EBEL, Circuit Judge:
A federal jury convicted Defendant-Appellant Leland Lapier of conspiracy to possess with intent to distribute methamphetamine under 21 U.S.C. § 846 ("Count One") in addition to a separate offense of possession with intent to distribute methamphetamine. Lapier challenges only his conspiracy conviction, arguing that (1) his conviction was not supported by sufficient evidence, as the evidence at trial established only "buyer-seller" relationships and not a conspiracy, and (2) the district court plainly erred in failing to give a "specific unanimity" instruction sua sponte. We reject the first argument but agree with the second. The record in this case reflects a genuine possibility that different jurors voted to convict on the basis of different conspiracies.
Count One alleged a single conspiracy "beginning in or before September 2011, and continuing through at least December 2012," but the evidence at trial showed at least two separate conspiracies — not the single conspiracy charged. This created a "genuine possibility of jury confusion" and the risk of a nonunanimous verdict, United States v. Echeverry, 719 F.2d 974, 975 (9th Cir.1983), as some jurors might have convicted on the basis of a conspiracy with Lapier's first supplier (Louis Kanyid), and other jurors might have convicted on the basis of a separate conspiracy with Lapier's second, subsequent supplier (Burt Boucher). Because this possibility was not cured with a specific unanimity instruction, it violated Lapier's constitutional right to a unanimous jury verdict. See, e.g., United States v. Gordon, 844 F.2d 1397, 1400-02 (9th Cir.1988) ("When there is such a genuine possibility of jury confusion or that a conviction may occur as the result of different jurors concluding that the defendant committed different acts, the trial judge is obligated to give curative instructions or submit special interrogatories to ensure a unanimous verdict."). Therefore, exercising jurisdiction under 28 U.S.C. § 1291, we REVERSE Lapier's conviction on Count One.
Count One of the indictment alleged a single, overarching conspiracy:
The evidence, viewed in the light most favorable to the government, see Jackson v. Virginia, 443 U.S. 307, 319, 99 S.Ct. 2781, 61 L.Ed.2d 560 (1979), showed the following. In late 2011, Louis Kanyid (based in Kennewick, Washington) began supplying meth to Lapier (based in Great Falls, Montana) for redistribution. Their relationship became more like a partnership over time. Kanyid "fronted" meth to Lapier, allowing Lapier to resell the drugs in Great Falls and repay Kanyid with the sale proceeds. Around March 2012, Kanyid increased his sales to Lapier and began working in Lapier's garage, helping Lapier prepare the meth for redistribution by dividing it into smaller quantities. Kanyid's wife testified that Lapier and Kanyid "were basically partners" by this point. Around this time, Lapier also met and began fronting meth to Mark Abresch, a user and subdealer.
In July 2012, Abresch was arrested after buying meth from Burt Boucher in Spokane, Washington. Abresch had turned to Boucher as an alternative supplier because sometimes Lapier did not have meth. After his arrest, the police allowed Abresch to stay out of jail in order to act as a confidential informant. The police had Abresch wear a wire during his interactions with Boucher and buy undercover from Boucher, but did not ask him to wear a wire around Lapier or to buy undercover from him. However, Abresch continued buying meth from Lapier and occasionally selling to Lapier.
In September 2012, Kanyid — Lapier's supplier — was arrested. By this point, Kanyid had sold an estimated 3-3.5 pounds of meth to Lapier over the course of their relationship, which lasted for nearly a year.
Kanyid's arrest meant that Lapier needed to find a new meth supplier. In October 2012, Paul Kessler, a customer, introduced Lapier to Boucher. Boucher began selling drugs to Lapier directly, fronting the drugs to Lapier and usually getting repaid within a day.
In December 2012, Boucher was arrested. Lapier continued to supply Abresch with meth until Abresch was arrested again in February 2013. Finally, Lapier was indicted and arrested in March 2013.
A federal jury convicted Lapier of one count of conspiracy to possess with intent to distribute methamphetamine and one count of possession with intent to distribute methamphetamine. At trial, Lapier moved twice for a judgment of acquittal on both counts, arguing that there was insufficient evidence to support his convictions. The district court denied both motions. Lapier timely appealed, challenging only his conspiracy conviction.
