United States Court of Appeals
                            FOR THE EIGHTH CIRCUIT
                                  ___________

                              Nos. 05-4468/06-1354
                                  ___________

United States of America,              *
                                       *
            Appellee/Cross-Appellant, *
                                       *
      v.                               *
                                       *
Steven Spears,                         *
                                       *
            Appellant/Cross-Appellee. *
                                       * On remand for reconsideration
___________________                    * from the United States Supreme
                                       * Court.
American Civil Liberties Union         *
Foundation Drug Law Reform Project; *
American Civil Liberties Union of      *
Iowa; Douglas A. Berman; Michael M. *
O’Hear; David N. Yellen; David M.      *
Zlotnick; Federal Public Defender of   *
Iowa,                                  *
                                       *
            Amici on Behalf of         *
            Appellant/Cross-Appellee. *
                                   __________

                             Submitted: February 13, 2008
                                Filed: June 23, 2008
                                 ___________
Before LOKEN, Chief Judge, WOLLMAN, MURPHY, BYE, RILEY, MELLOY,
SMITH, COLLOTON, GRUENDER, BENTON, and SHEPHERD, Circuit Judges,
en banc.
                             ___________

RILEY, Circuit Judge.

       After our earlier decision in this matter, United States v. Spears, 

(8th Cir. 2006) (en banc) (Spears I), the United States Supreme Court vacated and
remanded the case for reconsideration in light of Kimbrough v. United States, 552
U.S. ___, 

, ___ L.Ed.2d ___ (2007). See Spears v. United States, 

(Jan. 7, 2008). Upon reconsideration, we vacate Section II(C) of Spears I,
and we reinstate the background and remainder of the Spears I analysis. Based upon
the clarifications outlined below, we again reverse and remand for resentencing.

I.     DISCUSSION
       A.     Clarification
       In Spears I, we discussed recent precedent in which the Third Circuit concluded
“district courts may consider the crack/powder cocaine differential in the Guidelines
as a factor, but not a mandate, in the post-Booker1 sentencing process.” See Spears

(quoting United States v. Gunter, 

, 249 (3d Cir.
2006). We expressly declined either to adopt or endorse the Third Circuit’s
determination. See 

Kimbrough, our dicta was wrong.

        In light of Kimbrough, we now expressly adopt the determination that, “under
Booker, the cocaine Guidelines, like all other Guidelines, are advisory only . . . .”

. “A district judge must include the Guidelines range in
the array of factors warranting consideration. The judge may determine, however,
that, in the particular case, a within-Guidelines sentence is ‘greater than necessary’ to


      1
       United States v. Booker, 

(2005).

                                          -2-
serve the objectives of sentencing.” 

U.S.C. § 3553(a) (2000 ed. and
Supp. V)). The Supreme Court therefore concluded, “In making that determination,
the judge may consider the disparity between the Guidelines’ treatment of crack and
powder cocaine offenses.” 

Holding
       In Spears’s case, we did not need either to adopt or endorse the proposition
outlined in Gunter and now Kimbrough because the district court did not just
“consider the disparity between the Guidelines’ treatment of crack and powder
cocaine offenses,” 

, in conjunction with a proper
§ 3553(a) analysis. Rather, the district court substituted its own ratio, and did not
conduct any additional analysis under § 3553(a). The district court declared:

      [O]nce I made the decision I was going to vary on the [United States v.
      Perry, 

(D. R.I. 2005)] crack ratio basis, I didn’t really
      look at the other § 3553(a) factors because to me it was a moot question.
      I’ll only look at that issue should this sentencing be reversed on the
      20-to-1 ratio.

Spears 

(internal alteration omitted). The district court therefore
“impermissibly varied by replacing the 100:1 quantity ratio inherent in the advisory
Guidelines range with a 20:1 quantity ratio.” 

(emphasis added).
Additionally, “the district court did not vary from the advisory Guidelines range based
on an individualized, case-specific evaluation of the facts or of the defendant.” 

the overall goals of sentencing under § 3553(a) and conducting an
individualized assessment based upon the particular circumstances of a defendant’s
case, a district court may determine the 100:1 quantity ratio results in a harsher
sentence than necessary. See 

. However, the district
court may not categorically reject the ratio set forth by the Guidelines.




