United States v. Rafael Rojas-Reyes

In the United States Court of Appeals For the Seventh Circuit ____________________ Nos. 19-1074, 19-1110, 19-1126, & 19-1188 UNITED STATES OF AMERICA, Plaintiff-Appellee, v. HECTOR SAUL CASTRO-AGUIRRE, et al., Defendants-Appellants. ____________________ Appeals from the United States District Court for the Southern District of Indiana, Indianapolis Division. No. 1:16-cr-00123-TWP-DML — Tanya Walton Pratt, Judge. ____________________ ARGUED SEPTEMBER 15, 2020 — DECIDED DECEMBER 28, 2020 ____________________ Before FLAUM, ROVNER, and WOOD, Circuit Judges. WOOD, Circuit Judge. Illegal drugs often do not originate in the community where they are consumed, and so the drug business—like its legitimate counterparts—commonly in- cludes a complex distribution network. That was true of the arrangement before us, which involved large quantities of co- caine and methamphetamine that moved throughout the southwestern and northeastern United States. Eventually the 2 Nos. 19-1074 et al. government caught up with the participants. Among those it indicted were four who chose to go to trial: Jose Manuel Car- rillo-Tremillo, Hector Saul Castro-Aguirre, John Ramirez- Prado, and Rafael Rojas-Reyes. These four have joined in the present appeal, in which they challenge rulings the district court made at the guilt phase, as well as some of its sentencing decisions. For the most part, we find no reversible error. The only exception is Carrillo-Tremillo’s conviction for conspiracy to launder money, which we set aside. I A The defendants before us all played active roles in a cross- country drug organization: Castro-Aguirre served as the head of operations; Ramirez-Prado handled logistics, including providing cars and hotels for distributors and couriers; Rojas- Reyes coordinated sales in Indianapolis; and Carrillo-Tre- millo conducted sales in the northeast. To set the stage, we provide a brief overview of their activities from 2015 to 2016. Castro-Aguirre coordinated shipments of methampheta- mine—usually 30 pounds apiece—from a trailer park in Tuc- son, Arizona, to Avon, Indiana (a suburb of Indianapolis). Ramirez-Prado rented SUVs and booked hotels for couriers along the route. Once the packages reached their destination, Rojas-Reyes took over and sold the methamphetamine in In- dianapolis. Castro-Aguirre also handled cocaine sales in New Jersey and New York. Ramirez-Prado and other couriers transported bulk quantities of cocaine from Arizona and Indianapolis to New York. There, Castro-Aguirre fronted the drugs to Car- rillo-Tremillo, who would sell them and remit the proceeds to Nos. 19-1074 et al. 3 Castro-Aguirre. In much the same way, Castro-Aguirre fur- nished Carrillo-Tremillo with anywhere from 10 to 100 kilo- grams of cocaine on credit for distribution in Reading, Penn- sylvania. Castro-Aguirre arranged for couriers to pick up the proceeds from his various sellers and deliver the cash to him in Arizona. The final trip to Reading proved to be the downfall of the organization. The deal started out routinely, when Castro- Aguirre fronted 100 kilograms of cocaine to Carrillo-Tremillo. The drugs made it to Reading, but alert law enforcement of- ficers caught up with the couriers and stopped them near the Illinois-Missouri border. There the agents seized $2,400,000. That is just the bare outline of the operation. It had many moving parts, only some of which are important to this ap- peal. One key event involved the kidnapping and murder in 2016 of Angel Barrios-Moreno, who supplied the operation with drugs that he transported across the Mexican-U.S. bor- der. The evidence indicated that leaders of the infamous Sina- loa Cartel had ordered the hit. They did so because Barrios- Moreno failed to pay a debt to the cartel after law enforcement officials seized a major drug shipment. The cartel was not for- giving: during one of Barrios-Moreno’s trips in Mexico from Nogales to Sinaloa, members of the Sinaloa Cartel kidnapped Barrios-Moreno and two others—his nephew Adrian Barrios- Moreno and a friend Luis-Carlos Cebrero-Alvarez—and de- manded a $500,000 ransom. When Castro-Aguirre learned of the kidnapping, he im- mediately started to raise money for the ransom. He sent cou- riers to pick up cash and drugs from Rojas-Reyes and Ramirez-Prado. Once he had collected $250,000, Castro- 4 Nos. 19-1074 et al. Aguirre sent couriers to make a partial payment on the ran- som to a cartel member in New York. Despite these efforts, the cartel ultimately killed Barrios-Moreno, along with the other two men it had seized. B In January 2018, the government indicted twelve defend- ants. The four now before us proceeded to trial. The following chart provides the counts, charges, and verdicts for each one: Carrillo-Tre- Rojas-Reyes Ramirez- Aguirre Castro- Prado millo # Charge (Statute) Conspiracy to Distribute 500+ 1 grams of methamphetamine G1 G G G and/or 5+ kilograms of cocaine (21 U.S.C. §§ 841(a)(1) & 846) Engaging in a Continuing 2 Criminal Enterprise (21 U.S.C. G § 848(a), (b)(1)) Engaging in a Continuing 3 Criminal Enterprise (21 U.S.C. G § 848(a), (b)(1)) 4 Conspiracy to Launder Money G G G G (18 U.S.C. § 1956(h)) 1 The letter “G” indicates guilty. Nos. 19-1074 et al. 5 Distributing 500+ Grams of 5 Methamphetamine Mixture (21 G U.S.C. § 841(a)(1)) Distributing 50+ Grams of 6 Methamphetamine Mixture G (21 U.S.C. § 841(a)(1)) Possession of 50+ Grams of 8 Methamphetamine Mixture (21 G U.S.C. § 841(a)(1)) The jury found each defendant guilty of conspiracy to dis- tribute the controlled substances and conspiracy to launder money. As the chart indicates, Castro-Aguirre and Rojas- Reyes were also convicted on several additional charges. The district court sentenced all the defendants to lengthy terms in prison. The defendants have appealed from their convictions, their sentences, or both. We have organized their contentions as follows: Section II addresses evidentiary