United States v. Aaron Eric Williams , 472 F.3d 835 ( 2006 )

Menu:

456 F.3d 1353

UNITED STATES of America, Plaintiff-Appellant,
v.
Aaron Eric WILLIAMS, Defendant-Appellee.

No. 05-13205.

United States Court of Appeals, Eleventh Circuit.

July 21, 2006.

Peggy Morris Ronca, Jacksonville, FL, for U.S.

George Allen Couture and Rosemary T. Cakmis, Fed. Pub. Defenders, Clarence W. Counts, Jr., Asst. Fed. Pub. Def., Orlando, FL, for Wiliams.

Appeal from the United States District Court for the Middle District of Florida.

Before BLACK, PRYOR and COX, Circuit Judges.

BLACK, Circuit Judge:

1
The Government appeals Aaron Eric Williams' 204-month sentence imposed for possessing crack cocaine with intent to distribute, in violation of 21 U.S.C. § 841(a)(1). We conclude the district court erred in mitigating Williams' sentence based on (1) its generalized disagreement with Congress's policy of punishing crack cocaine offenders more severely than powder cocaine offenders through the 100-to-1 crack-to-powder drug quantity ratio, (2) its generalized disagreement with the Guidelines career offender enhancement, U.S.S.G. § 4B1.1, and (3) its belief that the Government manipulated Williams' sentence by arranging to purchase crack cocaine instead of powder cocaine. We further conclude that these errors are not harmless because Williams has failed to meet his burden of showing that, considering the record as a whole, the errors did not have a substantial effect on the court's choice of sentence. We, therefore, vacate the sentence as unreasonable and remand for the district court to sentence Williams based on the individualized facts and circumstances of Williams' case bearing upon the sentencing considerations enumerated in 18 U.S.C. § 3553(a).

I. BACKGROUND

2
In April 2003, a confidential informant told Drug Enforcement Administration (DEA) agents and Osceola County Sheriff's Deputies that Williams was selling crack cocaine from his home in Kissimmee, Florida. Based on this tip, DEA agents initiated a sting operation to purchase crack cocaine from Williams with the help of the confidential informant. In three separate transactions in April, May, and July of 2003, an undercover DEA agent purchased a total of 34.8 grams of crack cocaine from Williams.1 A grand jury subsequently indicted Williams for one count of possessing crack cocaine with intent to distribute, in violation of § 841(a)(1) and (b)(1)(C), and two counts of possessing five or more grams of crack cocaine with intent to distribute, in violation of § 841(a)(1) and (b)(1)(B)(iii). On the day of trial, but before the jury was empaneled, the Appellant filed a notice pursuant to § 851(a) of its intent to rely on Williams' prior felony drug convictions to seek enhanced punishment. On February 1, 2005, the jury found Williams guilty of all three counts in the indictment, making specific findings of the drug quantities involved in each transaction.

3
At the sentencing hearing, the district court adopted, without objection, the facts and Guidelines calculation set out in Williams' presentence investigation report (PSI). According to the PSI, Williams has an extensive criminal history, which began in 1991, when he was sixteen years old. His scored criminal history includes five convictions for possession of crack cocaine (two of which were committed with the intent to sell), four convictions for offenses involving possession of marijuana, one conviction for possession with intent to sell MDMA (a controlled substance also known as Ecstasy), and two convictions for battery of two women, one of whom was pregnant at the time. As a result, Williams had a total of 22 criminal history points, well above the 13 points needed to place him in the highest criminal history category, category VI.

4
The PSI calculated Williams' base offense level at 28, pursuant to U.S.S.G. § 2D1.1(c)(6).2 Williams' multiple prior felony convictions for drug offenses qualified him as a career offender under § 4B1.1(a). Because the statutory maximum for his offenses was life imprisonment, see 21 U.S.C. § 841(b)(1)(B), § 4B1.1 enhanced Williams' offense level to 37, see U.S.S.G. § 4B1.1(b)(A). Williams' advisory Guidelines range was, therefore, 360 months to life imprisonment.

5
The district court sentenced Williams to 204 months' incarceration on all counts, to run concurrently. The district court explained its choice of sentence at the sentencing hearing and in its subsequent "Memorandum Sentencing Opinion," which is published at United States v. Williams, 372 F.Supp.2d 1335 (M.D.Fla.2005). In explaining the sentence, the district court purported to track the sentencing factors outlined in 18 U.S.C. § 3553(a).

6
At the sentencing hearing, the district court expressed "concern about the discrepancy between powder and crack cocaine"—referring to Congress's policy of punishing crack cocaine offenders more severely than powder cocaine offenders— and viewed the cocaine sentencing disparity as bearing on "the nature and circumstances of the offense." See id. § 3553(a)(1). The district court suggested the harsher penalties for crack offenses "smacks of discrimination against blacks, African-Americans." It said crack cocaine and powder cocaine are "the same drug" with the same effects and that the different form of the drug "has never justified the substantial disparity in sentences." The court also condemned the DEA's conduct in investigating Williams, characterizing the sting operation as "basically set up by the Government to snare Mr. Williams." The district court stated, without elaboration, that this was also relevant to "the nature and circumstances of the offense."

7
In considering "the seriousness of the offense" and the need for the sentence "to provide just punishment," see id. § 3553(a)(2)(A), the district court disagreed with what it called the "compounding effects" of the Guidelines, which it said causes "incongruity and unjust results." The district court was referring to the interplay between the Guidelines career offender enhancement, U.S.S.G. § 4B1.1, and the enhanced statutory maximum provided in 21 U.S.C. § 841(b)(1)(B) for offenders with at least one prior felony drug conviction. The enhanced offense levels set out in § 4B1.1(b) for career offenders vary depending on the statutory maximum for the offense of conviction. See U.S.S.G. § 4B1.1(b). Because Williams had a prior conviction for a felony drug offense, his statutory maximum sentence under § 841(b)(1)(B) increased from 40 years' imprisonment to life imprisonment. See 21 U.S.C. § 841(b)(1)(B). The § 841(b)(1)(B) "enhancement" had the corresponding effect under § 4B1.1(b) of increasing Williams' offense level from 28 to 37. The district court saw the interaction between § 841(b)(1)(B) and § 4B1.1(b) as a "totally inappropriate way to consider the individual nature of an offense or a defendant's individual background" and stated it was "not going to do it."

