Gary J. Eschman pleaded guilty to several drug-related charges and a gun charge. Rejecting Eschman's objections to the quantity of drugs attributable to him for sentencing purposes and declining to reduce Eschman's sentence for acceptance of responsibility, the district court sentenced Eschman to almost twenty years in prison. Eschman appeals his sentence, and for the reasons stated herein, we reverse.

* Because Eschman does not challenge his conviction, we discuss only those facts relevant to his sentencing. Acting on a tip from a local Wal-Mart store that had sold Eschman's son-in-law Anthony Jines a large quantity of pseudoephedrine, a precursor to methamphetamine, agents from the Metropolitan Enforcement Group of Southwestern Illinois (MEGSI) conducted a search of the residence where Eschman lived with Anthony Jines and his daughter, Darlla Jines. The agents found, among other things, methamphetamine in the Jineses' bedroom, and two firearms and 6,400 30- milligram pseudoephedrine pills in Eschman's bedroom.1 Shortly after criminal charges were filed against him, his daughter, and his son-in- law, Eschman pleaded guilty to conspiracy to manufacture and possess with intent to distribute methamphetamine, in violation of 21 U.S.C. sec.sec. 841(a)(1), 846, maintaining a place for the manufacture of methamphetamine, in violation of 21 U.S.C. sec. 856(a)(1), and being a felon in possession of a firearm, in violation of 18 U.S.C. sec. 922(g)(1).2 In doing so, Eschman admitted that he "cooked" or manufactured methamphetamine in a shed next to the Jineses' residence.

Since the sentencing guidelines base the applicable offense level on the drug quantity involved, the key question at Eschman's sentencing was how much methamphetamine should be attributed to him. See U.S. Sentencing Guidelines (U.S.S.G.) sec. 2D1.1 (1998). Before his sentencing, the government (and later the Probation Department in its pre-sentence report) estimated the "production capacity" of Eschman's methamphetamine "laboratory" (i.e., the shed) to be 177 grams of actual (pure) methamphetamine based on a 100% "theoretical" yield of the pseudoephedrine pills found in his possession (i.e., converting pseudoephedrine into methamphetamine at a one-to-one ratio). Eschman, however, objected to this 100% conversion rate as a means for determining his base offense level.

At sentencing, Eschman introduced testimony from Dr. Terry Martinez, a chemist and professor at the St. Louis College of Pharmacy, who stated that a 100% conversion rate is merely theoretical and that professional chemists can only obtain a 90% yield using professional equipment. Based on a scientific study conducted by the Iowa Department of Public Safety ("Iowa study"), Dr. Martinez indicated that an average yield for a clandestine laboratory is from 40% to 50%. He stated that a clandestine laboratory can, at most, obtain an 80% yield. He characterized Eschman's lab as "primitive" and testified that no expert, in his view, could determine the possible yield of methamphetamine for Eschman's lab.

In rebuttal, the government offered the testimony of Virginia Kleekamp, a chemist with the Drug Enforcement Administration (DEA). Kleekamp testified that DEA chemists do a theoretical conversion rate of pseudoephedrine to methamphetamine of 100%, or one-to-one, adjusted only for the difference in molecular weight. She explained that the DEA uses a one-to-one theoretical conversion ratio because it is difficult to obtain an accurate measure of the production capacity of a clandestine laboratory. She admitted, as a practical matter, that it is impossible to obtain a 100% yield. She indicated that an average yield for a clandestine laboratory is from 40% to 60%, but she has noted yields as high as 85%. However, she did not dispute the findings of the Iowa study.

After hearing testimony from these two experts, the district court found Dr. Martinez's testimony not credible and accepted the one-to-one conversion ratio as a means to determine the applicable base offense level. The district court agreed with the pre-sentence report that the production capacity of Eschman's laboratory was 177 grams of pure methamphetamine. Based on the statements and testimony of Anthony Jines, who sold methamphetamine produced by Eschman, the district court also determined that Eschman manufactured, at least, 36 ounces (or 1020.6 grams) of mixture or substance containing methamphetamine.

On the basis of these drug amounts, the district court determined that Eschman had a base offense level of 34,3 which when increased by two levels for his possession of a firearm, resulted in an offense level of 36. The district court then denied Eschman a three-level reduction in his offense level for acceptance of responsibility, concluding that Eschman had not genuinely accepted responsibility under the sentencing guidelines. The court sentenced Eschman to two 235-month sentences for the drug- related convictions (the upper limit of the relevant sentencing range) and a 120-month sentence for the felon-in-possession conviction, all to be served concurrently.

On appeal, Eschman argues that the district court erred in calculating his base offense level by (1) holding him accountable for 177 grams of pure methamphetamine and (2) denying him a three- level reduction for acceptance of responsibility.

We review the district court's calculation of drug quantity, as well as its determination of whether a defendant has accepted responsibility, for clear error. See United States v. Johnson, 200 F.3d 529, 537 (7th Cir. 2000); United States v. Mancillas, 183 F.3d 682, 711 (7th Cir. 1999).