rulings, Section III resolves challenges to the underlying convictions, and Sec- tion IV briefly discusses the sentencing arguments. II. Evidentiary Rulings The defendants raise two primary evidentiary points: the first concerns the district court’s decision to admit evidence of cell-site locations collected pursuant to the Stored Communi- cations Act (“the Act”); and the second relates to the court’s decision to allow the jury to hear about the Barrios-Moreno kidnapping and murder. The standard of review for both points is generally deferential. To the extent the court made legal determinations, our review is de novo, United States v. 6 Nos. 19-1074 et al. Figueroa-Espana, 511 F.3d 696, 701 (7th Cir. 2007), but we re- view decisions whether to admit or exclude evidence only for abuse of discretion, United States v. Johnson, 916 F.3d 579, 586– 87 (7th Cir. 2019) (quoting United States v. Causey, 748 F.3d 310, 316 (7th Cir. 2014)). A. Cell-Site Location Information The defendants argue that the district court erred by deny- ing their motion to suppress cell-site location information (commonly called CSLI) that the government obtained using the procedures set out in the Act. They contend that this in- formation was collected in violation of the Fourth Amend- ment, because, midway through the trial, the Supreme Court ruled that the government must obtain a warrant if it wishes to obtain some CSLI. See Carpenter v. United States, 138 S. Ct. 2206 (2018). “CSLI is location information generated by cellular phone providers that indicates which cell tower a particular phone was communicating with when a communication was made.” Orin S. Kerr, The Effect of Legislation on Fourth Amendment Pro- tection, 115 MICH. L. REV. 1117, 1128 (2017). Any cell phone with a functioning battery regularly communicates with cell towers. The phone leaves behind a trail, which is refreshed “as frequently as several times per minute,” showing its rough location (within 50 meters or so of the actual spot). United States v. Curtis, 901 F.3d 846, 847 (7th Cir. 2018) (quot- ing Carpenter, 138 S. Ct. at 2219–21). With this information, the government has a near-perfect record of the phone’s location (and thus, presumptively, that of the user). Carpenter, 138 S. Ct. at 2217–21. Nos. 19-1074 et al. 7 The Act permits the government to obtain a court order authorizing the collection of “non-content information.” 18 U.S.C. § 2703(c). Before Carpenter, the government could ei- ther seek a warrant for this information or it could “obtain[] a court order for such disclosure under subsection (d) of this section.” Id. at § 2703(c)(A), (B). Subsection (d) specified that a court order could issue “only if the governmental entity offers specific and articulable facts showing that there are reasona- ble grounds to believe that ... the records or other information sought, are relevant and material to an ongoing criminal in- vestigation.” Id. at § 2703(d). That standard is significantly lower than the probable-cause requirement for a warrant. See, e.g., Terry v. Ohio, 392 U.S. 1 (1968). The defendants here do not contend that the government failed to present “specific and articulable facts” for its order. Instead, relying on Carpen- ter, they contend that they have a reasonable expectation of privacy—protected by the Fourth Amendment—in their CSLI, the government failed to obtain a warrant before it seized those records, and so suppression is called for. The district court did not disagree with the need for a war- rant, nor could it after Carpenter. But it recognized that this is not the end of the inquiry. The dispositive question, it held, was whether the government acted in good faith when, in- stead of securing a warrant, it relied on the court order. See United States v. Leon, 468 U.S. 897, 919 (1984); Heien v. North Carolina, 574 U.S. 54 (2014). If it acted in good faith, then sup- pression is not required. The district court found the good- faith rule applicable to these cases: at the time that the gov- ernment obtained the court order in July 2016, no court of ap- peals had concluded that CSLI could be collected only with a warrant, and the government had relied on the procedures set 8 Nos. 19-1074 et al. forth in the Act. See, e.g., United States v. Thousand, 558 F. App’x 666, 670 (7th Cir. 2014). The defendants present a bold theory in an effort to obtain suppression of the CSLI. They contend that Carpenter itself was a deviation from precedent. At the time, courts consist- ently held that a person has no reasonable expectation of pri- vacy in information held by third parties; they relied on United States v. Miller, 425 U.S. 435 (1976) (no reasonable ex- pectation of privacy in bank-held financial records), and Smith v. Maryland, 442 U.S. 735 (1979) (no reasonable expectation of privacy in phone-provider-held call logs). Carpenter held for the first time that this third-party rule had its limits and in particular did not apply to cell phones. 138 S. Ct. at 2217. De- fendants contend that the rationale of Carpenter—namely, that CSLI is uniquely expansive and invasive—also calls for us to override the good-faith exception and order suppression here. But it is one thing to speak of the scope of a rule, such as the third-party doctrine, and quite another to ask whether the purpose of the rule will be served by its application to good- faith missteps. We recognized just two years ago that the mis- taken use of the Act’s procedures, rather than a warrant, to collect CSLI is eligible for the good-faith exception to the or- dinary practice of suppression. See Curtis, 901 F.3d at 848 (cit- ing Krull, 480 U.S. at 349); see also Herring v. United States, 555 U.S. 135, 140 (2009). We recognized the deterrent role the ex- clusionary rule is designed to play, but we found no evidence that “legislators are inclined to subvert their oaths and the Fourth Amendment” when passing laws such as the Act. Cur- tis, 