We review de novo the district court's denial of Lapier's motion for judgment of acquittal based on insufficient evidence. See United States v. Mincoff, 574 F.3d 1186, 1191-92 (9th Cir.2009).
Lapier argues that the evidence the government presented at trial established only buyer-seller relationships between Lapier and Kanyid and Lapier and Boucher because neither supplier "had any input into, or interest in, what Lapier did with the methamphetamine they sold him — other than getting paid." According to the "buyer-seller rule,"
United States v. Moe, 781 F.3d 1120, 1123 (9th Cir.2015) (quotation marks omitted); see also United States v. Ramirez, 714 F.3d 1134, 1140 (9th Cir.2013) ("In the end, what we are looking for is evidence of a prolonged and actively pursued course of sales coupled with the seller's knowledge of and a shared stake in the buyer's illegal venture.") (quoting United States v. Thomas, 284 F.3d 746, 752 (7th Cir.2002)).
Lapier's buyer-seller argument fails because the evidence established that his relationships with both Boucher and Kanyid not only involved the purchase of drugs, but an agreement to further distribute them. See Moe, 781 F.3d at 1124-25. Both Kanyid and Boucher "fronted" drugs to Lapier, indicating that each supplier expected that Lapier would resell the drugs and use the proceeds of his sales to repay them. "[E]vidence of fronting may support a conviction for conspiracy to distribute a controlled substance." Mincoff, 574 F.3d at 1193. This evidence, combined with the large quantities of drugs Lapier bought, was sufficient for the jury to find that Kanyid and Boucher each knew Lapier would resell the meth and that each had a stake in Lapier's sales. There was also evidence that Kanyid helped Lapier prepare drugs for redistribution. This evidence is sufficient to support a jury verdict finding that Lapier entered into a conspiracy with Kanyid and a later, separate conspiracy with Boucher.
Because there was evidence of two different conspiracies, Lapier argues on appeal that the district court should
United States v. Marcus, 560 U.S. 258, 262, 130 S.Ct. 2159, 176 L.Ed.2d 1012 (2010) (quotation marks and alterations omitted).
"Normally, a general instruction on the requirement of unanimity suffices to instruct the jury that they must be unanimous on whatever specifications form the basis of the guilty verdict." Payseno, 782 F.2d at 835. However, a specific unanimity instruction is required if there is a "genuine possibility of jury confusion" or a possibility "that a conviction may occur as the result of different jurors concluding that the defendant committed different acts." Id. at 836 (quoting Echeverry, 719 F.2d at 975) (emphasis omitted). While jurors need not unanimously agree on the particular facts satisfying the overt act element of a conspiracy charge, see United States v. Gonzalez, 786 F.3d 714, 718-19 (9th Cir.2015), jurors must still unanimously agree that the defendant is guilty of participating in a particular conspiracy — i.e., of forming an agreement with at least one other particular individual to pursue a particular criminal goal, see, e.g., United States v. Monroe, 552 F.2d 860, 862 (9th Cir.1977) ("The crime of conspiracy is established once there is an agreement to engage in criminal activity and one or more overt acts are taken to implement the agreement."); United States v. Escobar de Bright, 742 F.2d 1196, 1199 (9th Cir.1984) ("A conspiracy is defined as an agreement between two or more people to commit an unlawful act, which arguably requires some form of a meeting of minds." (citations and quotation marks omitted)).
Lapier argues for the first time on appeal that the district court plainly erred in failing to give a specific unanimity instruction on the conspiracy count. Such an instruction was necessary in this case, he argues, because the indictment charged a single conspiracy while the evidence tended to establish several separate conspiracies, creating a genuine possibility of juror confusion and risking a nonunanimous verdict. The indictment alleged that Lapier participated with unnamed "others" in a single conspiracy "beginning in or before September 2011, and continuing through at least December 2012." However, the evidence at trial tended to show at least two separate conspiracies — one between Lapier and his first supplier, Kanyid, and a later one between Lapier and his second supplier, Boucher — but not a single overarching conspiracy. Some jurors might have convicted on the basis of the Lapier-Kanyid conspiracy and others might have convicted on the basis of the Lapier-Boucher conspiracy.
