 600 blocks of Morton Place and Orleans Place, N.E., with additional distribution in the separate Bates Street area of Northwest Washington. Between 1986 and 1989, police observed steady trafficking in cocaine and crack on the Strip, overseen by several co-conspirators who were tried in the Group I trial, including Emanuel Sutton, Jerry Millington, John Monford, James Antonio Jones, Keith Cooper, and Patrick McDonald. In 1988, police obtained search warrants for several stash houses on the Strip and carried out five separate searches, finding firearms and significant quantities of cocaine and/or money on each occasion. Shortly after one of these searches, the police found James Antonio Jones, Jerry Millington, and Patrick McDonald directing others in a clean-up of the house. On another occasion, a government agent purchased 13 grams of cocaine from appellant Melvin Stewart in the basement of one of the houses on the Strip, and, upon execution of a search warrant of the house that evening, police found Stewart in the house along with other co-conspirators and large quantities of drugs and money. In order to supply the Strip and the Bates Street operation, conspirators arranged large drug purchases from Los Angeles, using couriers such as appellant Willie Childress to ferry the money and drugs back and forth across the country. Edmond “lieutenants” such as Dave McCraw and James Minor took delivery of these drugs from the couriers and distributed them to safe houses on the Strip and Bates Street and to the apartments of various conspirators for storage, packaging, and further transport. On at least four occasions in the summer of 1988, McCraw and Minor retrieved cocaine from California couriers and drove it to Bates Street, where they were assisted by appellant Hardy in delivering the drugs to a house at Eighth Street and Rhode Island Avenue. A large cohort of conspirators assisted in storing and packaging the drugs for retail sale. For our purposes, the most relevant evidence is that appellant Stewart spent several hours packaging cocaine for retail sale, that Raynice and Jeffrey Thompson regularly packaged 1-2 kilograms of crack three times a week over a period of three months, and that Rachelle Edmond regularly stored and made available for retrieval drugs at the apartment she shared with Jerry Millington. Once the drugs were packaged, they were picked up by Edmond couriers such as cooperating witness Kathy Sellers, assisted on at least one occasion by Rachelle Edmond. The couriers would then deliver drugs daily to the stash houses early in the morning, unless told not to. These same couriers also retrieved the sales proceeds from the Strip and brought the money back to various locations, including Rachelle Edmond’s apartment and later her house. On one occasion in 1988, Sellers observed appellant Constance Perry in the basement of Rachelle Edmond’s house counting a large sum of money with a money counting machine. The government also presented evidence that appellant Perry brought or sent envelopes and paper bags full of old, dirty bills in small denominations for deposit into different credit union accounts once or twice a week. Once, she and Rayful Edmond III brought in $19,000 in old bills in exchange for a treasurer’s check. The government showed that Rachelle Edmond maintained a bank account with significant cash deposits during the time of her participation in the conspiracy. In sum, the government presented overwhelming evidence of a classic chain conspiracy operating between 1986 and 1989 to possess and distribute narcotics in the District of Columbia. The Individual Participants. We further conclude that there was sufficient evidence against each of the appellants to conclude that she or he agreed to further the purposes of this single conspiracy. As detailed below, there was for each appellant evidence from which the jury could infer that she or he knew of the scope of the conspiracy, that the individual’s own benefits depended on the related activities of the co-conspirators, and that, so knowing, she or he agreed to further the purposes of the conspiracy. Jeffrey and Raynice Thompson The Thompsons argue that there was insufficient evidence that they agreed to enter the conspiracy and that they knew “that their benefits depended on the entire venture’s success.” Appellants’ Brief at 102. The Thompsons note that they were not implicated in several of the conspiracy’s activities — they were never at the Strip, they did not make cash deposits at the bank, they never made drug purchases in California. Details of what the Thompsons did do, however, provided ample basis for the jury to conclude that they agreed to enter the conspiracy with knowledge of its scope. Government witness Rae Zanville testified that, on or about December 1987, and at the direction of Rayful Edmond III, Raynice Thompson helped her bag crack and powder cocaine. And, three times a week from January through March of 1988 both Thompsons, again at the direction of