8
With respect to the "history and characteristics of the defendant," see id. § 3553(a)(1), the district court noted Williams' "long history of criminal conduct" and stated Williams was, therefore, "going to be spending a bunch of time in prison." Turning to the need for the sentence "to afford adequate deterrence," see id. § 3553(a)(2)(B), the district court opined that sending a "petty drug dealer" like Williams to prison for 30 years was not the way to deter the illegal drug trade.

9
In its subsequent "Memorandum Sentencing Opinion," the district court again addressed the cocaine sentencing disparity. The district court stated it was "mindful of the substantial criticism" the disparity had garnered and that evidence suggested the disparity had a "discriminatory impact on African Americans of whom Williams is one." Williams, 372 F.Supp.2d at 1339 n. 8. The district court's disdain for the disparity factored into its choice of sentence in another way as well. Echoing its earlier statement that the DEA had snared Williams, the district court concluded a Guidelines sentence would not "promote respect for the law," see id. § 3553(a)(2)(A), because the DEA arranged a sting purchase of crack cocaine instead of powder cocaine to obtain a longer prison sentence. Williams, 372 F.Supp.2d at 1339. Without referring to any facts in the record, the district court found that a powder cocaine sale would have been consistent with Williams' prior drug sales. Id. To highlight the injustice it perceived in the Government's decision to purchase crack cocaine from Williams, the district court compared Williams' Guidelines range with the range applicable to defendants who, though not career offenders, have a criminal history category of VI and are convicted of selling the same quantity of powder cocaine.3 Id. The district court then explained, as it did at the sentencing hearing, that a Guidelines sentence was inappropriate for another reason, namely its disagreement with the career offender provision in § 4B1.1. Id. According to the district court, Williams' past criminal conduct was already accounted for in his category VI criminal history. Id. The "layering of Chapter 4 enhancements," the district court reasoned, "results in a double-compounding effect, increasing Williams' minimum guideline sentence . . . in light of the same criminal conduct." Id. The district court stated this "arbitrary compounding results in a guideline sentence much greater than that necessary" to achieve the sentencing goals enumerated in § 3553(a). Id.

10
Finally, the district court explained that although "Williams is a low-level drug dealer . . . convicted of selling relatively small amounts of crack cocaine," the "substantial term" of 204 months' incarceration was warranted by "the circumstances (crack versus powder cocaine) and Williams' long history of selling illegal drugs." Id. The Appellant subsequently appealed Williams' sentence as unreasonable.

II. STANDARD OF REVIEW

11
Our review of sentences after United States v. Booker, 543 U.S. 220, 125 S.Ct. 738, 160 L.Ed.2d 621 (2005), has two components. First, we consider challenges to the district court's calculation of the advisory Guidelines range. Second, we review the sentence for reasonableness. See United States v. Williams, 435 F.3d 1350, 1353 (11th Cir.2006) (determining whether the district court correctly calculated the Guidelines range before evaluating the reasonableness of the sentence).

A. Guidelines Calculation

12
"[A]s was the case before Booker, the district court must calculate the Guidelines range accurately." United States v. Crawford, 407 F.3d 1174, 1179 (11th Cir.2005). We review the district court's interpretation of the Guidelines de novo and accept its factual findings unless clearly erroneous. United States v. Jordi, 418 F.3d 1212, 1214 (11th Cir.2005). An error in the district court's calculation of the advisory Guidelines range warrants vacating the sentence, unless the error is harmless. See United States v. Scott, 441 F.3d 1322, 1329 (11th Cir.2006) (applying harmless error review to Guidelines miscalculation). A Guidelines miscalculation is harmless if the district court would have imposed the same sentence without the error. See id.

B. Reasonableness

13
If the Guidelines calculation is correct, or if the miscalculation is harmless, we consider whether the sentence is reasonable. When reviewing a sentence for reasonableness, we must evaluate whether the sentence achieves the purposes of sentencing as stated in 18 U.S.C. § 3553(a).4 United States v. Talley, 431 F.3d 784, 788 (11th Cir.2005). This evaluation must be made having "regard for . . . the factors to be considered in imposing a sentence, as set forth in [§ 3553(a)]; and . . . the reasons for the imposition of the particular sentence, as stated by the district court pursuant to the provisions of section 3553(c)." Booker, 543 U.S. at 261, 125 S.Ct. at 765 (quoting 18 U.S.C. § 3742(e)(3) (1994)); see also Williams, 435 F.3d at 1355 (stating that "when reviewing for reasonableness, we must consider both the § 3553(a) factors and the reasons given by the district court"). The party challenging the sentence bears the burden of establishing the sentence is unreasonable in light of the § 3553(a) factors. Talley, 431 F.3d at 788.

14
In order to tailor our reasonableness standard of review to the issues in this case, we must first identify the challenges Appellant makes to the reasonableness of Williams' sentence. First, Appellant argues the sentence is unreasonable, regardless of length, because it resulted from the district court's consideration of impermissible factors. Second, Appellant contends in the alternative that, even if Williams' sentence was not affected by legal errors, the length of the sentence is unreasonable because the record does not support a deviation from the Guidelines range.