This is a genuine possibility on the record of this case. Whether there is a "genuine possibility of jury confusion" depends on the circumstances of a particular
Lapier's argument that such a genuine possibility of jury confusion existed here is especially strong in light of a colloquy between the district court and the prosecutor during argument pertaining to Lapier's motion for acquittal at the end of the government's case but before the case was submitted to the jury. During that discussion, the district court indicated its belief that the government had introduced evidence of two separate conspiracies: one with Kanyid and a later one with Boucher. The government acknowledged that there might be two separate conspiracies: "Your Honor, that is possible. They were two separate conspiracies. However, the Defendant was involved in both, yes."
An independent review of the evidence leads to the same conclusion. There was evidence that Lapier conspired with Kanyid from December 2011 until September 2012, and that Lapier conspired with Boucher from October 2012 until December 2012. But there was no evidence that Kanyid and Boucher had any agreement with each other or acted pursuant to a single conspiracy involving them both and spanning the single time period charged in the indictment ("beginning in or before September 2011, and continuing through at least December 2012").
The jury in this case was given a general unanimity instruction: "Your verdict, whether guilty or not guilty, must be unanimous." Ordinarily, such an instruction is sufficient to protect the right to a unanimous verdict. See Payseno, 782 F.2d at 835. Here, however, the indictment was sufficiently broad and the evidence sufficiently complex as to create a risk that different jurors voted to convict on the basis of different facts establishing different offenses. In these circumstances, the district court was required to give a specific unanimity instruction sua sponte. See, e.g., id. at 836 ("When ... a conviction may occur as the result of different jurors concluding that the defendant committed different acts, the general unanimity instruction does not suffice. To correct any potential confusion in such a case, the trial judge must augment the general instruction to ensure the jury understands its duty to unanimously agree to a particular set of facts." (quoting Echeverry, 719 F.2d at 975)).
Because the evidence in this case tended to show multiple conspiracies instead
The government argues that any violation of Lapier's right to a unanimous verdict was not plain error because there was not a "high probability that the error materially affected the verdict." Anguiano, 873 F.2d at 1319. That seeks to frame the plain error analysis as requiring Lapier to show that but for the error, a jury would likely have acquitted him. However, that is not the proper way to analyze whether Lapier's "substantial rights" were violated in this context. Lapier had a constitutional right to a unanimous jury verdict. Having determined that the possibility of jury confusion about what acts formed the basis of his conviction was "genuine," we are "not free to speculate" about what his jury might have concluded had it been properly instructed. See Payseno, 782 F.2d at 837.
The cases on which the government relies are distinguishable because they involved circumstances in which there was little or no potential for jury confusion. The case most helpful to the government is Anguiano, 873 F.2d at 1318-21. In that case, the court held that the district court's failure to give a specific unanimity instruction sua sponte was not plain error because the indictment in that case "was drafted with sufficient clarity so as to preclude the possibility of juror confusion as to what facts must be found before Anguiano could be convicted," and that the case's factual context "was not so complex as to suggest that juror confusion was likely." Id. at 1320. The indictment in Anguiano named specific co-conspirators in a way that excluded the alternative conspiracy, and the dates charged in the indictment matched the charged conspiracy but not the alternative conspiracy. See id. Further, the evidence and argument presented at trial made it clear to the jury that the charged conspiracy to acquire marijuana related to a particular transaction. On that record, there was "little potential for juror confusion." Id.
Similarly, in United States v. Kennedy, 726 F.2d 546 (9th Cir.1984), a case involving an allegedly duplicitous indictment, the court found no plain error because "[t]here was no danger that the jury could convict... without reaching unanimous agreement on a given set of facts." Id. at 548.