                                         -3-
        This determination is directly in line with the Third Circuit’s reasoning in
Gunter, where the court explained, “we do not suggest (or even hint) that the Court
categorically reject the 100:1 ratio and substitute its own, as this is verboten.” 

(emphasis added). Nothing in Kimbrough suggests the district court
may substitute its own ratio for the ratio set forth in the Guidelines. Indeed, the
Supreme Court in Kimbrough explained “the court did not purport to establish a ratio
of its own.” 

.

II.    CONCLUSION
       We again affirm Spears’s conviction, and reverse Spears’s sentence and remand
for resentencing consistent with this opinion.

MURPHY, Circuit Judge, concurring.

       The Congressional sentencing ratio between powder and crack cocaine offenses
was incorporated into the sentencing guidelines by the first United States Sentencing
Commission. Although the Commission recently made some adjustments for crack
sentences in its 2007 amendments by reducing base offense levels, the underlying
guideline ratio between powder and crack remains the same. As an author2 of the
2002 Commission report to Congress that recommended reducing the disparity in
cocaine sentencing, I am aware of the large amount of evidence cited there which
undercut any continued validity for the policy reasons on which the disparate ratio
was originally based. In its decision in Kimbrough, the Supreme Court cited the
Commission's report and authorized sentencing judges to vary downward from the
advisory guidelines in crack cases on the basis of unwarranted 

. Since the Court's opinion in Kimbrough dealt only with guideline sentencing
(as did its other recent sentencing decisions in United States v. Booker, 

      2
       Chair of the United States Sentencing Commission, 1999-2004.

                                           -4-
(2005), and Gall v. United States, 

(2007)), the much criticized ratio
remains in place for all sentences affected by statutory mandatory minimums.

       Sentencing courts are faced with a dilemma in trying to fashion fair and
equitable sentences which are appropriate for the individual defendant, but which do
not create unwarranted disparities with sentences for other defendants and other
courts. See 18 U.S.C. § 3553(a)(6). Today's decision of our divided court reflects that
dilemma with thoughtful opinions on both sides of the issue, but a satisfactory
solution to the perceived unfairness of the great disparity in powder and crack
sentencing can only be accomplished by Congress. Only Congress has the power to
alter the cocaine sentencing ratio in the statutes which in turn affects the advisory
guidelines. Although legislative proposals for change have been introduced, Congress
has not yet been willing to act for change.

      One of the most basic principles of the Sentencing Reform Act of 1984, 18
U.S.C. § 3551 et seq., and a major reason for its bipartisan support was to reduce
sentencing disparity, to ensure that each federal judge would not set his or her own
individual sentencing policy. See S. Rep. No. 98-225, at 41-46 (1983); see also
United States v. Mickelson, 

, 1053-55 (8th Cir. 2006) (discussing
reduction of sentencing disparity as rationale for Sentencing Reform Act). The
approach advocated by the dissent would have just such a consequence of increasing
sentencing disparity, with the potential for each judge to develop his or her cocaine
policy ratio with dramatically different results within and between districts.

       The Supreme Court counseled in Kimbrough that a sentencing court should not
operate in a vacuum and "must take account of sentencing practices in other 

, and that it should also consider the disparate impact of the statutory
mandatory minimums as it weighs the "§ 3553(a) factors and any unwarranted
disparity created by the crack/powder ratio itself." 

suggesting that
a sentencing judge may develop his or her own policy ratio for the range of cocaine

                                          -5-
cases as the dissent proposes, in Kimbrough the Court approved a sentencing approach
based on the "particular circumstances" of the defendant's case. 

        The first step for the sentencing court in cocaine cases still is "properly
calculating and considering the advisory Guidelines range"; after that the court should
proceed to consider the individual circumstances of the crime and the defendant with
reference to the § 3553(a) factors. 

Recognizing the unfairness that the 100
to 1 ratio can have, the Court authorized district judges "to conclude when sentencing
a particular defendant that the crack/powder disparity yields a sentence 'greater than
necessary' to achieve § 3553(a)'s purposes." 

did not, however,
authorize district judges to develop their own policy ratio to be applied in every
cocaine case. In fact, the Court approvingly noted that the district court it was
reviewing

      did not purport to establish a ratio of its own. Rather, it appropriately
      framed its final determination in line with § 3553(a)'s overarching
      instruction to impose a sentence sufficient, but not greater than necessary
      [. . .] to do justice in this case.