1. Reasons for the Sentence

15
With respect to Appellant's first argument, we agree that a sentence can be unreasonable, regardless of length,5 if the district court's selection of the sentence was substantially affected by its consideration of impermissible factors. This is so, because our reasonableness inquiry is not confined to reviewing whether there are facts and circumstances found in the record that would justify the length of the sentence imposed. See United States v. Webb, 403 F.3d 373, 383 (6th Cir.2005) ("[R]eview for reasonableness is not limited to the length of the sentence." (quotation omitted)). The reasons given by the district court for its selection of a sentence are important to assessing reasonableness. United States v. Jimenez-Beltre, 440 F.3d 514, 519 (1st Cir.2006) (en banc) (stating the emphasis in reviewing the reasonableness of a sentence "will be on the provision of a reasoned explanation, a plausible outcome and—where these criteria are met— some deference to different judgments by the district judges on the scene"). A sentence based on an improper factor fails to achieve the purposes of § 3553(a) and may be unreasonable, regardless of length. United States v. Moreland, 437 F.3d 424, 434 (4th Cir.2006) ("A sentence may be substantively unreasonable if the court relies on an improper factor or rejects policies articulated by Congress or the Sentencing Commission.").

16
Because the party challenging the sentence bears the burden of demonstrating that the sentence is unreasonable, the party challenging the sentence bears the initial burden of establishing that the district court considered an impermissible factor at sentencing. Whether a factor is impermissible is a question of law that we will review de novo. See United States v. Robinson, 935 F.2d 201, 203 (11th Cir. 1991) ("The application of the law to sentencing issues is subject to de novo review."). If such an error exists and was preserved for appeal, we will vacate the sentence and remand, unless the error is harmless. See 28 U.S.C. § 2111; Fed. R.Crim.P. 52(a). The party defending the sentence has the burden of establishing the error was harmless. See, e.g., United States v. Mathenia, 409 F.3d 1289, 1291-92 (11th Cir.2005).

17
In considering whether an error is harmless, we apply our traditional harmless error standard: "A ``non-constitutional error' is harmless if, viewing the proceedings in their entirety, a court determines that the error did not affect the sentence, ``or had but very slight effect.' If one can say ``with fair assurance . . . that the sentence was not substantially swayed by the error,' the sentence is due to be affirmed even though there was error." Id. at 1292 (quoting United States v. Hornaday, 392 F.3d 1306, 1315-16 (11th Cir.2004) (quoting Kotteakos v. United States, 328 U.S. 750, 762, 763, 66 S.Ct. 1239, 1246, 1248, 90 L.Ed. 1557 (1946))); see also Williams v. United States, 503 U.S. 193, 203, 112 S.Ct. 1112, 1120-21, 117 L.Ed.2d 341 (1992) (holding that once the party challenging the sentence shows the district court relied on an invalid factor at sentencing, "a remand is appropriate unless the reviewing court concludes, on the record as a whole, that the error was harmless, i.e., that the error did not affect the district court's selection of the sentence imposed"); United States v. Paley, 442 F.3d 1273, 1278 (11th Cir.2006) (stating a district court's misinterpretation of the Guidelines is harmless if the district court would have imposed the same sentence absent the error); United States v. Jones, 1 F.3d 1167, 1171 (11th Cir.1993) ("A sentencing error is harmless if the record as a whole shows that the error did not affect the district court's selection of the sentence imposed."). Consistent with this standard, a district court's consideration of an impermissible factor at sentencing is harmless if the record as a whole shows the error did not substantially affect the district court's selection of the sentence imposed.

18
If the error is not harmless, we will go no further, and will vacate the sentence and remand for the district court to impose a sentence based on the individualized facts and circumstances of the defendant's case bearing upon the sentencing considerations enumerated in § 3553(a). If, on the other hand, the party defending the sentence is successful in showing the error did not substantially affect the district court's selection of the sentence, we must then resolve whether the sentence is reasonable in light of the § 3553(a) factors and the reasons given by the district court.

19
In sum, to succeed on a claim that an impermissible factor affected the sentence, the party challenging the sentence has the initial burden of establishing that the district court considered an impermissible factor in fashioning the sentence. If we conclude after a de novo review that the district court considered an impermissible factor at sentencing, and if the error was preserved, the burden shifts to the party defending the sentence to show, based on the record as a whole, that the error is harmless, i.e., that the error did not substantially affect the court's choice of sentence. If the error is not harmless, we will vacate the sentence as unreasonable and remand for the district court to resentence the defendant based on the individualized facts and circumstances of the defendant's case bearing upon the sentencing considerations enumerated in § 3553(a). If the error is harmless, we will review the sentence for reasonableness in light of the § 3553(a) factors and the reasons given by the district court.

2. Unreasonable Length of the Sentence

20
Appellant argues that even if the district court considered only permissible sentencing factors, the length of Williams' sentence is unreasonable because the facts and circumstances of Williams' case do not warrant any deviation from the advisory Guidelines range. In essence, Appellant contends that, assuming the sentence is based on only permissible factors, the district court nevertheless made a clear error of judgment in weighing those factors in Williams' case.

21
We review the length of a sentence for reasonableness in light of the facts and circumstances of the defendant's case reflecting the sentencing considerations in § 3553(a). Talley, 431 F.3d at 788. "Review for reasonableness is deferential." Id. And we must bear in mind "that there is a range of reasonable sentences from which the district court may choose." Id. The weight to be accorded any given § 3553(a) factor is a matter committed to the sound discretion of the district court. See United States v. Fernandez, 443 F.3d 19, 32 (2d Cir.2006). We will not substitute our judgment in weighing the relevant factors because "[o]ur review is not de novo." Talley, 431 F.3d at 788.