901 F.3d at 849 (quoting Krull, 480 U.S. at 349). Nor did we see any reason why the application of the exclusionary rule to Nos. 19-1074 et al. 9 evidence seized before Carpenter was decided, in good-faith reliance on a court order pursuant to the Act, would provide additional deterrence. Id. We are not inclined to revisit Curtis. We thus find that the CSLI obtained in this case was properly admitted, because the government, following the procedures set forth in the Act, gathered it in good faith. B. Motion in Limine We turn next to the defendants’ challenge to the district court’s denial of their motion in limine to suppress evidence of their gang membership and evidence that the Sinaloa Car- tel was behind Barrios-Moreno’s kidnapping and murder. Once again, they face an uphill battle, as we defer to the dis- trict court’s assessments both of relevance and potential inap- propriate prejudice. United States v. Johnson, 916 F.3d 579, 588 (7th Cir. 2019). 1. Gang Membership The defendants first contend that their relationship to the Sinaloa Cartel was irrelevant to show that they participated in the charged offenses. At trial, Yesenia Samaniego, an admitted participant in their drug-distribution network, testified that Castro-Aguirre indicated to her that he was involved with a cartel. Her husband, Jesus Samaniego, corroborated this ac- count, explaining that Castro-Aguirre “basically” told him that these defendants were working for the Sinaloa Cartel. In their motion in limine, the defendants argued that the proposed evidence of gang affiliation was inadmissible to show involvement in the conspiracies at issue. The district 10 Nos. 19-1074 et al. court, however, rejected their argument and allowed its ad- mission. Its decision to do so was not an abuse of discretion. We have long held that “‘[e]vidence of gang affiliation is admissible in cases in which it is relevant to demonstrate the existence of a joint venture or conspiracy and a relationship among its members’ and each defendant’s knowledge of and participation in the drug conspiracy.” United States v. Alviar, 573 F.3d 526, 536 (7th Cir. 2009) (alteration in original) (quot- ing United States v. Westbrook, 125 F.3d 996, 1007 (7th Cir. 1997)). The critical distinction is between use of gang affilia- tion for the limited purpose of showing the existence of the conspiracy and the connections among the various actors (which is permissible), and use of gang affiliation to show ac- tual involvement in the conspiracy (which is not). See United States v. Suggs, 374 F.3d 508, 517 (7th Cir. 2004). The district court here respected that line. The evidence of the defendants’ loose affiliation to the Sinaloa Cartel was relevant to show their association and the existence of a drug-distribution con- spiracy. 2. Kidnapping and Murder of Barrios-Moreno In contrast, we are troubled by the district court’s decision to admit the evidence of the kidnapping and murder of Bar- rios-Moreno and his colleagues. This was explosive material, and it had little to do with the conduct for which the defend- ants were being tried. At trial, the government elicited testi- mony from Yesenia that Angel Barrios-Moreno, a supplier for the defendants, was kidnapped by the Sinaloa Cartel for a debt he had accrued after the government seized drugs that had been bound for him. The cartel demanded a $500,000 ran- som. The government’s theory was that this ransom moti- vated certain of the defendants’ actions in the ensuing days. Nos. 19-1074 et al. 11 The ransom, prosecutors urged, put pressure on the defend- ants to collect cash to free Barrios-Moreno. Yesenia testified that they managed to raise $250,000 in relatively short order. But the government did not link that effort to the operation of the drug organization that lay at the heart of this case. Even less relevant was the government’s decision to try to supplement the ransom testimony with copies of a news arti- cle that included pictures of the three dead men: Barrios- Moreno and his two colleagues. The government contended that the jury needed to understand that Barrios-Moreno was brutally murdered by cartel members who were not part of this case, because the jury would otherwise be at a loss to ex- plain why the defendants stopped using Barrios-Moreno as a drug source and found a new seller. But the government never explained why the identity of the drug source mattered to the charges in this case. The defendants requested that the district court allow them to stipulate that Barrios-Moreno was killed, and they asked that the photos not be admitted because they were gruesome and had no effect apart from “inflam[ing] the jury.” The court compromised, keeping out one of the photos and admitting the other two. In so ruling, it commented that “there’s nothing gruesome. It’s just three bodies.” It then gave a limiting instruction to the jurors, telling them that the evi- dence was admissible “solely as background material to es- tablish the effect that the kidnapping and murder of these in- dividuals in Mexico had on the alleged drug trafficking and money laundering activities of the defendants in the United States.” Yesenia was not the only witness to address the kidnap- ping and homicide. The government also asked a DEA agent, 12 Nos. 19-1074 et al. Matthew Holbrook, to review the same news articles. During the direct examination, the government placed the photo- graphs of the three dead bodies on display, as well as multiple photos of Barrios-Moreno’s gravesite. The government even asked the district court to permit it to circulate physical copies of the photographs of Barrios-Moreno’s gravesite to the jurors because they “were a little fuzzy when you place them on the projector.” The district court obliged. The government asked two other witnesses to add context to Barrios-Moreno’s death. First, Jesus Samaniego testified that Barrios-Moreno was kidnapped because he owed a debt to