Eschman argues that the district court erred by holding him accountable for 177 grams of pure methamphetamine based on the amount of pseudoephedrine found in his possession. Specifically, Eschman contends that the district court did not have a reliable factual basis for the drug quantity calculation.

In cases "[w]here there is no drug seizure or the amount seized does not reflect the scale of the offense, the court shall approximate the quantity of the controlled substance." U.S.S.G. sec. 2D1.1, cmt. 12 (1998). In making this approximation, the district court may consider "the price generally obtained for the controlled substance, financial or other records, similar transactions in controlled substances by the defendant, and the size or capability of any laboratory involved." Id. Because there was no seizure of the substance charged in the offense, the district court in this case applied sec. 2D1.1 to approximate the amount of methamphetamine Eschman could have produced in his laboratory based upon the quantity of pseudoephedrine found in his possession.4 The district court, in accepting the government's recommendation set forth in the pre-sentence report, found that Eschman could have converted the pseudoephedrine into methamphetamine at a one-to-one conversion ratio (i.e., a 100% yield). We will uphold this finding to the extent it is based on reliable evidence. Johnson, 200 F.3d at 537.

From our review of the record, the district court's finding lacks an evidentiary basis. Both parties' experts testified that a 100% yield is merely theoretical (in other words, unattainable). The experts also testified that although an 80-85% yield might be possible with a clandestine laboratory, yields in the range of 40%-60% were more probable. This data is confirmed by the Iowa study, which Eschman introduced at sentencing. While the government must prove the quantity of drugs attributable to Eschman only by a preponderance of evidence, United States v. Galbraith, 200 F.3d 1006, 1011 (7th Cir. 2000), the record is void of any evidence which would reasonably support the district court's decision to base its methamphetamine quantity calculation on a one-to- one conversion ratio.

Criminal defendants, like Eschman, have a due process right to be sentenced on the basis of reliable information, id. at 1012, and district courts cannot quantify yield figures without regard for a particular defendant's capabilities when viewed in light of the drug laboratory involved. See, e.g., United States v. Cole, 125 F.3d 654, 655 (8th Cir. 1997) (relevant inquiry is on what defendant, not "an average cook," is capable of yielding); United States v. Hamilton, 81 F.3d 652, 653-54 (6th Cir. 1996) (rejecting standardized drug conversion formulas in favor of individualized assessment of defendant's capabilities); United States v. Mahaffey, 53 F.3d 128, 132-33 (6th Cir. 1995) (same). Here, the record contains no evidence regarding the sorts of yields Eschman could, with his equipment and recipe, obtain in his methamphetamine laboratory (or, for that matter, even evidence regarding yields of similarly-situated defendants).5 See United States v. Shaffer, 993 F.2d 625, 629 (7th Cir. 1993) (court may approximate amount that laboratory could have produced based upon DEA chemist's testimony regarding chemical operations and materials found at "farmhouse" laboratory and production capacity of defendant's 12-liter flask when taking into account "sloppy" laboratory procedures); Mahaffey, 53 F.3d at 132 (court may approximate amount that laboratory could have produced based upon yields of similarly-situated defendants); United States v. Beshore, 961 F.2d 1380, 1383 (8th Cir. 1992) (court may approximate amount that laboratory could have produced based upon quantity of precursor chemicals, size of laboratory, and recipes to "cook" methamphetamine seized); United States v. Short, 947 F.2d 1445, 1456-57 (10th Cir. 1991) (court may approximate amount that laboratory could have produced based upon testimony of DEA chemist and characteristics of laboratory equipment seized). Thus, the district court erred in relying on the one-to-one conversion ratio when determining Eschman's base offense level.6

On remand, the district court must undertake a more precise inquiry into the quantity of methamphetamine attributable to Eschman, using reliable evidence to support its ultimate drug quantity calculation.

Eschman also challenges the district court's refusal to grant him a three-level reduction in offense level for acceptance of responsibility under U.S.S.G. sec. 3E1.1. The district court refused to grant the reduction because, in its view, Eschman had not genuinely accepted responsibility, but instead was simply attempting to manipulate the criminal justice system in order to reduce his sentence.

Under U.S.S.G. sec. 3E1.1, a defendant who clearly demonstrates acceptance of responsibility for his or her offense is entitled to a reduction in his or her offense level. To qualify for the reduction a defendant must, in a timely manner, (1) demonstrate sincere remorse or contrition, (2) truthfully admit the conduct comprising the offense, and (3) neither falsely deny nor frivolously contest relevant conduct. United States v. Mancillas, 183 F.3d 682, 711 (7th Cir. 1999); United States v. Purchess, 107 F.3d 1261, 1269 (7th Cir. 1997); U.S.S.G. sec. 3E1.1, cmt. 1 (1998).

Here, Eschman pleaded very early in the case and never challenged the charges against him. Although he raised objections to the pre-sentence report at sentencing, he never expressed outright denials of relevant conduct and the district court found his challenge to the "production capacity" evidence not frivolous. Eschman also submitted a statement to the court acknowledging his wrongdoing and expressing regret, and at sentencing, he apologized for his conduct, stating that he was "sorry" for what he did. The district court, however, did not believe Eschman had genuinely accepted responsibility.