22
The district court's choice of sentence, however, is not unfettered. When reviewing the length of a sentence for reasonableness, we will remand for resentencing if we are left with the definite and firm conviction that the district court committed a clear error of judgment in weighing the § 3553(a) factors by arriving at a sentence that lies outside the range of reasonable sentences dictated by the facts of the case. See United States v. Martin, 455 F.3d 1227, 1238-42 (11th Cir.2006) (concluding a seven-day sentence for a multi-billion-dollar securities fraud was unreasonable); United States v. Crisp, 454 F.3d 1285, 1291-92 (11th Cir.2006) (vacating a sentence as unreasonable because of the district court's unjustified reliance on a single § 3553(a) factor to the detriment of the others); Moreland, 437 F.3d at 436 (concluding the district court committed "``a clear error of judgment by arriving at a sentence outside the limited range of choice dictated by the facts of the case'") (quoting United States v. Hawk Wing, 433 F.3d 622, 631 (8th Cir.2006)); United States v. Smith, 440 F.3d 704, 708 (5th Cir.2006) (holding a non-Guidelines sentence may be unreasonable where it "represents a clear error of judgment in balancing the sentencing factors"); United States v. Sebastian, 436 F.3d 913, 915 (8th Cir.2006) (holding a sentence may be unreasonable if the district court "commits a clear error in judgment in weighing the sentencing factors").

III. DISCUSSION

A. Guidelines Calculation

23
The first step in our review is to determine whether the district court properly interpreted and applied the Guidelines to Williams' case to arrive at a correct calculation of Williams' advisory Guidelines range. See Williams, 435 F.3d at 1353. There is no dispute about the district court's Guidelines calculation. The district court accurately calculated Williams' advisory Guidelines range using an enhanced offense level of 37 and criminal history category of VI, corresponding to an advisory Guidelines range of 360 months to life imprisonment. Moreover, the district court expressly considered the properly calculated Guidelines range in imposing a non-Guidelines sentence.

B. Reasonableness

1. Reasons for the Sentence

24
Appellant argues the district court committed multiple legal errors in applying the § 3553(a) factors to Williams' case. Specifically, Appellant contends the district court erred in three ways: (1) it rejected Congress's policy of punishing crack cocaine offenders more severely than powder cocaine offenders; (2) it refused to sentence Williams as a career offender, thereby rejecting Congress's policy of punishing recidivist drug offenders more severely; and (3) it erroneously concluded the Government engaged in sentencing manipulation by arranging to purchase crack cocaine instead of powder cocaine and erred in factoring this conclusion into Williams' sentence. According to Appellant, these impermissible considerations affected the district court's choice of sentence, rendering the sentence unreasonable, regardless of length. Because these contentions present questions of law as to the proper considerations at sentencing, we will review de novo whether the district court considered improper factors in fashioning Williams' sentence.

25
a. Impermissible Considerations Affecting the Sentence

26
i. The Cocaine Sentencing Disparity

27
Under 21 U.S.C. § 841 and § 2D1.1 of the Sentencing Guidelines, a defendant convicted of an offense involving "cocaine base" (i.e., crack cocaine)6 faces a longer possible sentence than a defendant convicted of an offense involving the same amount of powder cocaine, a chemically-similar substance. This disparity is commonly referred to as the "100-to-1" ratio, so named because of the relative quantities of each drug required to trigger the mandatory sentencing ranges in § 841(b). For example, § 841(b)(1)(B) provides that offenses involving 5 grams or more of crack cocaine or 500 grams or more of powder cocaine call for sentences in the range of 5 to 40 years' imprisonment. Where the defendant has at least one prior conviction for a felony drug offense, § 841(b)(1)(B) enhances the sentencing range to ten years to life in prison.7 Congress enacted harsher penalties for crack cocaine than for powder cocaine based on its conclusion that crack cocaine poses a greater threat to society. Specifically, Congress found crack cocaine (1) has a more rapid onset of action, (2) is more potent, (3) is more addictive, (4) is less expensive than powder cocaine, (5) has widespread availability, (6) more highly correlates with the incidence of violence and other crimes, (7) is more likely to have physiological effects, and (8) is more likely to attract users who are young or especially vulnerable. See United States v. Byse, 28 F.3d 1165, 1169 (11th Cir.1994) (quoting United States v. Thurmond, 7 F.3d 947, 953 (10th Cir.1993)); U.S. Sentencing Comm'n, Special Report to the Congress: Cocaine and Federal Sentencing Policy 118 (1995). The Sentencing Commission adopted the same 100-to-1 crack-to-powder cocaine ratio in establishing the Drug Quantity Table in U.S.S.G. § 2D1.1(c), which sets the offense levels for drug offenses. For example, § 2D1.1(c)(6) designates offense level 28 for offenses involving 20 to 35 grams of crack cocaine or 2 to 3.5 kilograms of powder cocaine.

28
In 1994, Congress directed the Sentencing Commission to conduct a study of the disparities in penalties for different forms of cocaine and to make recommendations about retaining or modifying the disparities. See Violent Crime Control and Law Enforcement Act of 1994, Pub.L. No. 103-322, § 280006, 108 Stat. 1796, 2097 (1994). Pursuant to this directive, the Commission issued a report to Congress in 1995 in which it agreed with Congress's finding that "crack cocaine poses greater harms to society than does powder cocaine," but concluded "it [could not] recommend a ratio differential as great as the current 100-to-1 quantity ratio." U.S. Sentencing Comm'n, Special Report to the Congress: Cocaine and Federal Sentencing Policy 195-96 (1995).8 One reason why the Commission rejected the 100-to-1 ratio was that Congress adopted it prior to the Guidelines taking effect, and the Commission believed that many, but not all, of the attendant additional harms of crack vis-à-vis powder cocaine (such as the increase in violent and other crimes) are now accounted for in the Guidelines. Id. at 196 ("[I]f Congress believed that certain factors warranted a 100-to-1 quantity ratio and if the subsequently adopted guidelines provided a punishment for some of those factors, then, as a logical matter, the ratio should be lowered by an amount commensurate with the extent to which these factors are addressed by the guidelines.").