Rayful Edmond III, cut and bagged one to two kilograms of crack at a time and handed the packaged drugs to Kathy Sellers for transport. The government also introduced evidence of the Thomp-sons’ otherwise unexplained affluence during this period: They invested tens of thousands of dollars in property in late 1987 and early 1988. From the quantity of drugs the Thompsons personally handled, the fact that they were knowingly preparing those drugs for street sale, and the evidence that they were doing so at the personal direction of Rayful Edmond III, the jury could have inferred an intent by the Thompsons to agree to and further a broad-scale conspiracy to distribute narcotics. Rachelle Edmond Rachelle Edmond argues as well that there was insufficient evidence that she knowingly and willingly became a participant in the Rayful Edmond-Tony Lewis conspiracy. She suggests that the government never succeeded in implicating her as any more than an associate of conspirators rather than as a participant. She was not charged with money laundering and maintains that evidence suggesting that she laundered drug money does not support her conspiracy conviction. Contrary to Edmond’s assertion, however, the government presented evidence of her participation in a wide range of the conspiracy’s activities, including core drug distribution. Rachelle Edmond lived with co-conspirator Jerry Millington, who was tried and convicted with the Group I defendants. Although Millington’s role in the conspiracy may have been greater than Edmond’s, there was extensive evidence that Edmond herself knew of the scope of the conspiracy and acted in several different ways to further it. Government witness Sellers testified that at her request Edmond spoke to Millington to facilitate Sellers’ entry into the conspiracy as a drug courier, that Edmond was with Mill-ington when Sellers came to their apartment to retrieve drugs and that, on a few occasions, Edmond herself gave cocaine to Sellers for delivery down the line. Sellers further testified that Edmond accepted drug proceeds from Sellers in packages of $1,000 to $10,000 when Millington was not at home. Government witness Zanville testified in the same vein. She said that on one or more occasions Edmond and Sellers together retrieved crack packaged for street sale from Zanville’s apartment, and that Edmond was with Millington when Zanville delivered packaged crack to their apartment. In addition, the government presented evidence of an FBI-recorded conversation between Rachelle Edmond and Zanville in which Edmond discussed the pending police investigation and efforts by herself and others to avoid surveillance and detection. During this conversation Edmond revealed detañed knowledge of the scope of the conspiracy — stating, among other things, that co-conspirators possessed guns, dealt in 36-pound quantities of cocaine, and had California connections. Finally, the government introduced evidence of Edmond’s unexplained wealth during this period. Notwithstanding her limited income as a hairdresser and the unemployment of her live-in co-conspirator Millington, she maintained accounts with thousands of dollars in cash deposits in 1987-88, purchased a 1987 Volvo with a $6,000 down payment, put a total of $46,000 in down payments on two houses, and paid an $11,000 mobüe telephone Mil during the same period. This evidence in toto is more than adequate to support the jury’s finding that Edmond was a knowing and voluntary participant in the conspiracy who agreed to the furtherance of its purposes. Melvin Stewart Melvin Stewart also challenges the sufficiency of the evidence tying him to the conspiracy. Stewart notes that a single sale of drugs is insufficient to support a conviction of conspiracy and argues that the core evidence against him consisted of just such a single transaction. The government agrees that evidence that Stewart sold 13 grams of cocaine to an undercover police officer in the basement of 407 M Street, N.E., is “[e]entral to the proof of Stewart’s involvement in the conspiracy.” Government’s Brief at 100. As the government further notes, however, this was far from the only piece of evidence connecting Stewart to the conspiracy. Police testified that when they executed a search warrant at 407 M Street on the same evening that they had purchased the drugs from Stewart, Stewart was at the house along with other conspirators, a large amount of cocaine, and thousands of dollars. Six hundred eighty dollars was found on Stewart himself, even though he was purportedly unemployed. In addition, other witnesses testified to Stewart’s involvement in the conspiracy on separate occasions. Government witness Denise Johnson told how one day in the summer of 1987 she and several other conspirators, including Stewart, packaged cocaine for street sale. Another government witness, James Mathis, who served as an undercover courier for the conspiracy, testified that he received from Stewart a payment of approximately $5,000 for a drug sMpment