(emphasis added). The role of the sentencing judge is not the
same as that of a legislative policymaker. The sentencing focus must be on the
"particular defendant" rather than on a whole category of cases. See 

reasons, I concur in Judge Riley's opinion. Perhaps the specter of
myriad cocaine sentencing policies with uneven results across the country will move
Congress to renew its interest in this critical subject which affects people's lives in a
way many perceive to be inequitable and unjust.




                                          -6-
COLLOTON, Circuit Judge, with whom WOLLMAN, BYE, MELLOY and SMITH,
Circuit Judges, join, dissenting.

       In Kimbrough v. United States, 

(2007), the Supreme Court
declared that “it would not be an abuse of discretion for a district court to conclude
when sentencing a particular defendant that the crack/powder disparity yields a
sentence ‘greater than necessary’ to achieve § 3553(a)’s purposes, even in a mine-run
case.” 

(emphasis added). The Court thus established that even when a
particular defendant in a crack cocaine case presents no special mitigating
circumstances – no outstanding service to country or community, no unusually
disadvantaged childhood, no overstated criminal history score, no post-offense
rehabilitation – a sentencing court may nonetheless vary downward from the advisory
guideline range. The court may do so based solely on its view that the 100-to-1 ratio
embodied in the sentencing guidelines for the treatment of crack cocaine versus
powder cocaine creates “an unwarranted disparity within the meaning of § 

, and is “at odds with § 3553(a).” 

The only fact necessary
to justify such a variance is the sentencing court’s disagreement with the guidelines
– its policy view that the 100-to-1 ratio creates an unwarranted disparity.3


      3
       Effective November 1, 2007, an amendment to the guidelines reduced the base
offense level associated with each quantity of crack by two levels, thus resulting in a
reduced crack/powder ratio that varies by quantity. The Supreme Court described this
as a “modest amendment,” observed that the amended guidelines “still produce
sentencing ranges keyed to the mandatory minimums” from the Anti-Drug Abuse Act
of 1986, and quoted the Sentencing Commission’s view that it is “‘only . . . a partial
remedy’ for the problems generated by the crack/powder disparity,” with a
“comprehensive solution” attainable only through “appropriate legislative action by
Congress.” 

& n.10. It seems unlikely, therefore, that
the Court would view the amended crack cocaine guideline as exemplifying “the
Commission’s exercise of its characteristic institutional role,” such that “closer
review” of a sentencing judge’s decision to vary from the amended guideline “may
be in order.” 

; cf. 

(Scalia, J., concurring).

                                         -7-
       This appeal presents the question whether a district court, in varying from the
advisory guidelines because the 100-to-1 ratio results in a sentence that is “greater
than necessary,” may determine the extent of the variance by reference to the
sentencing range that would apply if the guidelines used a different crack/powder ratio
– in this case, 20-to-1. Our court today says that a district court may not do so. There
is at least superficial appeal to this conclusion, based largely on two passages in
Kimbrough. The district court in that case, the Court observed, “did not purport to
establish a ratio of its own,” and “homed in on the particular circumstances of
Kimbrough’s 

. But Kimbrough juxtaposed a sentencing court
that “purport[s] to establish a ratio of its own” with a court that “appropriately
frame[s] its final determination in line with § 3553(a)’s overarching instruction to
‘impose a sentence sufficient, but not greater than necessary’ to accomplish the
sentencing goals advanced in § 3553(a)(2).” 

Court did not hold that
a district court is forbidden to consider an alternative crack/powder ratio as a means
of determining what sentence is “sufficient, but not greater than necessary” to comply
with the purposes of § 3553(a)(2); that question was not presented. On reflection, I
cannot identify anything about a sentencing court’s consideration of an alternative
ratio for that purpose that conflicts with § 3553(a) or with the Court’s remedial
guidance in the wake of United States v. Booker, 

(2005). To the
contrary, permitting a sentencing court to proceed in this fashion promotes the
statutory purposes of honesty, uniformity, and transparency in sentencing.

