Opinion
No. 06-2064.
Argued October 2, 2007.
Decided June 17, 2008.
Appeal from the United States District Court for the Western District of Wisconsin. No. 05-CR-0138-C. Barbara B. Crabb, Chief Judge.
Timothy M. O'Shea, Office of the United States Attorney, Madison, WI, for Plaintiff-Appellee.
Hannah V. Garst, Chicago, IL, for Defendant-Appellant.
Before FRANK H. EASTERBROOK, Chief Judge, DANIEL A. MANION, Circuit Judge, and ANN CLAIRE WILLIAMS, Circuit Judge.
ORDER
James Harper pleaded guilty to possessing crack with intent to distribute and was sentenced to 204 months. See 21 U.S.C. § 841(a)(1). He challenges his sentence on appeal, arguing that the district court miscalculated his drug quantity, and that the difference in offense levels for like quantities of crack and powder cocaine in the sentencing guidelines resulted in an unreasonable sentence. We vacate Harper's sentence and remand for resentencing in light of Kimbrough v. United States, ___ U.S. ___, 128 S.Ct. 558, 169 L.Ed.2d 481 (2007).
Harper's story begins with Larry McGee, who was arrested in August 2005 for selling nearly 30 grams of crack to an undercover agent over a five-month period. McGee decided to cooperate with authorities and named Harper as his main source, estimating that Harper had sold him one ounce of crack every three or four days for a year, totaling 90 to 121 ounces (2,551 to 3,430 grams). The next day McGee made a phone call to Harper and requested two ounces of crack. Harper agreed to meet McGee, and officers arrested Harper at the meeting point with 15.43 grams of crack in his possession.
Both McGee and Harper were charged in federal court. McGee pleaded guilty to possessing crack with intent to distribute. At some point prior to sentencing, McGee reduced his initial estimate of Harper's drug activity, claiming that investigators had misunderstood him during the interview and that he actually obtained less crack from Harper. Apparently, though, McGee never offered an alternative estimate or retracted his statement that Harper was his primary source for crack. Thus, McGee was sentenced on the basis of his initial estimate. McGee appealed his sentence, but this court granted his attorney's motion to withdraw and dismissed the appeal because of the absence of a nonfrivolous issue for appeal. United States v. McGee, 216 Fed.Appx. 580 (7th Cir. 2007). Meanwhile, the probation officer who prepared the presentence report in Harper's case acknowledged that McGee had recanted his initial estimate. Accordingly, the probation officer decided not to rely exclusively on the information from McGee and instead estimated that Harper had sold McGee one ounce of crack every week for eight months for a total of 907 grams; although the probation officer considered McGee's initial estimate, the estimated quantity was based largely on the undercover buys from McGee and a statement from Charles Benson, who told investigators that he was McGee's courier during those eight months. Thus, the probation officer concluded that Harper's relevant conduct included the 907 grams plus the 15 grams he possessed when arrested, totaling 922 grams.
At sentencing the district court found that Harper's drug quantity was at least 500 grams but less than 1.5 kilograms of crack. In making this finding, the court relied on the probation officer's recommendation, McGee's initial estimate, the speed with which Harper responded to McGee's buy request, Harper's previous drug-dealing activities, the undercover buys from McGee, and two previous incidents in which Harper was stopped with $568 and $23,931, respectively. The court acknowledged McGee's later recantation, but noted that McGee's earlier estimate formed the basis for his own sentence. Thus, the court found that Harper's base offense level was 36 under U.S.S.G. § 2D1.1(c)(2). Although Harper had two prior felony convictions for controlled substance offenses that made him a career offender with a base offense level of 34 under U.S.S.G. § 4B 1.1(b), his greater base offense level of 36 under § 2D1.1(c)(2) superseded the career offender figure. After a three-level downward adjustment for acceptance of responsibility, see U.S.S.G. § 3E1.1, Harper's total offense level of 33 and criminal history category of VI yielded an imprisonment range of 235 to 293 months.
Harper argued that the district court should impose a belowrange sentence because of what he characterized as an unwarranted disparity between crack and powder sentences. The court responded:
[I]f you want a discretionary sentence and it's based on the discrepancy between crack and powder, I simply don't agree that that is — If it were up to me, that would be a desirable difference. It would be desirable not to have a difference between powder and crack, but those decisions have been made by Congress.