29
The Commission subsequently proposed Guidelines amendments that would eliminate entirely the sentencing disparity between crack and powder cocaine. See Notice of Submission to Congress of Amendments to the Sentencing Guidelines, 60 Fed.Reg. 25,074, 25,076 (May 10, 1995). Although Congress stated "the current 100-to-1 quantity ratio may not be the appropriate ratio," it rejected the Commission's proposals because "the evidence clearly indicates that there are significant distinctions between crack and powder cocaine that warrant maintaining longer sentences for crack-related offenses" and "gross sentencing disparities" would result if the proposals were to take effect without Congress lowering the statutory mandatory minimum penalties. H.R.Rep. No. 104-272, at 4 (1995), reprinted in 1995 U.S.C.C.A.N. 335, 337; see also Federal Sentencing Guidelines, Amendment, Disapproval, Pub.L. No. 104-38, § 1, 109 Stat. 334, 334 (1995).

30
The Commission issued a second report in 1997, again at the direction of Congress. U.S. Sentencing Comm'n, Special Report to the Congress: Cocaine and Federal Sentencing Policy (1997). In this report, the Commission reiterated its earlier finding that "although research and public policy may support somewhat higher penalties for crack than for powder cocaine, a 100-to-1 quantity ratio cannot be justified." Id. at 2. The Commission recommended that Congress adjust the mandatory sentencing ranges to reflect a 5-to-1 ratio. Id. Congress, however, took no action.

31
In 2002, at the request of the Senate Judiciary Committee, the Commission issued a third report on the crack-to-powder disparity. U.S. Sentencing Comm'n, Report to the Congress: Cocaine and Federal Sentencing Policy (2002). In the report, the Commission "firmly and unanimously" declared the 100-to-1 drug quantity ratio "is unjustified and fails to meet the sentencing objectives set forth by Congress." Id. at 91. The Commission made four principal findings. First, it found that current penalties exaggerate the relative harmfulness of crack cocaine.9 Id. at 93-97. Although the Commission found that crack was the most addictive form of cocaine because of the method of ingestion, it concluded this difference alone did not warrant the 100-to-1 ratio. Id. at 94. Second, the Commission concluded the current penalties sweep too broadly and apply most often to lower-level offenders, creating disparate penalties in comparison to similar powder cocaine offenders and overstating the culpability of most crack cocaine offenders. Id. at 97-100. Third, the Commission found the penalties overstate the seriousness of most crack cocaine offenses and fail to provide adequate proportionality.10 Id. at 100-02. Fourth, the Commission found that the current penalties impact minorities most severely, fostering disrespect for the criminal justice system. Id. at 102-03. Based on these findings, the Commission recommended that Congress revise the mandatory sentencing range for crack and powder cocaine to implement a 20-to-1 drug quantity ratio. Id. at 107. The Commission also asked Congress for guidance in changing the Guidelines structure to better target the most serious drug offenders. Id. at 108. Again, Congress did not act on the Commission's recommendations.

32
With this background in mind, we turn to the district court's treatment of the cocaine sentencing disparity. The district court disapproved of the severity of Congress's disparate treatment of crack cocaine offenders relative to powder cocaine offenders. At sentencing, the district court expressed its belief that the disparity "smacks of discrimination" and that the difference between crack cocaine and powder cocaine "has never justified the substantial disparity in sentences." Even though the district court did not completely reject Congress's policy of imposing harsher penalties on crack offenders, it took into account its personal disagreement with Congress's judgment as to how much harsher the penalties for crack offenders should be. To the extent the district court did so, it considered an impermissible factor in fashioning Williams' sentence.

33
The First Circuit's decision in United States v. Pho, 433 F.3d 53 (1st Cir.2006), and the Fourth Circuit's decision in United States v. Eura, 440 F.3d 625 (4th Cir.2006), are particularly instructive.11 The district courts in both cases categorically rejected the 100-to-1 drug quantity ratio because they believed it overstated what the penalties ought to be for crack cocaine offenders relative to powder cocaine offenders. Pho, 433 F.3d at 58-59; Eura, 440 F.3d at 631-32. In sentencing the defendants below the advisory Guidelines range, the district courts failed to mention any facts concerning the defendants as individuals that would have warranted non-Guidelines sentences, but instead relied on the general inequities they perceived existed in the 100-to-1 ratio. Pho, 433 F.3d at 64; Eura, 440 F.3d at 634. The First and Fourth Circuits vacated the sentences, concluding that district courts are bound by Congress's policy judgments concerning the appropriate penalties for federal offenses. Pho, 433 F.3d at 62-63; Eura, 440 F.3d at 633-34. Both courts held that sentences must be based on individualized aspects of the defendant's case that fit within the § 3553(a) factors, and not on generalized disagreement with congressional sentencing policy. Pho, 433 F.3d at 64-65; Eura, 440 F.3d at 634.

34
We agree with the First and Fourth Circuit's conclusions. Congress's decision to punish crack cocaine offenders more severely than powder cocaine offenders is plainly a policy decision. It reflects Congress's judgment that crack cocaine poses a greater harm to society than powder cocaine. We have repeatedly held Congress's disparate treatment of crack cocaine offenders is supported by a rational basis. See, e.g., Byse, 28 F.3d at 1168-71 (rejecting equal protection challenge that the crack-to-powder cocaine disparity constitutes intentional race discrimination); United States v. Sloan, 97 F.3d 1378, 1383-84 (11th Cir.1996) (holding the sentencing disparity is supported by a rational basis).12 The 100-to-1 drug quantity ratio not only reflects Congress's policy decision that crack offenders should be punished more severely, but also reflects its choice as to how much more severe the punishment should be. Federal courts are not at liberty to supplant this policy decision. See Pho, 433 F.3d at 62-63; Eura, 440 F.3d at 633; see also Mistretta v. United States, 488 U.S. 361, 364, 109 S.Ct. 647, 650-51, 102 L.Ed.2d 714 (1989) ("Congress, of course, has the power to fix the sentence for a federal crime, and the scope of judicial discretion with respect to a sentence is subject to congressional control." (internal citation omitted)). Although the two drugs may be chemically similar, their effect on society is not the same, and it is not for the courts to say just how much worse crack cocaine is than powder cocaine. This is simply an impermissible sentencing consideration. As the Seventh Circuit aptly put it: "§ 3553(a) . . . does not include a factor such as ``the judge thinks the law misguided.'" Miller, 450 F.3d at 275.

