Summary
finding that, though there is no "mandatory script for sentencing," the District Court must demonstrate that it has carefully considered and analyzed the sentencing factors
Summary of this case from U.S. v. HasanOpinion
No. 09-1844.
Argued March 8, 2010.
Filed April 20, 2010.
Appeal from the United States District Court for the District of New Jersey, William J. Martini, J.
John F. Romano, (argued), George S. Leone, Office of United States Attorney, Newark, NJ, for Appellant.
Louise Arkel, (argued), David A. Holman, Office of Federal Public Defender, Newark, NJ, for Appellee.
Hector Merced pleaded guilty to a drug possession charge and was sentenced to five years of imprisonment. That sentence was well below the prison term recommended by the Sentencing Guidelines. The United States appeals, claiming that the District Court committed procedural errors in determining Merced's sentence. We agree with the government. Accordingly, we will vacate the judgment of sentence and remand for resentencing.
I.
On June 27, 2007, Merced sold 49.1 grams of crack cocaine to an undercover police officer. Merced received $1,500 from the sale, and expressed willingness to do more deals in the future. He was arrested on January 14, 2008. He pleaded guilty to one count of distributing and possessing with intent to distribute five grams or more of crack cocaine, in violation of 21 U.S.C. § 841(a) and (b)(1)(B). That crime carries a five-year mandatory minimum sentence. 21 U.S.C. § 841(b)(1)(B).
Thus, the District Court had jurisdiction under 18 U.S.C. § 3231.
A presentence report (PSR) was prepared. The PSR recounted Merced's extensive criminal history, which included:
• a 1997 conviction for possession of controlled dangerous substances ("CDS") — specifically, 41 bags of crack cocaine and a bag of marijuana — with intent to distribute within 1000 feet of school property;
• two 1998 convictions for prowling public places;
Merced possessed marijuana on both occasions.
• a 1999 conviction for possession of CDS (15 bags of cocaine) with intent to distribute within 1000 feet of school property;
• a 2001 conviction for receiving a stolen vehicle;
• a 2001 conviction for possession of CDS (52 bags of heroin and 8 bags of crack cocaine) with intent to distribute within 1000 feet of school property; and
— a 2006 conviction for conspiracy to distribute CDS (23 bags of marijuana).
Due to his criminal history, Merced qualified as a career offender under U.S.S.G. § 4B1.1. That provision applied because (1) Merced was more than 18 years old when he participated in the June 27, 2007 crack cocaine deal, (2) his crime was a controlled substance offense, and (3) he had two or more prior felony convictions that were controlled substance offenses. See U.S.S.G. § 4B1.1(a). Merced's career offender status increased his offense level to 34, id. § 4B1.1(b); a three-level reduction for acceptance of responsibility lowered that level to 31. Id. § 3E1.1. The PSR placed him in Criminal History Category VI. Id. § 4B1.1(b). His combined offense level and Criminal History Category yielded an advisory Guidelines range of 188 to 235 months. Absent the career offender provision, his base offense level would have been 28, his adjusted offense level 25, and his Guidelines range 110-137 months.
Both parties filed memoranda before the sentencing hearing. Merced argued for a below-Guidelines sentence. He acknowledged that the Guidelines range of 188-235 months had been correctly calculated, but argued that the resulting Guidelines sentence was "exceedingly harsh" and not justified under the sentencing factors of 18 U.S.C. § 3553(a). His argument was two-fold. First, based on Kimbrough v. United States, 552 U.S. 85, 128 S.Ct. 558, 169 L.Ed.2d 481 (2007), he attacked the crack cocaine Guidelines and the career offender provision of § 4B1.1 on policy grounds. He claimed that because neither reflected the Sentencing Commission's "exercise of its characteristic institutional role," his Guidelines range, which was a product of those two provisions, was a "poor touch-stone[] in any § 3553(a) inquiry." Second, he argued that consideration of the sentencing factors in his specific case counseled in favor of leniency. Merced claimed that he was a casualty of a "`perfect storm' of discouraging forces: a splintered family, economic struggle, and an increasingly punitive criminal justice system." Merced explained that his life "imploded" at the age of 12, when two ex-convict uncles moved into his home and got his mother hooked on crack cocaine. According to Merced, the mother's addiction led to domestic strife, and eventually caused Merced's father to abandon the family and move to Puerto Rico. Merced claimed that this troubled childhood pushed him down the wrong path and led him to a life of criminal activity for which he was not wholly responsible. Additionally, he asked the court to consider his strong relationship with his longtime girlfriend and their 10-year-old son, as well as letters from various family members attesting to his positive attributes. Merced did not request a sentence of any particular length, but he argued that even a 10-year sentence was "simply not necessary" to satisfy the sentencing goals of § 3553(a).
Those factors are:
(1) the nature and circumstances of the offense and the history and characteristics of the defendant;
(2) the need for the sentence imposed —
(A) to reflect the seriousness of the offense, to promote respect for the law, and to provide just punishment for the offense;
(B) to afford adequate deterrence to criminal conduct;
(C) to protect the public from further crimes of the defendant; and
(D) to provide the defendant with needed educational or vocational training, medical care, or other correctional treatment in the most effective manner;
(3) the kinds of sentences available;
(4) the kinds of sentence and the sentencing range established for —
(A) the applicable category of offense committed by the applicable category of defendant as set forth in the guidelines . . . .
(5) any pertinent policy statement —
(A) issued by the Sentencing Commission . . . . subject to any amendments made to such policy statement by act of Congress . . . .