the cartel. This was supposedly to provide context for the Castro-Aguirre group’s need for a new supplier after Barrios- Moreno was killed. The government asked Jesus if he knew who killed Barrios-Moreno, and Jesus responded that the son of Joaquin (“El Chapo”) Guzmán—the head of the Sinaloa Cartel—was behind the killing. Jesus also provided details about the efforts to repay the ransom. Shortly after Jesus testified, and well after the district court’s limiting instruction to the jurors specifically informing them that none of the defendants participated in these mur- ders, the district court informed counsel that multiple ju- rors—fearing for their safety—asked whether the defendants had access to their personal information. The district court brushed their concerns aside, suggesting that the fear likely arose “because weʹve talked about the cartel and things like that.” Over the defendants’ renewed objection, it found that the evidence was relevant and not prejudicial, and it ex- pressed the view that “this type of question is normal in al- most every jury trial.” Nos. 19-1074 et al. 13 Even then, the testimony about the murders did not stop. The government next presented Julio Cesar Cebrero-Alvarez, a taxi driver in Nogales (Mexico) who had driven Castro- Aguirre and Barrios-Moreno. Cebrero-Alvarez stated that he was present for the events that led to Barrios-Moreno’s mur- der and retold the story in great detail. He told the jury that he took a trip to Sinaloa from Nogales with several people: Barrios-Moreno, Angel’s nephew Adrian, and Cebrero-Alva- rez’s brother Luis Carlos. At the outset he was in a car with Barrios-Moreno, while Luis Carlos and Adrian rode in a sep- arate car. At some point Barrios-Moreno switched to the other car, and Cebrero-Alvarez, now driving alone, periodically called the other car to check in. Cebrero-Alvarez became con- cerned when Barrios-Moreno stopped answering. This prompted him to visit Barrios-Moreno’s wife, who told him that Barrios-Moreno’s sister Elvia had just called and in- formed her that her husband and the others had been kid- napped. Cebrero-Alvarez concluded by reporting that the po- lice informed him that they had found the bodies and asked him to identify them, but that he turned around when he was within 10 kilometers of the meeting place because he was “afraid maybe it was a trick” and that he would be killed too. We struggle to see the relevance of this highly prejudicial evidence, particularly in light of the jurors’ expressed concern about their safety. No one ever accused the defendants who were on trial with any of these murders, or even with violent histories. More troubling still, the jurors voiced their concern after receiving the limiting instruction designed to mitigate any undue prejudice from the evidence of Barrios-Moreno’s kidnapping and murder. We see no reason why the govern- ment needed to elicit testimony from at least four witnesses explaining in great detail how these men were kidnapped and 14 Nos. 19-1074 et al. murdered, why it needed to introduce multiple photographs of the deceased bodies or the gravesite, or why it would not have been sufficient for the court simply to tell the jury that Barrios-Moreno had left the conspiracy, or disappeared, or was killed. We are not impressed with the government’s arguments that the jury would have been left with an evidentiary gap without this extensive evidence. The government contends that the evidence explains certain travel to pay off the ransom and that the jurors would have been unable to figure out why the defendants would seek a new source and recruit new cou- riers without it. It relies on a 25-year-old case from this court in which we held that the government could introduce evi- dence of a triple murder in which no defendant participated, committed by the Zapata organization, because that incident explained why a participant “ended up cooperating with the government after being an active, loyal member of the con- spiracy himself.” United States v. Pulido, 69 F.3d 192, 200 (7th Cir. 1995). We have no quarrel with the result in Pulido, but it was based on significantly different facts. There, the government showed how the murders of the three men prompted their as- sociate to cooperate with the government. Here, no such link exists among the contested evidence, the defendants, and the charges. The government contended in the district court that any prejudice from this evidence is mitigated because Yesenia Sa- maniego testified that she feared for her life if she returned to Mexico, but it did not raise this argument in its appellate brief. In any event, this theory is not particularly persuasive. Only after the district court denied the defendants’ motion in limine Nos. 19-1074 et al. 15 did the defendants ask Yesenia whether she was afraid of be- ing killed for her testimony by members of the Sinaloa Cartel if she returned to Mexico, or if the government offered to “get [her] a visa to stay in the United States” in exchange for her testimony. Yesenia answered that she was afraid of Sinaloa members in Mexico, not any of the defendants, who are in the United States. In light of the tenuous connection between this evidence and the case before us and the prejudice reflected in the jurors’ concerns for their safety, we find that the district court abused its discretion by admitting this evidence. This finding is not, however, an automatic ticket to rever- sal. We also must decide whether the error was harmless. United States v. Vasquez, 635 F.3d 889, 898 (7th Cir. 2011). “The test for harmless error is whether, in the mind of the average juror, the prosecution’s case would have been ‘significantly less persuasive’ had the improper evidence been excluded.” Id. (quoting United States v. Emerson, 501 F.3d 804, 813 (7th Cir. 2007)). Given the overwhelming nature of the evidence presented at trial regarding the defendants’ guilt, we cannot say