35
Williams, however, asserts that the crack versus powder cocaine sentencing disparity is a valid consideration under § 3553(a)(6), which requires courts to consider "the need to avoid unwarranted sentence disparities among defendants with similar records who have been found guilty of similar conduct." Several district courts that have addressed the issue agree with Williams' position. See, e.g., United States v. Fisher, ___ F.Supp.2d ___, ___, 2006 WL 2542916 (S.D.N.Y. 2005); United States v. Smith, 359 F.Supp.2d 771, 781 (E.D.Wis.2005). Powder cocaine offenders, however, have not been found guilty of similar conduct in any relevant sense. Congress has determined that crack offenders and powder offenders are not similarly situated, and that the disparities caused by its choice of the 100-to-1 drug quantity ratio are warranted. See Pho, 433 F.3d at 64 ("Congress plainly believed that not all cocaine offenses are equal and that trafficking in crack involves different real conduct than trafficking in powder . . . . Clearly, then, Congress intended that particular disparity to exist, and federal courts are not free to second-guess that type of decision."). The district court's rejection of the 100-to-1 drug quantity ratio, therefore, cannot be justified under § 3553(a)(6).

36
Williams also contends the district court did not impermissibly usurp Congress's policy judgment because Williams was sentenced within the statutory range. He argues the 100-to-1 ratio embedded in the Guidelines is not Congress's policy, but the Sentencing Commission's policy, one the Commission has unanimously rejected. He suggests that the district courts can exercise their sentencing discretion to reject the advisory crack cocaine Guidelines without running afoul of Congress's policy judgment.

37
Williams is incorrect in suggesting the 100-to-1 ratio embedded in the Guidelines is merely the Sentencing Commission's policy and not Congress's policy. In determining the threshold quantities for triggering the statutory sentencing ranges in § 841(b), Congress decided on a 100-to-1 differential, and the Sentencing Commission was left no choice but to employ the same ratio in crafting the various Guidelines ranges within those statutory ranges. See id. at 63 ("As the Sentencing Commission recognized when it superimposed the guidelines on the statutory framework, it would be illogical to set the maximum and minimum sentences on one construct and then to use some other, essentially antithetic construct as the basis for fashioning sentences within the range."). Indeed, Congress rejected the Commission's proposal that would have equated the drugs for Guidelines purposes because of the gross sentencing disparities that would result if the Guidelines did not employ the same drug quantity ratio as the statutory scheme. See H.R.Rep. No. 104-272, at 4 (1995), reprinted in 1995 U.S.C.C.A.N. 335, 337. The same is true if instead of equating the two drugs, the Guidelines used a different drug quantity ratio, say a 20-to-1 ratio. See Pho, 433 F.3d at 63-64. If the Guidelines used a 20-to-1 ratio, a first time offender convicted of selling 50 grams of crack cocaine (the equivalent of one kilogram of powder cocaine in our hypothetical) would have a Guidelines range of 63 to 78 months' imprisonment, see U.S.S.G. § 2D1.1(c)(7), but would have a mandatory minimum sentence of 120 months' incarceration under § 841(b)(1)(A). In contrast, a first time offender convicted of selling 49 grams of crack cocaine (the equivalent of 980 grams of powder cocaine in our hypothetical) would also have a Guidelines range of 63 to 78 months' imprisonment, but would not be subject to the 120-month mandatory minimum. See id. § 841(b)(1)(B). In this scenario, the difference of one gram of crack cocaine would result in a sentencing disparity of at least 42 months. Thus, the statutory minimums and maximums and the Guidelines reflect Congress's policy decision to punish crack offenses more severely than powder cocaine offenses by equating one gram of crack to 100 grams of cocaine.

38
The same unwarranted disparities between similarly situated defendants would result if a district court were permitted to use its discretion to disregard the 100-to-1 ratio. Thus, a district court's rejection of the 100-to-1 ratio embedded in the Guidelines not only countermands Congress's policy choice, but also undermines sentencing uniformity in direct contravention of § 3553(a)(6)'s command that district courts seek to avoid unwarranted sentencing disparities between similarly situated defendants. See Pho, 433 F.3d at 63-64; Eura, 440 F.3d at 633 ("[G]iving a sentencing court the authority to sentence a defendant based on its view of an appropriate ratio between crack cocaine and powder cocaine would inevitably result in an unwarranted disparity between similarly situated defendants in direct contradiction to the specific mandate of 18 U.S.C. § 3553(a)(6)."). Of course, some disparity between similarly situated defendants is an inevitable result of Booker. See Booker, 543 U.S. at 263, 125 S.Ct. at 766-67 ("We cannot and do not claim that use of a ``reasonableness' standard will provide the uniformity that Congress originally sought to secure."). This inevitable disparity, however, should only be the product of the district court's discretion in weighing individualized § 3553(a) factors in a given case, not the consequence of the district court's "general, across-the-board policy considerations." Pho, 433 F.3d at 62.

39
Williams is correct that a sentence below the Guidelines range in a crack cocaine case may be reasonable, so long as it reflects the individualized, case-specific factors in § 3553(a). It may be that for some of the reasons stated in the Sentencing Commission's reports, the Guidelines range in a given crack case overstates the seriousness of the particular defendant's offense or that individualized mitigating factors counsel against a Guidelines sentence. See Eura, 440 F.3d at 637 (Michael, J., concurring) ("While the Commission's findings alone cannot justify a below-guidelines sentence, in certain cases they can help sentencing courts analyze the § 3553(a) factors and select a sentence that is ``sufficient, but not greater than necessary' to punish, deter, and rehabilitate the defendant."). But to say Congress's choice of a 100-to-1 drug quantity ratio is never justified is a categorical rejection of congressional policy, not an individualized, case-specific consideration. Congress concluded the 100-to-1 ratio is justified, and the courts have no authority to change that.

