United States v. Scott Carnell

In the United States Court of Appeals For the Seventh Circuit ____________________ No. 19-2207 UNITED STATES OF AMERICA, Plaintiff-Appellee, v. SCOTT A. CARNELL, Defendant-Appellant. ____________________ Appeal from the United States District Court for the Southern District of Illinois. No. 4:18-cr-40066-JPG-1 — J. Phil Gilbert, Judge. ____________________ ARGUED FEBRUARY 26, 2020 — DECIDED AUGUST 28, 2020 ____________________ Before ROVNER, WOOD and BARRETT, Circuit Judges. ROVNER, Circuit Judge. Scott Carnell pled guilty to a con- spiracy to distribute a mixture containing methamphetamine. The United States Sentencing Guidelines (U.S.S.G.) distin- guish between mixtures involving run-of-the-mill metham- phetamine and methamphetamine that is at least 80% pure. U.S.S.G. 2D1.1, note C. The latter the Guidelines refer to as “ice,” and that definition carries with it sentences that are 2 No. 19-2207 substantially higher than those for non-ice methampheta- mine. 1 Carnell claims that the government failed to meet its burden of proving that the substance in which he dealt was ice methamphetamine, and therefore he should have been sentenced as though he was involved in a conspiracy to dis- tribute methamphetamine that is less than 80% pure. I. The government indicted Scott Carnell, Kayla Kempfer, Jordan Vuichard, and Jarrett Hood on one count of conspiring to distribute 50 grams or more of a mixture or substance con- taining methamphetamine, pursuant to 21 U.S.C. §§841(a)(1), 841(b)(1)(B) and 846. Carnell pled guilty to the charge. Drug distribution structures can be as complicated as some corpo- rate structures, so we have provided a condensed graphic ver- sion below. 1 For a thoughtful discussion of the disparity in sentencing between meth- amphetamine and ice, see, United States v. Hendricks, 307 F. Supp. 3d 1104, 1105 (D. Idaho 2018). No. 19-2207 3 It is important to note that Nelson, Higgins, and Hughes were not part of this charged conspiracy. But all three co-defend- ants, Kempfer, Vuichard, and Hood, who were part of the conspiracy, admitted to obtaining methamphetamine from the defendant. The pre-sentence investigation report (PSR) concluded that Carnell’s relevant conduct, for sentencing pur- poses, involved 2.37 kilograms of ice. Ultimately the court adopted this quantity finding from the PSR, which, along with increases and decreases for criminal history and ac- ceptance of responsibility, landed Carnell with an offense level of 36 and a sentencing guideline range of 168-210 months. The district court judge imposed a sentence of 192 months. During the sentencing phase, Carnell objected to the PSR’s classification of the drugs as “ice.” He argued that the word “ice” was used colloquially by members of the conspiracy and others with whom he bought and sold drugs to refer to any methamphetamine, rather than a specific form of 80% pure methamphetamine. Although Carnell initially disputed the quantity determination to which he should be held 4 No. 19-2207 accountable, he eventually withdrew his objection to the 2.7 kilogram quantity and objected only to the classification of the drugs as “ice.” In response, the government asserted that it had sufficient evidence to establish, by a preponderance of the evidence, that the defendant was involved with ice as it is defined in U.S.S.G. 2D1.1, note C. In addition, four months after Carnell’s objections to the PSR, the government filed a supplemental response to Carnell’s objections and included lab reports reporting the purity of methamphetamine seized from Carnell’s alleged supplier (who was not part of this con- spiracy). We will describe the facts surrounding that seizure below. Although the government describes its evidence identify- ing the drug as “ice” witness by witness, we think it most use- ful to divide the government’s evidence into three categories. The first is nomenclature. In page after page, the government detailed the many instances in which the defendant and his co-defendants used the word “ice” to describe the drug they procured and distributed. For example, in his post-arrest statement, Carnell continually referred to the drugs he re- ceived and distributed as “ice.” For reasons we elucidate be- low, we need not detail the various evidence that the defend- ant and his co-conspirators bought and sold drugs that they called “ice,” as we conclude that the vernacular use of the word “ice” is insufficient to meet the government’s burden of proving that the drugs were, in fact, ice as defined by 2D1.1, note C of the Guidelines. The second category of evidence involves the physical properties and user-described quality of the drugs. Detective Donald Krull of the Randolph County Sheriff’s office, one of the investigators of this conspiracy, testified that No. 19-2207 5 “methamphetamine is sometimes more of a powdery sub- stance and ice is like a crystalline-type substance.” R. 130 at 9. On the other hand, he also testified that it is possible for d- methamphetamine hydrochloride in a crystalline form to be less than 80% pure, and that one cannot tell the purity of ice by looking at the substance. 