he brought from Los Angeles to WasMngton, D.C. From these incidents, the jury could reasonably infer Stewart’s knowledge of the conspiracy’s scope and his agreement to further its aims. Willie Childress Chüdress also argues that there was insufficient evidence to support his conviction for conspiracy. He maintains that the evidence identifying him as the California courier who brought 50 kilograms of cocaine to Dave MeCraw and James Minor at the Crystal City Days Inn in August 1988 was too thin, and that even if all the evidence presented were believed, he could not be tied to the overall conspiracy. Although the government’s evidence against Childress is admittedly more circumstantial than for the other appellants, it nevertheless suffices to support Childress’s conviction. The evidence against Childress centered around a delivery of 50 kilograms of cocaine in August 1988, accepted by James Minor and Dave McCraw, and a similar delivery the week before about which there was no eyewitness testimony. Government witness Minor testified that he and Dave McCraw received a 50-kilogram package from an older bearded man with glasses, whom McCraw referred to as “Captain.” Minor further testified that McCraw told him that he had received another 50-kilogram delivery from the same man a week earlier. Although Minor made no in-court identification of Chil-dress, several pieces of circumstantial evidence permitted the jury to conclude that Childress was in fact the same “Captain” who made the deliveries: (1) Wiretap evidence showed that Rayful Edmond III tried unsuccessfully to reach the occupant of Room 226 at the Crystal City hotel on August 16. The next day Edmond was taped talking to co-conspirator Tony Lewis about giving something to two older people who had arrived in town in a van, one of whom Edmond called “Captain Willie.” (2) The hotel records of the Crystal City Days Inn showed that a person signing his name “Willie Chil-dress,” listing his home state as California and his car as a Chevrolet van checked into Room 226 on August 5 and stayed more than one night. Although the hotel room records after August 7 were lost, telephone logs indicated that throughout the entire period and as late as August 15, calls were made from Room 226 to a California number that Chil-dress claimed elsewhere as his own. (3) The signature on Childress’s driver’s license and that on the Days Inn records were both in evidence, and the government argued that they were identifiably from the same man. (4) In an interview with a Missouri Highway Patrol officer, Childress said that his nickname was “Cap.” (5) At trial, Childress had a beard and wore glasses. From this substantial network of evidence tying Childress to the observed drug delivery of August 1988 and the prior recollected delivery, the jury could conclude that Childress was, indeed, the “Captain” who made the two deliveries. Childress further argues that even if the jury could find that he was the “Captain” who made the drug delivery received by Minor, it could not, from this transaction alone, find him guilty of conspiracy. Cf. United States v. Kimmons, 917 F.2d 1011, 1016 (7th Cir.1990) (“The relationship of buyer and seller absent any prior or contemporaneous understanding beyond the mere sales agreement does not prove a conspiracy.”) (internal citation omitted); accord United States v. Morris, 836 F.2d 1371, 1374 (D.C.Cir.1988). But see United States v. Medina, 944 F.2d 60, 65 (2d Cir.1991) (“The rationale for holding a buyer and a seller not to be conspirators is that in the typical buy-sell scenario, which involves a casual sale of small quantities of drugs, there is no evidence that the parties were aware of, or agreed to participate in, a larger conspiracy.”). However, the jury was entitled to credit Minor’s testimony that the same “Captain” made two drug deliveries of 50-kilogram magnitude to the Edmond organization. Two deliveries of this magnitude suggest a continuity of relationship between Childress and the Edmond organization and support the inference that Childress knew that the organization to which he was delivering such a sizeable amount of drugs must involve a substantial distribution network. Evidence of these two deliveries is thus sufficient to prove his agreement to participate in the conspiracy. Columbus Daniels Daniels likewise “adopts” the argument that there was insufficient evidence to support his conspiracy conviction. Appellants’ Brief at 170. Daniels makes no particular argument in support of this claim and, in the face of record evidence of Daniels’ substantial participation in the conspiracy, it is unsustainable. Government witness Desiree Murphy testified that on one occasion and at Rayful Edmond Ill’s personal direction, Daniels retrieved between 89 and 93 kilograms of crack from a courier’s apartment in Crystal City after the courier had been apprehended in Los Angeles. In addition, police frequently observed Daniels on the Strip. On one occasion, Daniels fled from a Merce