that the jury would have found the prosecutor’s case significantly less persuasive had the court kept out the evidence relating to Bar- rios-Moreno. Yesenia and Jesus Samaniego provided uncon- troverted testimony connecting Castro-Aguirre, Ramirez- Prado, and Rojas-Reyes to multiple large drug transactions and identifying Carrillo-Tremillo as a major distributor. These two testified about each of the trips from Arizona to In- dianapolis, as well as the trips from Arizona and Indianapolis to New York, New Jersey, and Pennsylvania. The government presented travel and financial records tying Carrillo-Tremillo 16 Nos. 19-1074 et al. and Ramirez-Prado to the major drug routes the conspirators used, records from phones the DEA lawfully seized contain- ing photographs and evidence of other connections among the defendants, drugs and cash seized from the defendants’ cars and homes, and cell-site location data corroborating the other evidence showing the defendants’ proximity during the relevant events. We therefore find that the evidentiary error we have addressed is harmless. See FED. R. EVID. 103(a); FED. R. CRIM. P. 52(a). III. Carrillo-Tremillo’s Convictions We next address Carrillo-Tremillo’s appeal of the district court’s denial of his motion under Federal Rule of Criminal Procedure 29 for acquittal on Count 1 (conspiracy to distrib- ute) and Count 4 (conspiracy to launder money). We review the denial of a motion for judgment of acquittal de novo, ask- ing only “whether evidence exists from which any rational trier of fact could have found the essential elements of a crime beyond a reasonable doubt.” United States v. Johnson, 874 F.3d 990, 998 (7th Cir. 2017). We conclude that the district court correctly denied the motion to dismiss with respect to Count 1, because there was sufficient evidence to prove a con- spiracy, rather than a simple buyer-seller arrangement. In contrast, we reverse the district court’s denial of the motion to dismiss with respect to Count 4, because we can find no evi- dence that would permit a rational trier of fact to connect Car- rillo-Tremillo to a conspiracy to launder money. A. Conspiracy to Distribute To prove a conspiracy to distribute a controlled substance, the government must show “that (1) two or more people agreed to commit an unlawful act, and (2) the defendant Nos. 19-1074 et al. 17 knowingly and intentionally joined in the agreement.” United States v. Johnson, 592 F.3d 749, 754 (7th Cir. 2010) (citing United States v. Rollins, 544 F.3d 820, 835 (7th Cir. 2008)). We have identified five commonly occurring facts that permit the in- ference of a conspiracy to distribute: “[1] sales on credit or consignment, [2] an agreement to look for other customers, [3] a payment of commission on sales, [4] an indication that one party advised the other on the conduct of the other’s business, or [5] an agreement to warn of future threats to each other’s business stemming from competitors or law-enforcement au- thorities.” Johnson, 592 F.3d at 755–56. Among these, fronting (i.e. supplying on credit, for later reimbursement), has become the most telling factor. Id. Fronting, which occurs only if the supplier and the next- level distributor have a shared economic stake in the ultimate transaction, is a sign that we are looking at a conspiracy, not two independent actors pursuing their own ends. A typical buyer-seller relationship—even one involving a long history of transactions—involves one or more on-the-spot exchanges of cash for drugs. United States v. Lechuga, 994 F.2d 346, 349 (7th Cir. 1993) (en banc) (lead opinion); United States v. Colon, 549 F.3d 565, 567 (7th Cir. 2008). In contrast, the prosecution can establish a conspiracy by proving that a larger agreement, including a sale on credit, links the initial sale to the down- stream distribution efforts. United States v. Avila, 557 F.3d 809, 816 (7th Cir. 2009). This type of evidence tends to show a mu- tual interest in the success of each party’s operations. Carrillo-Tremillo contends that the district court improp- erly denied his motion for judgment of acquittal on the drug conspiracy charge because, as he sees it, the government pro- vided only circumstantial evidence of a simple buyer-seller 18 Nos. 19-1074 et al. relationship. But first, there is nothing wrong with circum- stantial evidence. And second, there was more to the govern- ment’s case than he suggests. At trial, Yesenia Samaniego testified that she saw Carrillo- Tremillo enter a hotel room with a small suitcase. She was in the room with Castro-Aguirre. She then left the room and waited outside for ten minutes, during which time she did not see anyone else enter or leave the room. She returned to the room after Carrillo-Tremillo left and saw a large pile of money that was not there before on the bed. She concluded by stating that Castro-Aguirre told her that Carrillo-Tremillo was a courier for their cocaine distribution. Jesus Samaniego corroborated his wife’s account. He tes- tified that he watched Carrillo-Tremillo walk into a hotel in Scottsdale, Arizona. Carrillo-Tremillo was carrying the same duffle bag that Castro-Aguirre had filled with cocaine, as Je- sus watched. Jesus also reported seeing a large pile of money on the bed in the hotel room after Carrillo-Tremillo visited Castro-Aguirre there. Like Yesenia, Jesus testified that Castro- Aguirre specifically told him that Carrillo-Tremillo gave him the money. The government also presented evidence, includ- ing airline and hotel receipts from Arizona, Pennsylvania, New York, and New Jersey, linking Carrillo-Tremillo to the distribution routes. Carrillo-Tremillo attacks the evidence provided by the Sa- maniegos as unreliable, because neither of them personally saw him deliver the money. He also contends that without more context, their testimonies can prove only that he had a buyer-seller relationship with Castro-Aguirre, not a conspira- torial one. He believes that the bank and travel records do not help either. He relies on our decision in Colon, where we held Nos. 19-1074 et al. 19 that the government merely “describe[d] a routine buyer- seller relationship” when it presented evidence of routine cash purchases of distribution quantities of drugs. 