2 R. 130 at 38–39. According to detective Krull’s testimony, drug users are experts at deter- mining the quality of drugs and would stop purchasing from a supplier who cut the drugs with impurities. He testified that none of the defendants reported that they cut the drugs (it warrants noting, however, that neither did he report that any of the defendants stated that they did not cut the drugs). At various times during the course of the investigation, all three co-defendants, Vuichard, Kempfer, and Hood, were ar- rested in possession of drugs. Testing of those drugs revealed that they contained methamphetamine. The entities testing those drugs, however—the Illinois State Police Laboratory and the St. Louis County, Missouri Police—lacked the capa- bility to determine purity. And although the lab reports de- scribed each of the substances as “crystalline,” none of them could be identified as having come from Carnell. Vuichard 2 A 1995 amendment to the Guidelines “delete[d] the distinction between d- and l-methamphetamine in the Drug Equivalency Tables in the Com- mentary to § 2D1.1. L-methamphetamine, which is a rather weak form of methamphetamine, is rarely seen and is not made intentionally, but rather results from a botched attempt to produce d-methamphetamine. Under this amendment, l-methamphetamine is treated the same as d-metham- phetamine (i.e., as if an attempt to manufacture or distribute d-metham- phetamine).” U.S.S.G. 2D1.1, Commentary to § 2D1.1 captioned “Applica- tion Notes” (Nov. 1, 1995). See also, United States v. McEntire, 153 F.3d 424, 431 (7th Cir. 1998) (citing U.S.S.G. App. C, Amendment 518 at 423 (Nov. 1, 1995)). 6 No. 19-2207 alleged that the drugs found upon his arrest came from Hig- gins, and no one identified the source of the drugs seized from Hood. At the end of the day, the district court judge refused to admit these lab reports. He did, however, admit a lab re- port regarding drugs confiscated from Kempfer’s purse dur- ing a March 2018 traffic stop in which Carnell was driving. Kempfer claimed that the drugs came from Carnell, but Car- nell stated that they came from Nelson. In any event, those drugs were described as “crystalline” in the laboratory report, but no level of purity could be confirmed, and therefore the information contained in that report is not particularly help- ful. The remainder of the government’s evidence in this cate- gory—physical properties and user-stated quality—came from the testimony of the three co-defendants. Kempfer, Vuichard, and Hood all testified that no one complained about the quality of the ice that they received from Carnell. Nevertheless, Kempfer’s testimony was all over the map about whether she or anyone else could determine the purity of ice. She noted that quality and purity, even among what she considered “ice,” might differ depending on the dealer. She stated, “if it’s ice, you are going to get ice. It’s good pu- rity.” R. 130 at 80. But she also testified contradictorily that “[i]ce is a crystallized substance. It usually comes in shards. It’s pretty clear, unless it’s dirty, which usually [sic] it’s not as good.” R. 130 at 82. And then, almost immediately after that statement, on re-cross, she stated that ice that was dirty is not of a lower quality and then said, “Dirty—when I say dirty, it just has a darker color to it a little bit, but the purity is still good.” Id. Hood also noted that ice could vary in quality and purity. He indicated that he had, on at least one occasion, re- ceived ice from the defendant that was heavily cut with MSM No. 19-2207 7 and thus was not pure. 3 Hood estimated that in that one pur- chase from Carnell, only about 19 of the 28 grams (67.85%) was “good.” Id. at 101. The final category of evidence consists of two laboratory reports in which the Federal Drug Enforcement Agency (DEA) tested the methamphetamine sample and determined it to be 100% pure 4. The first of these lab reports came from a sample of drugs from Vuichard. Vuichard testified that he had two different sources for his ice, Carnell and Lewis Hig- gins. The DEA-tested sample came from drugs that Vuichard obtained from Higgins as part of a different conspiracy. The government had an incentive to link Carnell’s untested ice to the lab-tested ice that Vuichard obtained from Higgins, but its best evidence was that Vuichard testified that he could “just tell” from “do[ing] a lot of dope” that the ice supplied by Hig- gins and Carnell were “both the same quality.” R. 130 at 87, 94. During the sentencing hearing, Vuichard testified that the purity level for the ice he received from Carnell was 3 MSM is the common commercial name for the chemical methylsulfonyl- methane, which is used as a cutting agent for methamphetamins. Pure MSM is an odorless, white, crystalline powder that is highly soluble and mixes readily with most substances without leaving a residue. Metham- phetamine cut with MSM often appears to be uncut because after the chemicals are combined and the mixture cools, the MSM recrystallizes, re- sembling pure methamphetamine. National Drug Intelligence Center, “In- formation Bulletin: Crystal Methamphetamine,” August 2002. https://www.justice.gov/archive/ndic/pubs1/1837/1837t.htm. (Last vis- ited, July 30, 2020). 