40
In short, the district court erred in mitigating Williams' sentence based on its personal disagreement with Congress's policy decision to employ a 100-to-1, crack-to-powder drug quantity ratio in punishing crack cocaine offenders more severely than powder cocaine offenders. In so doing, the district court impermissibly usurped Congress's authority to set sentencing policy and failed to properly consider § 3553(a)(6)'s directive to avoid unwarranted sentence disparities between similarly situated defendants. We agree, moreover, with the Fourth Circuit that "allowing sentencing courts to subvert Congress' clearly expressed will certainly does not promote respect for the law, provide just punishment for the offense of conviction, or result in a sentence reflective of the offense's seriousness as deemed by Congress." Id. at 633 (majority opinion).

41
ii. Career Offender Guideline Provision

42
Appellant next argues the district court erred in refusing to sentence Williams as a career offender. There is no dispute that Williams qualified as a career offender under U.S.S.G. § 4B1.1. At sentencing, however, the district court stated the career offender enhancement "is a totally inappropriate way to consider the individual nature of an offense or a defendant's individual background" and said it was not going to sentence Williams as a career offender. In its sentencing memorandum, the district court again explained what it considered to be the "arbitrary compounding" effect of the career offender enhancement. Williams, 372 F.Supp.2d at 1339. This, too, was error.

43
In creating the Sentencing Commission and charging it with establishing sentencing policies and practices for the federal criminal justice system, Congress directed the Commission to:

44
assure that the guidelines specify a sentence to a term of imprisonment at or near the maximum term authorized for categories of defendants in which the defendant is eighteen years old or older and—

45
(1) has been convicted of a felony that is—

46
(A) a crime of violence; or

47
(B) an offense described in section 401 of the Controlled Substances Act (21 U.S.C. 841) . . .; and

48
(2) has previously been convicted of two or more prior felonies, each of which is—

49
(A) a crime of violence; or

50
(B) an offense described in section 401 of the Controlled Substances Act (21 U.S.C. 841). . . .

51
28 U.S.C. § 994(h). Section 994(h) reflects Congress's policy that repeat drug offenders receive sentences "at or near" the enhanced statutory maximums set out in § 841(b). See United States v. LaBonte, 520 U.S. 751, 762, 117 S.Ct. 1673, 1679, 137 L.Ed.2d 1001 (1997) (holding "the phrase ``at or near the maximum term authorized' . . . requires a court to sentence a career offender ``at or near' the ``maximum' prison term available once all relevant statutory sentencing enhancements are taken into account").

52
Congress's goal was not simply to punish offenders with prior criminal histories more severely than first time offenders; Congress also wanted to target specific recidivism, particularly repeat drug offenders. There is no question Williams is a recidivist drug dealer. To the extent the district court believed Williams' prior criminal history was adequately taken into consideration in his criminal history category of VI, it ignored Congress's policy of targeting recidivist drug offenders for more severe punishment. The district court, therefore, erred in mitigating Williams' sentence based on its disagreement with the career offender Guidelines provision.

53
iii. Sentencing Factor Manipulation

54
Finally, Appellant argues the district court erred in mitigating Williams' sentence based on its belief that the DEA "snared" Williams by arranging to purchase crack cocaine from him, when a powder cocaine purchase would have been consistent with his prior drug sales. The district court's decision to mitigate Williams' sentence because of the DEA's conduct in investigating Williams calls to mind a claim we have referred to as "sentencing factor manipulation." United States v. Sanchez, 138 F.3d 1410, 1414 (11th Cir.1998).

55
A sentencing factor manipulation claim "``requires us to consider whether the manipulation inherent in a sting operation, even if insufficiently oppressive to support an entrapment defense, . . . or due process claim, . . . must sometimes be filtered out of the sentencing calculus.'" Id. (quoting United States v. Connell, 960 F.2d 191, 194 (1st Cir.1992)). The claim focuses on the Government's conduct and "points to ``the opportunities that the sentencing guidelines pose for prosecutors to gerrymander the district court's sentencing options and thus, defendant's sentences.'" Id. (quoting Connell, 960 F.2d at 194).

56
We need not decide whether a finding of sentencing factor manipulation is a valid mitigating consideration under § 3553(a) because, even if it can be, it was not an appropriate consideration here. There is no question Williams' arrest and conviction was the result of a valid sting operation, and the DEA no more "snared" Williams or engaged in sentencing factor manipulation than in any other sting operation. To say a district court may factor into a crack cocaine offender's sentence the bare fact that the Government chose to purchase crack instead of powder cocaine, without more, would undermine Congress's policy of punishing crack cocaine offenders more severely and impermissibly interfere with the executive branch's performance of legitimate law enforcement practices. Contrary to the district court's conclusion, it does not promote respect for the law to imply Government misconduct from the mere fact that the Government chose to purchase crack cocaine from a crack dealer instead of any other controlled substance.13 To the extent the district court considered this fact as a mitigating consideration, it erred as a matter of law. This is not to say that sentencing manipulation may never be a valid consideration in sentencing. In this case, however, it was error to mitigate Williams' sentence based on the fact that the DEA purchased crack cocaine from Williams instead of powder cocaine.

57
b. Harmless Error

58
Because Appellant objected to the district court's consideration of these impermissible factors, it is Williams' burden under our traditional harmless error standard to show, based on the record as a whole, that the errors did not substantially affect the district court's choice of sentence. See Mathenia, 409 F.3d at 1292. Williams has failed to meet his burden.