549 F.3d at 567. Carrillo-Tremillo’s reliance on Colon is mistaken. There we held that the government would have successfully proven a conspiracy if it had provided evidence that the supplier “‘fronted’ cocaine to the defendant (a factor mentioned in al- most all the cases) rather than being paid in cash at the time of sale. With fronting, the seller becomes the buyer’s creditor, adding a dimension to the relationship that goes beyond a spot sale for cash.” Id. at 569. In the present case, the government did just what we had called for in Colon. The jury was entitled to conclude that the Samaniegos’ testimony and the physical evidence connecting Carrillo-Tremillo to the defendants were enough to prove fronting, which in turn can be used to prove a conspiracy. Carrillo-Tremillo’s theory is not completely improbable, but it was the jury’s job to decide what happened, and this jury drew inferences favorable to the government. We do not ask whether a rational trier of fact could agree with the defend- ant’s construction of the evidence. Rather, we ask whether a rational trier of fact could find that Carrillo-Tremillo partici- pated in the conspiracy to distribute. We conclude that it could, and so we affirm the district court’s denial of Carrillo- Tremillo’s motion for judgment of acquittal on Count 1. B. Conspiracy to Launder Money A charge of conspiracy to launder money requires proof that the defendant knowingly was involved with “at least one other person [who has] agreed with him to commit” an act of 20 Nos. 19-1074 et al. money laundering, United States v. McBride, 724 F.3d 754, 756 (7th Cir. 2013), and that the defendant “knew the proceeds used to further the scheme were derived from illegal activity,” United States v. Arthur, 582 F.3d 713, 718 (7th Cir. 2009). Proof of the underlying offense of money laundering requires “that a rational trier of fact could have concluded from the record that [the defendant] knowingly used the proceeds from a specified unlawful activity in financial transactions that [1] were intended to promote the continuation of the unlawful activity, or [2] were designed to conceal or disguise the pro- ceeds of the unlawful activity.” Id. (quoting United States v. Turner, 400 F.3d 491, 496 (7th Cir. 2005); 18 U.S.C. § 1956(a)(1), (h)). There must be evidence of activity that is separate from the underlying predicate offense that generates proceeds to be laundered, in order successfully to charge a defendant with involvement in a conspiracy to commit money laundering. The point of the money-laundering offense is to “criminalize[] a transaction in proceeds, not the transaction that creates the proceeds.” United States v. Mankarious, 151 F.3d 694, 705 (7th Cir. 1998) (emphasis added). Even when “the predicate of- fense and the conspiracy to launder money cannot be sepa- rated with surgical precision, they are clearly distinct.” United States v. Diamond, 378 F.3d 720, 727 (7th Cir. 2004). The evi- dence against a defendant will not support a conviction for conspiracy to commit money laundering when it demon- strates involvement only in a conspiracy that generated pro- ceeds but does not demonstrate involvement in any subse- quent conspiracy to launder those proceeds. We pause here to clarify the kinds of activity that signals involvement in a conspiracy to commit money laundering. Nos. 19-1074 et al. 21 We have never provided an exhaustive list of the transactions that give rise to a money laundering conspiracy, and we de- cline to do so now. But we have explained that “certain types of transactions may be indicative of a design to conceal,” in- cluding, for example, “transactions surrounded in unusual secrecy, structured transactions, depositing ill-gotten funds into another’s bank accounts, using third parties to conceal the real owner, or engaging in unusual financial moves which culminate in a transaction.” Turner, 400 F.3d at 497. In Dia- mond, we upheld a conviction for conspiracy to launder money where the evidence showed that the defendant en- dorsed checks that a coconspirator deposited into his personal accounts, repeatedly purchased cashier’s checks with fund money (often to evade IRS reporting), withdrew fund money in cash because cash is untraceable, wired money to an off- shore bank account to purchase an oil refinery, and wired money to her own offshore bank account. Diamond, 378 F.3d at 727–28. In contrast, in United States v. Malone, the defendant received payment for sending drivers with vans full of drugs stuffed in band equipment, so that he could arrange for the delivery of the drugs, receive cash payments for them, and return the funds to a drug-distribution group. 484 F.3d 916, 921–22 (7th Cir. 2007). We held that the evidence was insuffi- cient to support a conviction for conspiracy to launder money because the defendant was a mere conduit; he did not partic- ipate in any transactions designed to conceal the proceeds that were intended to promote the continuation of the unlaw- ful activity. Id. The government argues that Carrillo-Tremillo is guilty of participating in the conspiracy to launder money because he knowingly conducted financial transactions with the intent to 22 Nos. 19-1074 et al. promote the continuation of the unlawful activity. Its evi- dence is scant. All we see is that Carrillo-Tremillo paid Castro- Aguirre the proceeds he collected from drugs that had been fronted to him. The government argues that this payment ad- vanced the conspiracy because it kept Carrillo-Tremillo’s credit good with Castro-Aguirre, thereby ensuring a contin- ual flow of additional supplies. This evidence falls short. The evidence