4 The lab reports list the purity as 100% ± 4%. See R-106-4. We understand this to mean that the substance might be anywhere from 96-100% pure, as a substance cannot be more than 100% pure. For simplicity, we will refer to this as 100% pure. 8 No. 19-2207 “probably 85 and up.” R. 130 at 94. When asked where he got the number 85%, he stated, “one of the police officers, Ralph Jones, told me that the ice I bought off of Lewis Higgins was above 85 percent pure … So, there was comparison for me right there.” R. 130 at 94–95. We have no evidence to indicate how Vuichard would be able to compare quality to that level of precision and whether his drug use palate was sophisti- cated enough to distinguish between, for example, a sample that is 79% pure and one that is 80% pure or more. The second DEA lab result came from drugs seized after the arrest of a different defendant in a different conspiracy— albeit a defendant closely linked to Carnell. On January 9, 2018, someone at a Walgreens store called the police to report that two people had arrived together in a van and were acting suspiciously. The responding police found Carnell and Jessica Hughes, the latter of whom they arrested after determining that she had an active warrant. In a search of her purse inci- dent to the arrest, they found three empty syringes, $1,200 in cash, an electronic scale, and a small resealable plastic bag which contained an off-white crystal substance. After Hughes consented to a search of the van, in the console between the two front seats, the police found a plastic bag filled with sev- eral smaller resealable plastic bags containing an off-white, crystal-like substance, and several small empty plastic bags. At the time of the arrest Hughes stated, “That’s my meth in there. He has nothing to do with it.” R. 106-2 at 9. The police submitted those drugs to a DEA laboratory which determined that the drugs from the center console consisted of 24.083 grams of d-methamphetamine hydrochloride with a purity level of “100% +/- 4%.” R. 106-4. The drugs found in Hughes’ purse weighed .671g and had the same purity. Hughes was No. 19-2207 9 not charged in this conspiracy and was not one of Carnell’s co-defendants in this case. Because the government had no evidence of the purity of any of the drugs involved in the conspiracy, it had a strong incentive to connect Carnell to the drugs from Hughes’ arrest in Missouri. Hughes did not testify at Carnell’s sentencing. Instead, over Carnell’s objection, Jackson County Sheriff Rob- ert Burns testified in court about the content of Hughes’ an- swers from an interview held two days before Carnell’s sen- tencing. Sheriff Burns testified that during the interview, Hughes told him that she and Carnell had been selling a four- ounce “front” at a Red Roof Inn prior to the Walgreens arrest. In other testimony, Vuichard and Hood testified that they bought ice from a young woman named “Jess” in the parking lot of a Red Roof Inn in Missouri, and that Carnell had accom- panied Jess in the van and handed Vuichard the drugs and took the money. R. 130 at 89–90. According to Vuichard, Car- nell told him the next day that Jess was arrested shortly there- after. The government did not explicitly link this occurrence with the Walgreens arrest, but the implication in the govern- ment’s brief is that it was the same event. II. The only question to resolve on appeal is whether the dis- trict court properly found that the government met its burden of proof in demonstrating that Carnell sold d-methampheta- mine with a purity of 80% or more. “The Government’s bur- den at sentencing is substantially lower than at trial,” as a sen- tencing court “must only find that a preponderance of reliable evidence supports the drug quantity finding.” United States v. Tankson, 836 F.3d 873, 881 (7th Cir. 2016) (internal quotation omitted). Therefore, “[a]t sentencing, the government must 10 No. 19-2207 prove that the methamphetamine attributed to the defendant is more likely than not” methamphetamine as described in the Guidelines. United States v. McEntire, 153 F.3d 424, 432 (7th Cir. 1998). In this case we must look to see whether the gov- ernment had a preponderance of reliable evidence that the methamphetamine it attributed to Carnell for a relevant con- duct determination was “ice” methamphetamine as that term is defined under the Guidelines Section 2D1.1, note C. A. Evidence of the purity level of the methamphetamine. The sentencing guidelines define ice as “a mixture or sub- stance containing d-methamphetamine hydrochloride of at least 80% purity.” U.S.S.G. 2D1.1, note C. We begin with a car- dinal rule of statutory construction: courts “must give effect, if possible, to every clause and word of a statute.” Loughrin v. United States, 573 U.S. 351, 358 (2014) (internal citation omit- ted). When “Congress includes particular language in one section of a statute but omits it in another” a court must pre- sume that Congress intended a difference in meaning. Rus- sello v. United States, 464 U.S. 16, 23 (internal citation omitted). These rules apply equally to the Guidelines. United States v. Gawron, 929 F.3d 473, 478 (7th Cir. 2019). Our interpretation of the Guidelines