NONPRECEDENTIAL DISPOSITION
To be cited only in accordance with
Fed. R. App. P. 32.1
United States Court of Appeals
For the Seventh Circuit
Chicago, Illinois 60604
Argued October 2, 2007
Decided June 17, 2008
Before
FRANK H. EASTERBROOK, Chief Judge
DANIEL A. MANION, Circuit Judge
ANN CLAIRE WILLIAMS, Circuit Judge
No. 06‐2064
UNITED STATES OF AMERICA, Appeal from the United States District
Plaintiff‐Appellee, Court for the Western District of
Wisconsin.
v.
No. 05‐CR‐0138‐C
JAMES W. HARPER,
Defendant‐Appellant. Barbara B. Crabb,
Chief Judge.
O R D E R
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, 128 S. Ct. 558 (2007).
No. 06‐2064 Page 2
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 F. App’x 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. § 4B1.1(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
No. 06‐2064 Page 3
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 below‐range 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
No. 06‐2064 Page 4
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
No. 06‐2064 Page 5
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, No. 06‐4123,
2008 WL 782739, at *1 (7th Cir. Mar. 26, 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, 128 S. Ct. 558, 564, 575 (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, 2008 WL 782739, at *1.
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 Kimbrough. 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.
No. 06‐2064 Page 6
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 REMAND 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, 128 S. Ct. 1134 (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.