. . . .
. . . I want to make it clear that I understand that I have the discretion to go below the guidelines, but what I'm saying is that I wouldn't go below the guidelines simply because of an argument that . . . there should not be a differential between crack and powder cocaine.
But the court did go below the range and imposed a sentence of 204 months, citing Harper's age and life expectancy.
On appeal Harper argues that the district court erred in relying on McGee's repudiated estimate of his crack purchases in determining the drug quantity. We review a district court's drug-quantity calculation for clear error, United States v. White, 360 F.3d 718, 720 (7th Cir. 2004), and will affirm absent "a definite and firm conviction that a mistake has been made," United States v. Fudge, 325 F.3d 910, 919-20 (7th Cir. 2003). The government must prove the facts underlying an upward adjustment by a preponderance of the evidence. United States v. Hines, 449 F.3d 808, 815-16 (7th Cir. 2006). Because the Federal Rules of Evidence do not apply at sentencing, the district court can consider a wide range of information — including hearsay — that might have been inadmissible at trial, so long as the information bears "sufficient indicia of reliability to support its probable accuracy." United States v. McEntire, 153 F.3d 424, 435 (7th Cir. 1998).
Harper argues that the district court erred by not resolving the inconsistency between McGee's earlier and later statements and by failing to explain its reasons for adopting his earlier estimate. Because of the danger that a recanted statement is unreliable, a sentencing court must adequately investigate its probable accuracy before accepting it. See United States v. Cross, 430 F.3d 406, 411 n. 2 (7th Cir. 2005). If a court decides to credit one statement over another inconsistent statement by the same witness, it should address the contradiction and articulate its reasons for doing so. United States v. Span, 170 F.3d 798, 803 (7th Cir. 1999); United States v. Duarte, 950 F.2d 1255, 1266 (7th Cir. 1991). Statements against penal interest are presumptively reliable, though, and a court can find sentencing facts against a defendant based on statements against penal interest made by others. See United States v. Johnson, 342 F.3d 731, 734 (7th Cir. 2003); United States v. Szakacs, 212 F.3d 344, 352-53 (7th Cir. 2000) (finding sentencing facts based on statements against penal interest made by co-defendants). Here the court acknowledged that McGee disavowed his initial estimate of Harper's drug activities but emphasized that McGee himself was sentenced on the basis of his earlier estimate. Additionally, in its statement of reasons, the court noted that McGee's earlier statements to investigators were made "without a proffer agreement, and ultimately to his own detriment." Finally, for sentencing purposes, the court held Harper responsible for only 500 grams of crack, less than 20% of McGee's lowest initial estimate. Although a richer description of the court's analysis would have been desirable at the sentencing hearing, the court adequately probed the government's evidence and explained its reasons for adopting McGee's earlier statements.
Harper's reliance on United States v. Robinson, 164 F.3d 1068 (7th Cir. 1999)"in which we suggested, but did not require, that a sentencing judge conduct an evidentiary hearing to question under oath a witness whose hearsay statements provide the only evidentiary support for the lion's share of the drug quantity attributed to the defendant — is misplaced. Id. at 1070. In Robinson we observed that a witness who provided uncorroborated, nonsensical statements could not be relied upon to calculate drug quantity. Id. at 1070-71. Robinson is distinguishable from this case, though, because McGee's earlier statements are not improbable, nor are they uncorroborated. Robinson merely demonstrates a preference for live testimony in the most extreme cases, but this preference does not warrant vacating Harper's sentence.
Although Harper emphasizes the role of McGee's testimony in the district court's drug-quantity determination, the court also relied upon the probation officer's recommendation, the speed with which Harper was able to respond to McGee's buy request, Harper's history of drug-dealing, the undercover buys from McGee (which Harper does not dispute he supplied to McGee), and two previous incidents in which Harper was stopped with $568 and $23,931, respectively, without a plausible explanation. Harper contends that his history of drug-related activity and the unexplained money cannot support the court's specific drug-quantity determination, citing United States v. Acosta, 85 F.3d 275 (7th Cir. 1996), and United States v. Beler, 20 F.3d 1428 (7th Cir. 1994). But those cases simply state that evidence of previous drugs sales, standing alone, does not corroborate specific drug-quantity amounts. See Acosta, 85 F.3d at 282-83; Beler, 20 F.3d at 1436-37; see also Cross, 430 F.3d at 411 (concluding that money and drugs seized from defendant's home and his prior arrests for drug-related activity corroborated witness testimony concerning quantity of drugs purchased by defendant). And here the evidence of previous drug-related activity is not alone; it is supported by McGee's statement, the undercover buys, the speed with which Harper responded to McGee's buy request, and the unexplained money. If McGee's statements provided the only basis for the court's drug-quantity calculation, this would be a closer case. See, e.g., Beler, 20 F.3d at 1430-37. But the sentencing court made its decision based upon a range of evidence, and Harper's argument that the court failed to conduct a sufficiently searching inquiry into the government's evidence is unpersuasive. Accordingly, the court did not commit clear error in its calculation of Harper's drug quantity.