59
Williams points to nothing in the record showing the errors did not substantially affect the district court's choice of sentence. A review of the sentencing transcript and the district court's sentencing memorandum instead shows the district court devoted the overwhelming majority of its explanation of the sentence to expressing its disagreement with the cocaine sentencing disparity, the career offender provision, and the DEA's decision to purchase crack cocaine from Williams. Although the record reflects that the district court also considered individualized facts and circumstances of Williams' case—such as the relatively small amount of crack cocaine involved in his offenses—the district court's explanation of Williams' sentence was so permeated by its consideration of impermissible factors that we are unable to conclude the errors did not have a substantial effect on its choice of sentence. The errors are, therefore, not harmless, and we must vacate the sentence and remand for the district court to resentence Williams solely on the basis of the individualized facts and circumstances of Williams' case bearing on the § 3553(a) factors.

2. Unreasonable Length of the Sentence

60
Having concluded Williams' sentence is unreasonable because it is based on impermissible factors, we do not reach Appellant's alternative argument that, assuming the district court considered only proper factors in crafting Williams' sentence, the length of the sentence is nevertheless unreasonable because the record does not justify any deviation from the advisory Guidelines range. We express no opinion as to whether there are individual facts and circumstances in Williams' case that would make a 204-month sentence reasonable.

IV. CONCLUSION

61
We conclude, after a de novo review, that the district court considered impermissible factors in crafting Williams' sentence. Because Williams failed to show, based on the record as a whole, that the errors did not substantially affect the district court's choice of sentence, we conclude the errors are not harmless. We, therefore, vacate Williams' sentence as unreasonable without reaching Appellant's alternative argument that the record does not justify the length of the sentence imposed. On remand, the district court must resentence Williams based on the individual facts and circumstances of Williams' case bearing on the § 3553(a) factors. We express no opinion as to what sentence the district court should impose after properly applying the § 3553(a) factors.

62
VACATED AND REMANDED.

Notes:

1
Williams sold the undercover agent 4.6 grams of crack cocaine in April, 10.1 grams of crack cocaine in May, and 20.1 grams of crack cocaine in July

2
All references to the Sentencing Guidelines refer to the Guidelines effective November 1, 2004

3
The advisory Guidelines range for a defendant convicted of selling 34.8 grams of powder cocaine with a criminal history category of VI is 37 to 46 months' imprisonmentSee U.S.S.G. § 2D1.1(c)(13); ch. 5, pt. A. Notably, the district court did not compare Williams' sentence to the range applicable to a career offender convicted of selling 34.8 grams of powder cocaine: 262 to 327 months' imprisonment. See 21 U.S.C. § 841(b)(1)(C); U.S.S.G. § 4B1.1(b); ch. 5, pt. A.

4
The § 3553(a) sentencing factors include the nature and circumstances of the offense and the history and characteristics of the defendant; the need for the sentence to reflect the seriousness of the offense, promote respect for the law, and provide just punishment for the offense; the need to deter crime, protect the public, and provide the defendant with educational or vocational training, or medical care; the kinds of sentences available; the Sentencing Guidelines range; pertinent policy statements of the Sentencing Commission; the need to avoid unwarranted sentencing disparities; and the need to provide restitution to victims. 18 U.S.C. § 3553(a);United States v. Winingear, 422 F.3d 1241, 1246 (11th Cir.2005).

5
We realize a sentence could be alleged to be unreasonable on grounds other than length. Here, the Government contests only the length of Williams' sentence, so we confine our discussion to length. Our opinion should not be read to foreclose other possible challenges to the reasonableness of the terms of a sentence

6
Section 841 refers to "cocaine base," and the Sentencing Guidelines define cocaine base to mean crack cocaineSee U.S.S.G § 2D1.1(c), n.D.

7
Similarly, § 841(b)(1)(A) sets the sentencing range for offenses involving 50 grams or more of crack cocaine or 5 kilograms or more of powder cocaine at ten years to life in prison. If the defendant has at least one prior conviction for a felony drug offense, § 841(b)(1)(A) requires a sentence between 20 years and life in prison

8
The Sentencing Commission's reports are available at http://www.ussc.gov/reports.htm

9
For example, the Commission pointed to studies showing the effects of prenatal crack cocaine exposure were the same for powder cocaine and that the epidemic of young users and distributors "never materialized to the extent feared." U.S. Sentencing Comm'n, Report to the Congress: Cocaine and Federal Sentencing Policy 93-97 (2002)

10
For example, the Commission stated that although studies showed harmful conduct (such as violence) occurs more often in crack cocaine offenses than in powder cocaine offenses, "it occurs in only a relatively small minority of crack cocaine offenses." U.S. Sentencing Comm'n, Report to the Congress: Cocaine and Federal Sentencing Policy 100 (2002). Thus, "to the extent that the 100-to-1 drug ratio was designed to account for the harmful conduct . . ., it sweeps too broadly by treating all crack cocaine offenders as if they committed these various harmful acts, even though most crack cocaine offenders in fact had not."Id. Although the Commission recognized that "some differential in the quantity-based penalties for crack cocaine and powder cocaine is warranted" because sentencing enhancements did not account for the fact that "trafficking in crack cocaine is associated with somewhat greater levels of systemic crime," it opined that this consideration did not justify the 100-to-1 ratio. Id. at 101-02.

11
See also United States v. Miller, 450 F.3d 270, 275 (7th Cir.2006) (agreeing with Pho and Eura).

12
Post-Booker, we have rejected a defendant's challenge to the reasonableness of his sentence predicated on the disparity between crack and powder cocaine. See United States v. Marlin, 147 Fed.Appx. 122, 124 (11th Cir. 2005) (unpublished opinion).

13
We note that all five of Williams' prior felony cocaine convictions listed in the PSI involved crack cocaine; two of those were convictions for possessing crack cocaine with intent to sell