The government also cites United States v. Bosch, 951 F.2d 1546 (9th Cir.1991), United States v. LeMaux, 994 F.2d 684 (9th Cir.1993), and United States v. Hernandez-Escarsega, 886 F.2d 1560 (9th Cir. 1989), in support of its argument. All are distinguishable. In Bosch, the defendant argued that the district judge was biased toward the prosecutor and so should have recused himself; the court found no plain error because "a close analysis" of the record "shows that the judge did not favor the prosecutor in any way." 951 F.2d at 1548.
Hernandez-Escarsega and LeMaux were continuing criminal enterprise (CCE) cases. In neither case was a specific unanimity instruction given. However, in Hernandez-Escarsega, the jury returned separate convictions on two of the three required predicate acts needed to prove a CCE, and the evidence was overwhelming as to nine additional acts that the jury could have used as the basis for the third predicate act. The court concluded on
In some cases finding no error or no plain error when no specific unanimity instruction was given, the district court effectively gave a specific unanimity instruction without labeling the instruction as such; the jury returned specific verdicts or interrogatories that showed unanimity; or the instructions specified that the jury must unanimously find the defendant guilty of the single, specific conspiracy charged.
As noted previously, there is ordinarily no need to give a specific unanimity instruction in addition to the general unanimity instruction. Each of these cases turns on its potential, if any, for jury confusion and a non-unanimous verdict. In our case, by contrast, we have nearly a perfect storm: (1) the indictment was broadly worded and did not name the co-conspirators; (2) the indictment covered a specific time frame marked by specific dates that included both conspiracies; (3) the evidence adduced at trial credibly showed at least two separate conspiracies within the text and scope of the indictment; (4) the district court itself concluded, after hearing the government's case, that the evidence showed two conspiracies;
The Ninth Circuit recently discussed an issue related to this case in United States v. Gonzalez, 786 F.3d 714 (9th Cir.2015). Gonzalez reaffirmed that "a general unanimity instruction alone is insufficient `if it appears "that there is a genuine possibility of jury confusion or that a conviction may occur as the result of different jurors concluding that the defendant committed different acts."'" Id. at 717 (quoting United States v. Chen Chiang Liu, 631 F.3d 993, 1000 (9th Cir.2011)). However, the court said there that a specific unanimity instruction was not required with respect to the overt act element of a conspiracy — in other words, the jury need not unanimously agree on the particular overt act that furthered a "unanimously agreed-upon
The Gonzalez court's citation of Echeverry and Payseno, and its emphasis on the specific unanimity instruction that was given by the district court, make clear that Gonzalez did not intend to disturb the rule that jurors must unanimously agree that the defendant participated in a particular conspiracy with at least one particular conspirator and an agreed-upon criminal object. Where the indictment is broad enough to encompass multiple separate and distinct conspiracies, the evidence adduced at trial tends to show multiple conspiracies, and both the district court and the prosecutor agreed that the evidence could support multiple conspiracies, a "genuine possibility of juror confusion" would seem to exist. If so, the district court must give a specific unanimity instruction sua sponte requiring the jurors to unanimously agree on which conspiracy the defendant participated in. See Payseno, 782 F.2d at 834-37; Gilley, 836 F.2d at 1211-13.
The government argues (for the first time on appeal)
The government offered no evidence that Boucher ever formed an agreement with Kanyid or that Boucher shared some common purpose with Kanyid (the two did not know each other and never met). It
To show a conspiracy, "the circumstances must lead to an inference that some form of overall agreement exists," and that "each defendant knew or had reason to know of the scope of the conspiracy and ... reason to believe that their own benefits were dependent upon the success of the entire venture." Kenny, 645 F.2d at 1335 (citations omitted). Here, there is no evidence of any "common purpose of a single enterprise" linking Kanyid and Boucher in any collective agreement with the alleged rim, which consisted of separate buyers. Canella v. United States, 157 F.2d 470, 476 (9th Cir.1946) (quoting Kotteakos v. United States, 328 U.S. 750, 769, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946)).
The district court's failure to give a specific unanimity instruction sua sponte was plain error warranting reversal because it created a genuine possibility of jury confusion and of a nonunanimous verdict. We therefore