Harper next argues that his sentence is unreasonable because the district court did not believe that it had the authority to impose a reduced sentence based on a disagreement with the 100:1 crack-to-powder ratio in the sentencing guidelines. At the time of Harper's sentencing, the law in this circuit was that a sentencing judge could not stray from the 100:1 ratio when calculating the correct guidelines range. See United States v. Taylor, 520 F.3d 746, 746-47 (7th Cir. 2008); United States v. Jointer, 457 F.3d 682, 686-88 (7th Cir. 2006). A judge could, however, consider criticism of the 100:1 ratio when determining the appropriate actual sentence in light of the factors set forth in 18 U.S.C. § 3553(a), but only insofar as that criticism was immediately relevant to the circumstances of an individual defendant. See Jointer, 457 F.3d at 686-88. But the Supreme Court recently announced that, even in a routine case, a judge "may consider the disparity between the Guidelines' treatment of crack and powder cocaine offenses" when determining whether a within-guidelines sentence is greater than necessary to achieve the goals of sentencing. Kimbrough v. United States, ___ U.S. ___, 128 S.Ct. 558, 564, 575, 169 L.Ed.2d 481 (2007). And we have since characterized the 100:1 ratio as "not a statutory dictate, but merely a judgment, entitled to respect but not to uncritical acceptance, made by the Sentencing Commission as an input into fixing guideline ranges for crack offenders." Taylor, 520 F.3d at 747.
Kimbrough issued after oral argument in this case, so we ordered the parties to submit supplemental briefs concerning its effect on Harper's appeal. Harper and the government agree that some of the district court's comments at sentencing are inconsistent with Kimbrovugh. The government conceded as much at oral argument, stating that Judge Crabb was "not so sure about the 100:1 ratio" but adhered to it nonetheless. Yet the government insists that Harper's sentence is reasonable because the district court imposed a below-guidelines sentence that happens to fall within the guidelines range he would have received as a career offender. This argument is unpersuasive. We have no reason to believe that the district court felt that the specific sentence imposed was appropriate regardless of the applicable guidelines range.
Because we cannot discern from the record whether the district court would have imposed a lesser sentence had it known that it was free to disagree with the 100:1 ratio, a remand is appropriate. The sentencing transcript admits both possibilities. And this appeal is distinguishable from United States v. White, 519 F.3d 342 (7th Cir. 2008), in which we decided that a remand was unnecessary "given the lower court's firm statement that it would have imposed the same sentence even if there were no Guidelines, thus making clear that the crack/powder disparity reflected in the Guidelines in no way affected the court's sentencing decision." Id. at 349 (emphasis added). We cannot be so certain here. Therefore we VACATE Harper's sentence and EEMAND for resentencing in light of Kimbrough.
We also note that although the district court "shall apply the guidelines . . . that were in effect on the date of the previous sentencing of the defendant prior to the appeal," 18 U.S.C. § 3742(g)(1), "current guidelines can be used to help guide a judge who is minded to sentence the defendant outside the applicable guidelines range," United States v. Sriram, 482 F.3d 956, 961 (7th Cir. 2007), vacated on other grounds, ___ U.S. ___, 128 S.Ct. 1134, 169 L.Ed.2d 946 (2008). Today U.S.S.G. § 2D1.1(c) prescribes a base offense level of 34 — as opposed to 36 — for a drug quantity of at least 500 grams but less than 1.5 kilograms of crack. Id. § 2D1.1(c)(3) (2007). Harper also faces a base offense level of 34 because he was a career offender, and Kimbrough and the revised sentencing guidelines had no effect on his career offender status.