       There can be no doubt that the district court acted within its discretion by
granting a downward variance based on its view that the 100-to-1 ratio resulted in a
sentence greater than necessary to achieve the purposes of § 3553(a). Kimbrough says
this is allowed, even in a mine-run case. Nor, given the impermissibility of
proportionality review after Gall v. United States, 

(2007), is there
serious doubt that the district court’s considerable discretion would allow it to impose
a sentence of 240 months in this case, if it merely announced its view that a variance
of that degree was necessary to account for disparity caused by the 100-to-1

                                          -8-
crack/powder ratio. The remaining question is whether the court impermissibly relied
on the advisory range that would be indicated by a 20-to-1 crack/powder ratio as the
specific justification for the extent of the variance. The court says this is “verboten,”
but never really explains why this approach conflicts with § 3553(a) or the Sentencing
Reform Act. I believe the better view is that the statutory purposes of “honesty,”
“uniformity,” see 

, and transparency, see Rita v. United States,

, 2468-69 (2007), are actually furthered by the district court’s
approach and frustrated by the court’s contrary rule.

        If a district court is not permitted to make its decision about the extent of a
variance by reference to a different crack/powder ratio, then how should the court
proceed? Rita says a sentencing court should articulate reasons for its decision, so that
reviewing courts and the public are assured that the sentencing process is a “reasoned
process.” 

The district court’s explanation for varying to the degree
indicated by a 20-to-1 ratio surely is a reasoned basis for selecting the final sentence
in this case. Everyone can understand how and why the court arrived at the extent of
the variance. The district court, incorporating United States v. Perry, 

, 301-03 (D.R.I. 2005), see S. Tr. 32, relied on the same studies and criticisms of
the 100-to-1 ratio discussed favorably in Kimbrough, including the 2002 Sentencing
Commission report that recommended a 20-to-1 ratio. See 

, 575-76 (“[T]he [district] court alluded to the Sentencing Commission’s
reports criticizing the 100-to-1 ratio,” and “properly . . . accorded weight to the
Sentencing Commission’s consistent and emphatic position that the crack/powder
disparity is at odds with § 3553(a).”). How is the sentencing process improved if the
court is forbidden to rely consistently on the detailed analysis of the Sentencing
Commission to determine the magnitude of a variance? The goal of a reasoned
sentencing process is not advanced by requiring a sentencing court instead to make
a seat-of-the-pants judgment in each mine-run case about how much to vary from the
advisory guidelines based solely on the “crack/powder disparity.” 

                                          -9-
       Our court’s decision also seems likely to undermine the Sentencing Reform
Act’s interest in uniformity. To be sure, the decisions in Booker, Kimbrough, and
Gall make it less likely that the system will achieve one type of uniformity sought by
Congress: similar sentences for defendants with similar offense conduct and similar
criminal histories who are sentenced in different courts. Some judges now will elect
to vary from the advisory guidelines on crack cocaine, e.g., United States v. McGhee,

, 1051 (8th Cir. 2008) (per curiam), while others will apply the ratios
in the advisory guidelines, e.g., United States v. Moore, 

, 580 (8th Cir.
2008), and the deferential appellate review described in Gall and Kimbrough will not
iron out these differences. See 

(“We cannot and do not claim
that use of a ‘reasonableness’ standard will provide the uniformity that Congress
originally sought to secure.”).4 But there is still an interest in uniform treatment of
similarly-situated defendants by the same sentencing judge. If a particular judge
sentences numerous defendants in mine-run crack cocaine cases, and if that judge
holds the policy view that the crack/powder disparity yields a sentence greater than


      4
       I fully agree with Judge Murphy’s observation that “one of the most basic
principles of the Sentencing Reform Act of 1984” was that each individual judge
would not set his or her own sentencing policy. Ante, at 5 (concurring opinion); see
United States v. Maloney, 

, 669 (8th Cir. 2006) (observing that “approval
of the district court’s analysis in this case, and the degree of variance adopted, has
great potential to result in excessive sentencing disparities, as other district judges
apply their own personal sentencing philosophies”). But it is not “the approach
advocated by [this] dissent” that would permit each judge to set “his or her own
individual sentencing policy” in cocaine cases, and have the “consequence of
increasing sentencing disparity.” Ante, at 5. As our court’s post-Kimbrough decisions
already illustrate, it is the decision in Kimbrough itself that will have that effect. This
dissent simply acknowledges the reality that Kimbrough permits district judges to
apply their individual policy views in cocaine sentencing, 

(quoting the government’s concession that “courts may vary [from Guidelines ranges]
based solely on policy considerations, including disagreements with the Guidelines”),
and declines to prohibit judges from selecting one reasonable way of applying those
views.