thus “begin[s] with the text of the provision and the plain meaning of the words in the text.” United States v. Hill, 645 F.3d 900, 907 (7th Cir. 2011) (quoting United States v. Arnaout, 431 F.3d 994, 1001 (7th Cir. 2005)). We additionally consider the Guidelines’ Application Notes “as part of the Guidelines themselves, and not mere commentary on them.” Arnaout, 431 F.3d at 908; see also Stinson v. United States, 508 U.S. 36, 38 (1993) (holding that the commentary to a Guideline is as binding as a Guideline). No. 19-2207 11 Starting with these principles, we must make sense of each word of the Guidelines’ definition of ice as “a mixture or sub- stance containing d-methamphetamine hydrochloride of at least 80% purity.” U.S.S.G. 2D1.1, note C. We note that no other substance in the Guidelines’ drug chart in 2D1.1 is de- fined by its level of purity in this way (although the purity of PCP, amphetamine, hydrocodone and oxycodone are ac- counted for in other ways, see footnote 5 below). It is clear that the Sentencing Commission intended that this difference have teeth. The commentary to the Guideline states as follows: (C) Upward Departure Based on Unusually High Purity.—Trafficking in controlled sub- stances, compounds, or mixtures of unusually high purity may warrant an upward departure, except in the case of PCP, amphetamine, meth- amphetamine, hydrocodone, or oxycodone for which the guideline itself provides for the con- sideration of purity. U.S.S.G. 2D1.1, commentary, application note 27(C). 5 In other words, the Guidelines note, and courts have held, that a 5 The purity of PCP, hydrocodone, oxycodone and amphetamine is ac- counted for by referring to the “actual” measures of these drugs. The Guidelines state: The terms “PCP (actual)”, “Amphetamine (actual)”, and “Methamphetamine (actual)” refer to the weight of the controlled substance, itself, contained in the mixture or substance. For example, a mixture weighing 10 grams containing PCP at 50% purity contains 5 grams of PCP (actual). In the case of a mixture or substance containing PCP, amphetamine, or methamphetamine, use the of- fense level determined by the entire weight of the mixture or substance, or the offense level determined by the 12 No. 19-2207 district court may not use purity as a reason for an upward departure in a methamphetamine case, as the Guidelines al- ready account for purity. See also, United States v. Cones, 195 F.3d 941, 944 (7th Cir. 1999). In short, purity matters for meth- amphetamine. It is tempting to turn to the body of case law that our cir- cuit has developed to distinguish between cocaine base and crack cocaine and apply those rules to distinguish between methamphetamine and ice. The crack versus cocaine case law developed in response to a chasmic disparity in the sentenc- ing for these two drugs. Historically, statutes and the Guide- lines treated one hundred grams of powder cocaine as the equivalent of one gram of crack, creating enormous sentenc- ing disparities for two chemically equivalent drugs, and thus a great need for courts to accurately determine which drug was at play in any particular crime. See Kyle Graham, Sorry Seems to Be the Hardest Word: The Fair Sentencing Act of 2010, Crack, and Methamphetamine, 45 U. Rich. L. Rev. 765, 766 (2011). 6 It turns out, however, that distinguishing between crack and cocaine was a complicated task. As explained by our court in United States v. Stephenson: Given our sophisticated crime laboratories, it might seem an easy task to determine whether a particular drug is crack or another form of weight of the PCP (actual), amphetamine (actual), or methamphetamine (actual), whichever is greater.” U.S.S.G. 2D1.1, note B. Methamphetamine has an additional pu- rity requirement for sentencing as “ice.” Id. at note C. 6In 2010, Congress attempted to narrow this disparity by enacting the Fair Sentencing Act of 2010, PL 111-220, August 3, 2010, 124 Stat 2372. No. 19-2207 13 cocaine base, but no chemical test can distin- guish between crack and cocaine base. Crack is merely one form of cocaine base—a form that arises as the end result of one method of turning the salt form of cocaine, cocaine hydrochloride (powder cocaine), back into a base form. See United States v. Edwards, 397 F.3d 570, 574 (7th Cir. 2005). Drug dealers alter the form of natu- rally occurring cocaine to offer drug users their preferred method of ingesting the chemical. Crack can be smoked, but not snorted or in- jected; powder cocaine can be snorted, but not smoked. [United States v. Booker, 70 F.3d 488, 490–91 (7th Cir. 1995)]. United States v. Stephenson, 557 F.3d 449, 452–53 (7th Cir. 2009). Moreover, this court “rejected rigid definitions of crack,” noting that “to employ such a rigid definition would invite those in the drug trade to make minor changes in structure, processing, or packaging to avoid the increased penalties for selling crack cocaine.” Id. at 453. As a result of these two fac- tors—the fact there is no chemical test that can distinguish crack from cocaine, and the fact that a rigid definition would invite manipulation—courts were forced to come up with a different solution for determining drug composition. The so- lution was to declare that “[t]he experts in this field are those who spend their lives and livelihoods enmeshed with the drugs—users, dealers, and law enforcement officers who spe- cialize in narcotics crimes.” Id. (citing United States v. Kelly, 519 F.3d 355, 364 (7th Cir. 2008); United States v. Bradley, 165 F.3d 594, 596 (7th Cir. 1999)). 