of a conspiracy to commit money laundering cannot be identical to the evidence of the underlying offense that generated the proceeds that need laundering. As we have noted in connection with Count 1, the government proved that Carrillo-Tremillo par- ticipated in a conspiracy to distribute controlled substances: Castro-Aguirre gave Carrillo-Tremillo drugs on credit, Car- rillo-Tremillo distributed those drugs, and Carrillo-Tremillo paid Castro-Aguirre back. It is precisely this element of front- ing that differentiates these transactions from a simple buyer- seller relationship. Now the government wants to say that the same evidence of fronting and reimbursement qualifies as money laundering, in the sense of financial transactions in- tended to promote the unlawful activity. What is missing here is any kind of transaction in the drug proceeds, as opposed to the actions that generate the pro- ceeds. The government did not prove that Carrillo-Tremillo conducted a financial transaction with those proceeds over and beyond paying his debt for the fronted drugs. In addition, it needed to point to an independent transaction that was (1) intended to promote the continuation of the unlawful activity or (2) designed to conceal or disguise the proceeds. Evidence of fronting, standing alone, is insufficient for purposes of this Count. The government’s theory would erase the distinction Nos. 19-1074 et al. 23 between the conspiracy to distribute controlled substances and money laundering. We have not accepted this approach, nor do we see any sign that Congress meant to create this per- fect overlap. We conclude, therefore, that the district court erred when it denied Carrillo-Tremillo’s motion for judgment of acquittal with respect to the conspiracy to commit money laundering. IV. Sentences We turn finally to the defendants’ challenges to their sen- tences. The challenges fall within two categories. First, Rojas- Reyes attacks his concurrent life sentences on constitutional grounds. Second, Ramirez-Prado and Carrillo-Tremillo argue that their below-guideline sentences are substantively unrea- sonable. A. Rojas-Reyes We review constitutional challenges, including challenges to the constitutionality of sentences, de novo. United States v. Figueroa-Espana, 511 F.3d 696, 705 (7th Cir. 2007); United States v. Nagel, 559 F.3d 756, 759 (7th Cir. 2009). The Eighth Amend- ment’s Cruel and Unusual Punishment Clause “prohibits sen- tences that are grossly disproportionate to the crime commit- ted.” Nagel, 559 F.3d at 762. Proving that a life sentence is grossly disproportionate to the crime committed is a difficult task, United States v. Brucker, 646 F.3d 1012, 1019 (7th Cir. 2011), because that standard is met only in “exceedingly rare” cases, Harmelin v. Michigan, 501 U.S. 957, 963 (1991); see also United States v. Gross, 437 F.3d 691, 692 (7th Cir. 2006). Three factors shed light on the question: “(i) the gravity of the of- fense and the harshness of the penalty; (ii) the sentences im- posed on other criminals in the same jurisdiction; and (iii) the 24 Nos. 19-1074 et al. sentences imposed for commission of the same crime in other jurisdictions.” Nagel, 559 F.3d at 762 (quoting Solem v. Helm, 463 U.S. 277, 292 (1983)). As our earlier chart illustrated, the jury found Rojas-Reyes guilty on six counts. The district court sentenced him to four concurrent life sentences for Counts 1, 3, 5, and 8. Count 3 car- ried a mandatory life sentence, and the district court applied the maximum sentence of life for Counts 1, 5, and 8. The court then imposed concurrent sentences of 240 months in prison for Count 4 and 480 months in prison for Count 6. Rojas- Reyes argues that his four concurrent life sentences are grossly disproportionate to the offenses charged because he was only 38 years old at the time of sentences, had no prior criminal history, and was sentenced for only non-violent of- fenses. Rojas-Reyes’s constitutional challenges are without merit. We have “repeatedly upheld mandatory life sentences im- posed pursuant to § 841(b)(1)(A) against Eighth Amendment challenges.” United States v. Ousley, 698 F.3d 972, 975 (7th Cir. 2012) (finding that mandatory life sentences for possession of more than 50 grams of crack cocaine with the intent to distrib- ute with at least two prior drug felonies does not constitute cruel and unusual punishment); see also United States v. Car- raway, 612 F.3d 642, 644–46 (7th Cir. 2010) (same); United States v. Strahan, 565 F.3d 1047, 1052–53 (7th Cir. 2009) (rejecting Eighth Amendment challenges to statutorily prescribed man- datory-minimum life sentences). The Supreme Court has like- wise explained that even mandatory minimums without con- sideration for “mitigating factors such as … the fact that [a defendant] had no prior felony convictions” do not violate the Nos. 19-1074 et al. 25 Eighth Amendment because “they are not unusual in the con- stitutional sense.” Harmelin, 501 U.S. at 994–95. Our sister cir- cuits have also upheld multiple concurrent life sentences against Eighth Amendment challenges. See Virgin Islands v. Gereau, 592 F.2d 192 (3d Cir. 1979); United States v. Anderson, 561 F.2d 1301 (9th Cir. 1977); United States v. Williams, 576 F.3d 1149 (10th Cir. 2009); United States v. Rolon, 511 F. App’x 883 (11th Cir. 2013). Our task is not to determine whether we, sitting in the seat of the district court, would have given a person in his late thir- ties with no prior record multiple concurrent life sentences for non-violent drug offenses. Instead, we ask only whether the district court’s sentencing decisions violate the Eighth Amendment. They did not. B. Ramirez-Prado Despite the fact that Ramirez-Prado received a sentence below the level suggested by the advisory Federal Sentencing Guidelines, he argues that it was unreasonably long. This is a hard case to make. When, as here, there is no procedural chal- lenge to a sentence, we ask only whether the sentence selected by the district court is substantively reasonable. United States v. Warner, 792 F.3d 847, 856 (7th Cir. 2015). In this connection, we must be “mindful that substantive reasonableness occu- pies a range, not a point, and that the sentencing judge is in the best position to apply the § 3553(a) factors to the individ- ual defendant.” Id. (quotation marks and citations omitted). As a result, we evaluate sentencing challenges for abuse of discretion and recognize that “sentence[s] within [guideline] range [are] presumptively reasonable.” United States v. Wal- lace, 531 F.3d 504, 507 (7th Cir. 2008). 