                                           -10-
necessary to achieve the purposes of § 3553(a) in a mine-run case, see 

, then the sentencing judge may well (and hopefully would) desire to
treat each defendant the same in that respect. The judge should be permitted to use
a consistent methodology to achieve a consistent degree of variance in similar cases.5

       The decision in this case places the conscientious sentencing judge in a
quandary. If the opinion means that a court may not even consider the guideline range
that would result from an alternative crack/powder ratio, then the sentencing judge
who wishes to vary from the guidelines based on the crack/powder differential is
deprived of a reliable and analytically rigorous means to ensure similar sentences for
similar defendants in the same court. If, on the other hand, the opinion permits the
court to consider the result that would be derived from an alternative crack/powder
ratio so long as the judge does not expressly rely on it, then the sentencing court must
refrain from providing a complete explanation for its sentence or, worse, engage in
subterfuge when setting forth its rationale. A system that encourages this sort of
behavior is at odds with the goal of the Sentencing Reform Act to promote honesty
in sentencing.

       Allowing sentencing courts to rely on the range indicated by an alternative
crack/powder ratio also helps the sentencing process to evolve. See 

. Booker, Gall, and Kimbrough have resulted in a sentencing system that
differs significantly from the system enacted by Congress. The statutes established
mandatory sentencing guidelines and de novo appellate review of many departures,

      5
       Kimbrough suggests that a sentencing court should consider arguments from
the government that the crack/powder ratio should be evaluated differently in a
particular case because of “sentencing practices in other courts” or “‘cliffs’ resulting
from statutory minimum sentences.” 

. As a procedural matter,
the government raised no such argument in this case, see S. Tr. 24-26, and, as a matter
of substantive reasonableness, the Supreme Court did not indicate whether or to what
extent these considerations must override uniform treatment of crack cocaine
defendants by a particular sentencing judge.

                                         -11-
with basic policy decisions made largely by the Sentencing Commission rather than
individual sentencing courts. The Court’s post-Booker remedial direction provides
for advisory sentencing guidelines and deferential abuse-of-discretion review of
departures and variances, thus transferring a good measure of authority to the
individual sentencing courts. As a consequence, Congress will be faced with choices
in the future about whether to retain the new system or to legislate a different system
that more closely approximates the former statutory plan. See 

(“The ball now lies in Congress’ court. The National Legislature is equipped to devise
and install, long term, the sentencing system, compatible with the Constitution, that
Congress judges best for the federal system of justice.”); 

(Souter, J., concurring) (“After Booker’s remedial holding, I continue to think that the
best resolution of the tension between substantial consistency throughout the system
and the right of jury trial would be a new Act of Congress: reestablishing a system
of statutory mandatory sentencing guidelines (though not identical to the original in
all points of detail), but providing for jury findings of all facts necessary to set the
upper range of sentencing discretion.”).

       The ability of Congress (and the Sentencing Commission) to evaluate the
system in place after Booker, Gall, and Kimbrough will be enhanced if sentencing
courts are permitted to articulate and apply consistent methodologies to justify their
variances from the advisory guidelines. Congress should know if some district judges
decline to vary from the 100-to-1 ratio (or from the ratios set forth in the guidelines
as amended in November 2007), while others vary in mine-run cases, based solely on
the crack/powder differential, to sentences consistent with, say, a 20-to-1 ratio, a 5-to-
1 ratio, or a 1-to-1 ratio. This information would help Congress to make an
enlightened decision about how to respond to post-Booker sentencing practices. The
Sentencing Commission would benefit from that information as it considers whether
to modify the advisory guidelines in response to the sentencing practices of individual
judges. See 

. The system works best if the process is



                                          -12-
transparent. There is reason to believe, however, that our court’s decision will serve
merely to obfuscate the sentencing process without changing the result in any case.