14 No. 19-2207 In sum, the case law that has evolved to distinguish one form of a drug from another developed in response to an at- tempt to distinguish crack from other forms of cocaine base where no lab test or rigid definition exists to distinguish be- tween the drugs. Ice, however, is different, not only because of the availability of lab tests and precise definitions, but also because of a Congressional choice about purity. See Cones, 195 F.3d at 944. In Cones, we pointed out that Congress made de- liberate choices about when a court should focus on purity and when it should not: Drug purity cannot reasonably be described as a circumstance that the Commission has over- looked or inadequately considered. Both the rel- evant statutes and the Guidelines use the for- mula “mixture or substance containing a detect- able amount” of a given drug. E.g. 21 U.S.C. § 841(b)(1)(A)(i). The possibility of converting to a uniform purity—whether 100% purity or “street-level” purity—was considered and de- liberately rejected. See United States v. Marshall, 908 F.2d 1312 (7th Cir.1990) (en banc), affirmed under the name Chapman v. United States, 500 U.S. 453 (1991); United States v. Neal, 46 F.3d 1405 (7th Cir.1995) (en banc), affirmed, 516 U.S. 284 (1996). When defendants who sold a highly di- lute drug objected that the “detectable amount” approach greatly magnifies their punishment compared with people who sell a more concen- trated drug, both this court and the Supreme Court responded in the cases just cited that this outcome is the result of deliberate choices by Congress and the Sentencing Commission. No. 19-2207 15 Statutes and Guidelines allow conversion to a uniform purity for PCP and methamphetamine, and the Guidelines now allow a conversion for LSD, which reinforces the conclusion that for other drugs Congress and the Commission have rejected a common-purity approach. Id. The Sentencing Commission did not reject a purity ap- proach for methamphetamine ice. To the contrary, it empha- sized it. The Guidelines dictate that “[u]nless otherwise spec- ified, the weight of a controlled substance set forth in the table refers to the entire weight of any mixture or substance con- taining a detectable amount of the controlled substance.” U.S.S.G. 2D1.1, note A. And then the Guidelines note the ex- ception for ice, which is defined as a “mixture or substance containing d-methamphetamine hydrochloride of at least 80% purity.” U.S.S.G. 2D1.1, note C. This is both subject to a rigid definition and testable. It is abundantly clear that the “80%” language has mean- ing. And although it makes sense in the context of crack and cocaine to define the users, dealers, and law enforcement of- ficers as the experts in the field at distinguishing between the drugs, it cannot carry the government’s burden in a case al- leging methamphetamine ice. The government must prove, albeit only by a preponderance of evidence at sentencing, that the substance was, in fact methamphetamine of at least 80% purity. We think it defies common sense that even the most experienced dealer, user, or police officer could somehow de- tect the difference between 79% pure methamphetamine and 80% pure methamphetamine. We therefore reject the Eighth Circuit’s conclusion that the same kind of evidence that 16 No. 19-2207 suffices for distinguishing between crack and cocaine—the experience of users, dealers and law enforcement officers, without more—suffices to meet the burden of proving that a particular drug is 80% pure methamphetamine. See United States v. Walker, 688 F.3d 416, 423–25 (8th Cir. 2012); United States v. Lugo, 702 F.3d 1086, 1090–91 (8th Cir. 2013); see also United States v. Cockerill, No. 99-4634, 2000 Westlaw 852608, at *1 (4th Cir. June 28, 2000) (in a case involving calculation of actual methamphetamine in a mixture—not ice—stating, “[t]he sentencing guidelines do not require absolute certainty about the amount of drugs or their purity when the drugs are not seized or the amount seized does not reflect the scale of the offense. In such cases, the district court may estimate the quantity, and may use whatever reliable evidence is availa- ble.”). Given this understanding of the government’s burden, we can eliminate the utility of several categories of evidence pre- sented by the government. For instance, we can eliminate all of the evidence in the nomenclature category in which the ar- gument is that “everyone referred to the drugs as ice.” We simply have no idea whether it is customary within this circle of users and dealers to refer to any methamphetamine of at least 65% or 75% or 79% purity as “ice,” or whether they con- fine that term to methamphetamine that would qualify as ice under the guidelines—that is, methamphetamine of at least 80% purity. It is possible that in this particular cohort, ice is used interchangeably for any form