In light of his early, consistent, and repeated acceptance of responsibility, it appears that the district court's determination was affected by the manner in which Eschman prefaced his statements. In accepting responsibility, Eschman stood firm in challenging the drug quantity calculation and our earlier discussion demonstrates that he had good reason for doing so. While we do not take lightly the district court's apparent concerns about Eschman's contrition, see U.S.S.G. sec. 3E1.1, cmt. 5 (1998) (district court's determination entitled to "great deference" on review), the court's finding appears to have been colored by Eschman's firm, but good faith challenge to the drug quantity calculation. Our decision rejecting the court's drug quantity calculation places the acceptance of responsibility question in a new light. On remand, the district court should reevaluate whether Eschman is entitled to a three-level reduction in his base offense level for acceptance of responsibility.

For the foregoing reasons, we VACATE Eschman's sentence and REMAND for resentencing in accordance with this opinion.

Easterbrook, Circuit Judge, concurring.

I join the court's opinion, which resolves the issues the parties presented for decision and holds that the record does not demonstrate that the conspirators could have turned their pseudoephedrine into methamphetamine of equal weight. As my colleagues observe, however, the conversion ratio is a "somewhat curious" subject to pursue. Maj. op. 890 n.4. Both the prosecutor and defense counsel misunderstand the import of the provision that affects Eschman's sentence.

Application Note 12 to U.S.S.G. sec.2D1.1 provides

Types and quantities of drugs not specified in the count of conviction may be considered in determining the offense level. See sec.1B1.3(a)(2) (Relevant Conduct). Where there is no drug seizure or the amount seized does not reflect the scale of the offense, the court shall approximate the quantity of the controlled substance. In making this determination, the court may consider, for example, the price generally obtained for the controlled substance, financial or other records, similar transactions in controlled substances by the defendant, and the size or capability of any laboratory involved.

The prosecutor contended, and the district judge concluded, that seizures of methamphetamine did not "reflect the scale of the offense", so the parties set out to determine "the size or capability of any laboratory involved." But instead of inquiring whether the laboratory was large, sophisticated, efficient, and so on--the keys to its ability to turn out methamphetamine, and therefore good clues to how much of that drug this operation had produced (and thus to the scale of the offense)--both litigants and judge asked instead how much methamphetamine could have been made using the stock of raw materials on hand when the police arrived. The district court concluded that the pseudoephedrine could have been used to make an equal weight of methamphetamine, but this finding is clearly erroneous, for it conflicts with expert testimony offered by both sides. Under Application Note 12, the finding also is irrelevant, because it does not demonstrate "the size or capability of any laboratory involved." It is like saying that the "size or capability" of an automobile body plant depends on how much aluminum can be found nearby on a given day. But that tells us only the plant's inventory, not its "size or capability"; many producers of automobiles (or drugs) practice just-in-time purchasing to curtail costs. An auto body plant produces many more cars per year (or even per week) than the aluminum on hand at a given moment can yield. Just so with drug manufacturing enterprises.

Application Note 12 is designed to match the penalty to the true scale of the drug operation. That the police discovered some inputs for drug production is happenstance and not a good indicator of long-term output; the object of the Note is to move away from "what was seized?" to "how big was this drug business?" Consider two situations. In the first, the police find a large and sophisticated laboratory, able to produce substantial quantities of high-purity methamphetamine, but do not find any precursor chemicals--perhaps because they are stored elsewhere, perhaps because they had been used recently. In the second, the police find crude equipment, a few pots and beakers that could be used to make only small quantities of low-purity drug, but also find a supply of pseudoephedrine because they arrive just before a "cook." The prosecutor's approach yields no enhancement under Application Note 12 in the first case, but a potentially substantial enhancement in the second. That's backward and turns the Guidelines into an engine of disproportion.

Perhaps the parties and the district judge were led to their position by an omission in Application Note 12: "size or capability of any laboratory involved" is an incomplete way of describing the scale of an enterprise. Manufacturing facilities are rated in capacity per unit of time. An auto plant produces X cars per day; a generating station produces so many megawatts of electricity continuously. Should the district court take account of the lab's likely production over the last month, or year, or the life span of a normal drug-manufacturing ring, or the span of this particular conspiracy, or the period of the statute of limitations? Application Note 12 does not say, which makes it impossible for courts to treat equally dangerous drug rings the same way. Some judges are bound to select a short period (because estimates based on the recent past are more accurate), while others select a long period (because that best reflects the total output). The Sentencing Commission should amend Note 12 to incorporate an accounting period. The most logical period is the span of this conspiracy, which is "the offense" whose relevant conduct the judge is supposed to estimate. U.S.S.G. sec.1B1.3(a)(1)(A). But the incompleteness of Application Note 12 as it stands does not justify replacing an estimate of capacity (and thus of past production) with an estimate of raw materials at the time of arrest.