26 Nos. 19-1074 et al. After the jury returned a verdict of guilty against Ramirez- Prado, the district court held a sentencing hearing for him. When all was said and done the district court calculated his recommended guideline range for Count 1 as 324 to 405 months; for Count 4, his “range” was 120 months.2 The court then applied a seven-year reduction below the low end of the Count 1 range, sentencing Ramirez-Prado to 240 months for Count 1 and a concurrent sentence of 120 months on Count 4. Ramirez-Prado contends that his sentence should have been lower yet—no more than a total of 120 months. We take this as an admission that 120 months was reasonable for Count 4, and so we focus on Count 1. Emphasizing that he had no prior criminal history, Ramirez-Prado contends that the sentence was improper because other defendants who confessed to their involvement in the conspiracy and testified at trial garnered lower sentences. Yesenia Samaniego received 60 months, and Jesus Samaniego, who had a higher criminal history, received 230 months. Ramirez-Prado argues that 18 U.S.C. § 3553(a)(6) requires the district court to consider sen- tencing disparities among defendants with similar records. He contends that the district court failed to do so. We are not persuaded. “Unlike the sentencing judge, we may presume on appeal that a within-guidelines sentence is reasonable.” United States v. Warner, 792 F.3d 847, 855–56 (7th 2We read the district court’s determination that the recommended sentencing range was a 120-month point as an acknowledgement of a stat- utory maximum. But 18 U.S.C. § 1956(a), (h) prescribes a 20-year (240- month) maximum, not 120 months (or 10 years). The government, how- ever, does not challenge the district court’s sentencing decision or take a cross-appeal, and so we evaluate the sentencing decisions as presented on the record below. Nos. 19-1074 et al. 27 Cir. 2015) (quoting United States v. Nania, 724 F.3d 824, 830 (7th Cir. 2013); Rita v. United States, 551 U.S. 338, 341 (2007)). Precisely because sentences that fall within the guidelines are presumptively reasonable, sentences that fall below the rec- ommended sentencing range are presumptively reasonable as well. See United States v. Anderson, 580 F.3d 639, 651 (7th Cir. 2009) (citing United States v. Wallace, 531 F.3d 504, 507 (7th Cir. 2008)). Indeed, we have never “deemed a below-range sen- tence to be unreasonably high.” United States v. Brown, 932 F.3d 1011, 1019 (7th Cir. 2019). Moreover, while we have rec- ognized that the district court must account for unwarranted sentencing disparities between codefendants, we have also found that (1) cooperation is a relevant and appropriate factor in sentencing disparities between codefendants and (2) “a dis- trict court that sentences within the Guidelines necessarily gives weight and consideration to avoiding unwarranted dispari- ties.” United States v. Pulley, 601 F.3d 660, 668 (7th Cir. 2010) (emphasis added). Here, the district court explicitly accounted for the dispar- ity in sentences between the Samaniegos and Ramirez-Prado, noting that the Samaniegos received larger downward vari- ances for their cooperation. Moreover, just as a district court sentencing within the guideline range necessarily is taking steps to avoid unwarranted disparities, so too does a district court that sentences below the guideline range. Cf. United States v. Anderson, 580 F.3d 639, 651 (7th Cir. 2009). This case is not destined to be the first one in which we find that a sentence substantially below the recommended guideline range is nonetheless unreasonably high. We there- fore affirm Ramirez-Prado’s sentence. 28 Nos. 19-1074 et al. C. Carrillo-Tremillo Carrillo-Tremillo’s situation is a bit more complex. The district court found that the recommended guideline range for Count 1 was 235 to 293 months and the statutory maxi- mum for Count 4 was 240 months pursuant to 18 U.S.C. § 1956(a), (h). It sentenced Carrillo-Tremillo to 215 months for Count 1 and 120 months for Count 4. Carrillo-Tremillo attempts to chart the same course as Ramirez-Prado, challenging the reasonableness of his 215- month sentence because of the sentencing disparity between himself and the Samaniegos, one of whom received a larger term than he. As with Ramirez-Prado, these arguments are without merit. But there’s more. The district court’s sentence of 215 months’ imprisonment for Count 1 was based on the offense level of 38, which in turn took into account its im- proper conviction with respect to Count 4, which we vacate today. Looking at Count 1 alone, the district court found that Carrillo-Tremillo’s offense level was 36. On remand, Carrillo- Tremillo is entitled to be resentenced solely on the basis of his conviction on Count 1. His sentence on Count 4 must be set aside in light of our ruling on the merits of that conviction. V In the end, the outcome of this complex prosecution re- mains largely undisturbed. We AFFIRM the convictions of all the defendants, except for Carrillo-Tremillo’s conviction on Count 4 for conspiracy to launder money. We also AFFIRM the sentences of all the defendants, once again with the exception of Carrillo-Tremillo. We VACATE Carrillo-Tremillo’s sentence in appeal No. 19-1110, and REMAND his case for further pro- ceedings consistent with this opinion.