of methamphetamine. See, e.g., United States v. Lee, 725 F.3d 1159, 1166–67 (9th Cir. 2013) (“The record indicates that all members of [this] drug traffick- ing scheme referred to the methamphetamine, regardless of its purity, as “ice.”). None of this evidence meets the No. 19-2207 17 government’s burden of proving by a preponderance of evi- dence that this particular ice was 80% pure. We can also eliminate all of the evidence in the category of “I am a heavy user and this seemed like high quality meth- amphetamine.” The district court relied primarily on this genre of evidence when deciding that Carnell was distrib- uting methamphetamine in the form of ice: One thing I have found through the years is drug dealers know their product. And when you have three of them testify, just from their testimony alone, even apart from the law en- forcement testimony from hearsay from their interviews, it’s clear that Mr. Carnell knew this was ice methamphetamine… . Common sense is overwhelming that this Defendant was aware that this was methamphetamine in the form of ice that he was distributing. R. 130 at 112–13. But as we have just described, we cannot say, as we can in the case of the cocaine and crack distinction, that users know their product to the required level of precision. For example, when Kempfer described the ice as “good pu- rity,” we have no idea what she meant by that. Nor do we have evidence even drug dealers and heavy users can detect the difference between 79% and 80% pure methamphetamine. Common sense suggests otherwise. The government references three cases to support its prop- osition that it can prove drug type through the testimony of those who are familiar with the drug, such as experienced po- lice officers, forensic chemists, informants, and those that buy, sell and use the drug. See Government’s Brief at 37, citing 18 No. 19-2207 United States v. Padilla, 520 F.3d 766, 771 (7th Cir. 2008); United States v. Anderson, 450 F.3d 294, 301 (7th Cir. 2006); Bradley, 165 F.3d at 596. All three of these cases, however, involved the distinction between crack and cocaine, and we have already ascertained that the Guidelines’ demands for an ice-based sentence are more exacting. Evidence that the users and dealers described the sub- stance as “crystalline,” or “glass like shards” also fails to meet the burden. These descriptions are only relevant if the gov- ernment also puts forth evidence from a chemist or other rel- evant expert that methamphetamine cannot form a crystalline structure below 80% purity. In fact, one of the witnesses that the government set forth as an expert, Retired Detective Krull, testified that it is possible for d-methamphetamine hydro- chloride in a crystalline form to be less than 80% pure. R. 130 at 38–39. 7 This leaves only two remaining pieces of evidence—the two methamphetamine samples that were tested by DEA la- boratories and found to be 100% pure. The first is the sample taken from Vuichard, who received the tested drugs from Higgins as part of a different conspiracy. The only connection between Carnell and those drugs is that Vuichard testified that, in his experience as a drug user, the drugs he received from Carnell were of the same quality as those he received from Higgins. But as we noted, we have no evidence that a 7At oral argument the government was unable to answer the court’s in- quiry as to what purity is required for the formation of crystalline meth- amphetamine. Oral Argument at 19:04-19:16. In any event, this infor- mation would have been outside of the record. No. 19-2207 19 drug dealer or user can distinguish the quality of drugs with such precision. The second is the sample of drugs seized from Hughes during an arrest in which she was travelling in a van with Carnell. This is a much closer case, especially given the impli- cation from the facts that Carnell and Hughes may have been selling drugs together earlier during the day. Nevertheless, at the time of arrest, Hughes insisted the drugs were hers alone, and Hughes’ drugs were not part of the Carnell/ Kempfer/Vuichard/Hood conspiracy for which Carnell would have been responsible. See United States v. Robinson, 964 F.3d 632, 638 (7th Cir. 2020) (“In a drug conspiracy, each conspirator is responsible not only for drug quantities directly attributable to him but also for amounts involved in transac- tions by co-conspirators that were reasonably foreseeable to him.”). The government never appeared to dot its i’s and cross its t’s on connecting Carnell to the alleged earlier drug deal with Vuichard and Hood. Moreover, because Hood tes- tified that he received both “good” and “not good” ice from Carnell, the fact that some ice in a separate conspiracy was “pure” is not sufficient to attribute 2.37 kilograms of 80% pure ice to Carnell. We review the sentencing court’s determination of the re- liability of evidence for an abuse of discretion, and the fact finding as to the quantity of drugs attributable to a defendant for clear error. United States v. Tankson, 836 F.3d 873, 881 (7th Cir. 2016). The evidence in this case, however, simply cannot support a finding that the methamphetamine was 80% pure. The district court abused its discretion