       In opposition to these points, the majority says that the sentencing judge may
not “categorically” reject the 100-to-1 ratio, ante, at 3, and the concurring opinion
suggests that the judge here acted as a “legislative policymaker” rather than as a
sentencing judge. Ante, at 6. But the district court did not purport to act as Congress
or even a “junior-varsity Congress.” The judge did not fashion a rule that must be
followed by any other judge. He did not bind even himself to apply the same view of
the crack/powder ratio in future cases – although for the reasons stated, it may be quite
sensible for a judge to act consistently in that respect. We have before us only one
sentencing decision in one case involving one defendant. The objection to
“categorical” decisionmaking or “legislative policymak[ing],” therefore, seems to
echo the argument of the government that was rejected in Kimbrough – namely, that
a district court “may vary from the 100-to-1 ratio if it does so based on the
individualized circumstances of a particular case,” but that a court “may not give any
weight to its own view that the ratio itself is inconsistent with the § 3553(a) 

n.13 (internal quotation omitted). Kimbrough settled that a judge
may give weight to his or her own policy views in crack cocaine cases, and while this
would have been impermissible under the mandatory guidelines before Booker, it is
not inherently inconsistent with the judicial role. “[S]entencing is a field in which the
Judicial Branch long has exercised substantive or political judgment,” Mistretta v.
United States, 

, 396 (1989), and for more than a century prior to the
Sentencing Reform Act, individual sentencing judges made the same sort of policy
judgment that is embodied in the district court’s decision in this case.

       Finally, to the extent the court holds that the district court committed procedural
error within the meaning of 

, I respectfully disagree. The
district court began by accurately determining the advisory guideline range, applying
the 100-to-1 crack/powder ratio set forth in the guidelines. S. Tr. 22; see United

                                          -13-
States v. Spears, 

, 1174 (8th Cir. 2006) (en banc). The court considered
all arguments from the government for traditional upward departures under the
guidelines or for a sentence greater than the advisory range based on § 3553(a). See

. Indeed, the court even granted an upward departure based on
its view that Spears’s criminal history score under-represented the severity of his
criminal history. S. Tr. 21-22. The court then considered arguments from Spears,
based on § 3553(a), for a more lenient sentence than recommended by the advisory
guidelines. Incorporating the rationale of United States v. Perry, the district court
concluded, consistent with Kimbrough, that the advisory guideline range for Spears
resulted in a sentence “greater than required to reflect the purposes of sentencing as
outlined in § 3553(a),” due to the “Guideline disparity” in the treatment of powder
cocaine and crack cocaine. 

; see 

.

       In determining how much “greater than necessary” was the advisory sentence,
the court calculated what the advisory range would be if the guidelines applied a 20-
to-1 crack/powder ratio, as suggested by the Sentencing Commission in its 2002
report. See 

(“[T]he Court believes a 20:1 ratio
effectively meets the criteria of § 3553 and the objectives of sentencing policy. . . .
This conclusion is supported by the findings and recommendations of the Sentencing
Commission . . . .”). This analysis led the district court to vary downward from the
advisory range of 324 to 405 months’ imprisonment to the statutory minimum
sentence of 240 months’ imprisonment.              The government suggested no
“individualized” reason that this degree of variance was inappropriate for Spears.
Rather, it advanced essentially the same argument rejected in Kimbrough. See S. Tr.
25 (discussing United States v. Tabor, 

, 1058-61 (D. Neb. 2005),
aff’d, 

(8th Cir. 2006), vacated, 

(2008)). Because the
sentencing court concluded that accounting for a disparity in sentencing caused by the
100-to-1 ratio required a variance down to the statutory minimum, the court found it
unnecessary to decide whether there was merit to other arguments advanced by Spears

                                         -14-
for downward variances based on other § 3553(a) factors. S. Tr. 26-28. The district
court did not, as the court implies, ante, at 3, ignore other factors in § 3553(a) that
might have affected the sentence. It simply declined to resolve questions unnecessary
to the final sentencing decision, a practice that we have not heretofore condemned as
inconsistent with Booker or § 3553(a).

      For these reasons, I would affirm the judgment of the district court.
                      ______________________________




                                         -15-