in finding this evi- dence reliable using the standards we apply to evidence of drugs for which the Guidelines do not require a particular 20 No. 19-2207 level of purity. This is not a matter of a credibility determina- tion by the district court. We do not doubt the testimony that the other defendants knew that they were buying and selling a substance that they considered to be “ice,” or the testimony that they thought that the ice was of high quality. Instead, our determination is that these vague descriptions do not meet the government’s burden of proof that the drug they were distrib- uting, along with Carnell, was “d-methamphetamine hydro- chloride of at least 80% purity,” as described by U.S.S.G. 2D1.1, note C. This is not to say that a lab report is always needed to meet the burden required by 2D1.1, note C. Nevertheless, even the Eighth Circuit, which disagrees with our newly proposed ap- proach, has noted that: This is not to diminish the value of chemical testing when reasonably practicable, particu- larly given the increased penalties for “ice” methamphetamine and the relatively high pu- rity level specified in the advisory Guidelines definition. Scientific testing of at least part of a quantity of suspected “ice” methamphetamine seized from a conspiracy is one of the strongest means by which the government can meet its burden of proving the methamphetamine at- tributed to a defendant is “ice” as defined in the Guidelines. See United States v. Verdin–Garcia, 516 F.3d 884, 896 (10th Cir. 2008) (explaining “[l]aboratory test results are perhaps more per- suasive evidence of amounts and purities than eyewitness testimony or wiretapped conversa- tions”). We also agree with the Third Circuit's No. 19-2207 21 observation that “where a written plea agree- ment is entered[,] questions of notice and proof at sentencing could be greatly minimized by simply including language in the plea agree- ment by which a defendant acknowledges the identity of the drugs involved.” United States v. Roman, 121 F.3d 136, 141 n. 4 (3d Cir.1997). Walker, 688 F.3d at 425 n.4. Nevertheless, we are not ruling out the possibility that there may be other evidence of purity. For example, the government could provide evidence connecting the visual description of the methamphetamine to the pu- rity—for example if the government had evidence that meth- amphetamine will not appear in crystalline form until it is at least 80% pure. But we cannot rely on the expertise of a drug user, dealer, or law enforcement officer alone to determine that methamphetamine is more than 80% pure as opposed to 79% pure, and the Guidelines clearly require a more exacting determination than it does for crack and cocaine, for example. We also leave for another day the question of whether all of the methamphetamine attributable to a defendant must be tested, and if not, what would constitute a reliably representa- tive sample. In this case, none of the ice attributable to the conspiracy had been tested. The two cases from our Circuit on which the government places weight shed light only around the edges of the ques- tion that confronts us today. In United States v. McEntire, 153 F.3d 424 (7th Cir. 1998), we allowed circumstantial evidence and the evidence of experts to fill in the gaps where the labor- atory testing was not able to distinguish between l-metham- phetamine and d-methamphetamine. See note 2, supra. In that case, however, the court accepted the testimony of a DEA 22 No. 19-2207 expert that l-methamphetamine is simply the result of a failed attempt at manufacturing d-methamphetamine, does not pro- vide a euphoric high, and therefore has no demand or value on the street, and in fact had been found in only .07% of all samples analyzed by DEA chemists in 10,710 cases over six- teen years. Id. at 429–30. In contrast, there is no similar expert testimony that met the government’s burden to prove that the methamphetamine in this case was at least 80% pure. Our decision in United States v. Castenada, 906 F.3d 691 (2018), fails to provide a clear precedent for establishing the government’s burden to demonstrate the identity of ice. Id. at 694. In Castenada, the court dedicated one short paragraph to the issue of determining whether the drug at issue could be identified as ice as opposed to generic methamphetamine. In Castenada, however, the answer was not relevant. The govern- ment charged Castenada’s co-defendants with distributing large quantities of mixtures containing methamphetamine, but charged Castenada himself with distributing a much smaller quantity of ice. Castenada argued he should not be charged with the distribution of ice when his co-conspirators had been charged with distributing a mixture of methamphet- amine. The district court, however, concluded that the higher quantity of generic methamphetamine was the equivalent of the lower quantity of ice attributed to Castenada for purposes of sentencing, and so the distinction was not critical to Caste- nada’s sentence. Id. at 694–95. Moreover, methamphetamine seized from a co-conspirator was tested and found to be 100% pure. Id. at 694. Here we have no laboratory tests from meth- amphetamine directly linked to this conspiracy, and the dis- tinction between a finding of ice or simply a mixture of meth- amphetamine would have an impact on Carnell’s sentence. No. 19-2207 23 Neither McEntire nor Castenada directly answer the ques- tion we face today, and therefore our decision here does not explicitly overrule our earlier precedent. It does, however, create conflict with the Eighth Circuit’s decision in United States v. Walker, 688 F.3d 416 (8th Cir. 2012). It appears as though other circuit courts have largely avoided addressing the question of the government’s burden in demonstrating that ice meets the definition set forth in the Guidelines. Some have touched upon the question of the sentencing disparity between methamphetamine and ice. See, e.g., United States v. Burgos-Vasquez, 784 F. App'x 663, 670 (11th Cir. 2019); United States v. Flores-Perez, 749 F. App'x 793, 798–99 (11th Cir. 2018); United States v. Kort, 440 F. App'x 678, 683–85 (10th Cir. 2011). Other courts have pointed out the disparity, but ultimately concluded that they did not need to determine whether the drugs at issue should be classified as methamphetamine or ice because the distinction did not matter to the sentence, as in Castenada. See, e.g., United States v. Martinez, 872 F.3d 293, 301– 03 (5th Cir. 2017); United States v. Patterson, 713 F. App'x 916, 918 (11th Cir. 2017); United States v. Fisher, 319 F. App'x 795, 797 (11th Cir. 2009); United States v. Varela, 586 F.3d 1249, 1251 n.3 (10th Cir. 2009); United States v. Verdin-Garcia, 516 F.3d 884, 897 (10th Cir. 2008) (“We therefore need not reach at all the question whether the evidence was sufficient to prove by a preponderance that the disputed narcotics alleged to be “ice” were in fact 80% pure.); United States v. Cook, 224 F. App'x 794, 799–800 (10th Cir. 2007). Because of our disagreement with the Eighth Circuit, we have circulated this opinion among all judges of this court in regular active service, pursuant to Circuit Rule 40(e). No judge favored a rehearing en banc on the question of whether at sentencing circumstantial evidence by users, dealers and 24 No. 19-2207 law enforcement that a drug appears to be ice based on look, smell, effect, nomenclature or the like will suffice to meet the government’s burden, by a preponderance of the evidence, that a drug is at least 80% pure methamphetamine. B. The confrontation clause issue. Although we now hold that the circumstantial evidence by users, dealers and law enforcement that a drug appears to be ice based on look, smell, effect, nomenclature or the like will not suffice to meet the government’s burden, by a pre- ponderance of the evidence, that a drug is at least 80% pure methamphetamine, we also hold that the district court did not err by admitting the government’s lab reports where the ana- lysts and scientists who drafted those reports were not made available for confrontation in the courtroom. “Sentencing courts have long enjoyed discretion in the sort of information they may consider when setting an appropriate sentence.” Dean v. United States, 137 S. Ct. 1170, 1175 (2017). Because the “Sixth Amendment’s confrontation clause does not apply to a sentencing proceeding, the court may rely on the testimony or other statements of a witness even if that witness has not been subject to cross-examination by the defendant.” United States v. Ghiassi, 729 F.3d 690, 695–96 (7th Cir. 2013); See also United States v. Campuzano-Benitez, 910 F.3d 982, 990 (7th Cir. 2018) (“[T]he Confrontation Clause in the Sixth Amendment does not apply at sentencing.). Carnell tries to do an end-run around these precedents, by stating that his inability to test the reliability of the lab reports by confronting the authors in- fringed upon his right to due process rather than his Sixth Amendment rights. But dressing his confrontation clause ar- gument in due process clothes gets him no farther. Early Su- preme Court cases also relied on the due process clause in No. 19-2207 25 stating that “once the guilt of the accused has been properly established, the sentencing judge, in determining the kind and extent of punishment to be imposed, is not restricted to evidence derived from the examination and cross-examina- tion of witnesses in open court but may, consistently with the Due Process Clause of the Fourteenth Amendment, consider responsible unsworn or ‘out-of-court’ information relative to the circumstances of the crime and to the convicted person’s life and characteristics.” Williams v. State of Okl., 358 U.S. 576, 584 (1959); see also Ghiassi, 729 F.3d at 695–96 (referencing opinions relying on both the Sixth Amendment’s Confronta- tion Clause and the Due Process Clause to support the fact that a sentencing court may rely on the testimony or other statement of a witness even if that witness has not been sub- ject to cross-examination by the defendant.) Given this prece- dent, we see no reason that we might disturb the district court’s discretion to allow in the laboratory reports despite the fact that the authors were not available for questioning. The judgment of the district court is REVERSED in part and AFFIRMED in part, and we remand for further proceed- ings consistent with this opinion.