(6) the need to avoid unwarranted sentence disparities among defendants with similar records who have been found guilty of similar conduct; and
(7) the need to provide restitution to any victims of the offense.
18 U.S.C. § 3553(a).
The government advocated a sentence within the Guidelines range and urged the court to reject Merced's Kimbrough attack on § 4B1.1. It argued that Kimbrough was inapposite because, unlike the crack cocaine Guidelines at issue in that case, the career offender provision that dictated Merced's recommended sentence reflected not only the Sentencing Commission's exercise of its unique institutional role, but also direct "congressional involvement in the setting of punishment for certain recidivists." (A. 53-55, citing 28 U.S.C. § 994(h)). It further argued that even if the Court accepted Merced's argument based on Kimbrough, it should not apply a replacement ratio less than the 20:1 ratio accepted by the Supreme Court in Spears v. United States, ___ U.S. ___, 129 S.Ct. 840, 842, 172 L.Ed.2d 596 (2009). According to the government, applying that ratio would generate a Guidelines range of 92 to 115 months, and "any ratio less than this would result in significant sentencing disparities, as a defendant who possessed merely an additional .9 grams of crack [i.e., 50 grams] . . . and had no criminal history whatsoever, would be subject to a mandatory minimum term of ten years' imprisonment, [under] 21 U.S.C. § 841(b)(1)(A)."
Finally, the government turned to the sentencing factors and argued that each relevant factor weighed in favor of a Guidelines sentence. It emphasized Merced's criminal history, which included at least five drug-related convictions, and the seriousness of his offense. See 18 U.S.C. § 3553(a)(1)-(2). It argued that a Guidelines-length sentence was necessary to protect the public and to deter both Merced and other drug dealers. See id. § 3553(a)(2)(B)-(C). The government also contended that a within-Guidelines sentence was necessary to avoid unwarranted sentencing disparities. see id. § 3553(a)(6). It argued that Merced's sentence "should be commensurate to other career offenders convicted of similar offenses in New Jersey and across the country."
The District Court held a sentencing hearing on February 24, 2009. At that hearing, the Court took particular interest in the application of the career offender provision. All parties agreed that Merced was eligible for career offender status under a mechanical application of § 4B1.1. The point of debate was not Merced's technical eligibility, but whether the Court should sentence him within the range generated by applying that provision. Throughout the hearing, the District Court was obviously wrestling with two competing realities. The first was that Merced was a 31-year-old career criminal who had not been deterred by the punishment he received for his prior convictions. The second was that many of the crimes that brought Merced within § 4B1.1 were "street level" offenses that the Court did not consider to be terribly serious. This tension between the numerosity and the severity of Merced's crimes led the District Court to engage in an on-one-hand-on-the-other-hand dialectic several times during the hearing. For example:
[Merced has] been a repetitive drug — street drug dealer, okay? He just doesn't get it. I mean, he's been on the street dealing drugs, relatively small amounts compared to what I often see, but he's been doing it repetitively. . . . And it begins at a very young age, it begins at 19. He's 31 years of age now. It begins at 19. A lot of what's in his record with the exception of four prior drug arrests, a lot of it is the type of stuff that you see with people who are dealing with drugs on the street level. I'm not minimizing it, but, you know, public — prowling, public places, a couple of municipal court violations.
After walking through the specifics of Merced's criminal history, the Court returned to this issue. He acknowledged that, "there are career offenders and there are career offenders. . . . But he hasn't learned." In the same vein, the Court later noted that "we're talking about a relatively small sale of crack cocaine at the street level. But it's his fifth time he's done this now. . . . I don't know when he's going to wake up."
In the midst of his oral ruminations, however, the District Court revealed another possible reason for its reluctance to sentence Merced as a career criminal. The Court told Merced that he needed to realize that "he can't make a living like this" because "another judge might . . . [apply] the career offender status recommendation. . . . I have a problem with that. I mean, I kind of reserve career offender status for violent, significant drug deals, that type of thing, even though the guidelines may advise that it's appropriate." (A. 76, emphasis added.) The Court did not explain either how it arrived at this personal sentencing policy, or why it believed that the contrary policies reflected in the Guidelines were out of line.
During the hearing, the District Court also took note of several § 3553(a)(1) factors that in its view favored a downward variance. He observed the relatively small quantity of crack involved in Merced's latest offense, the "street level" nature of Merced's previous crimes, the fact that Merced "probably [has] a drug problem himself," Merced's troubled childhood, and his strong relationships with his girl-friend and son. The District Judge noted that the longest prison term Merced had received for any of his previous crimes was, at most, four years, and reasoned that no matter what, he was facing "more [jail time] than he's done before." The government conceded that Merced was a relatively low-level dealer, and not "the main connect around here who is bringing in kilos of heroin and coke." It emphasized, however, that this was Merced's sixth drug crime, and remained adamant that Merced deserved a Guidelines sentence because he had not "learned his lesson" from previous arrests. The Court was not persuaded: "Looking behind and looking at his criminal record, which gets into the individual circumstances of his conduct and his history, it just is an excessive guideline recommendation, in this Court's opinion."
Merced used ccstacy and marijuana daily, and claimed that he was addicted to the latter.
After Merced read a letter to the Court accepting responsibility for his crime, the District Judge undertook a formal analysis of the § 3553(a) factors. We reprint this analysis in full, because just as important as what the Court said is what it did not say. Specifically, it made no mention of § 3553(a)(6), "the need to avoid unwarranted sentence disparities among defendants with similar records who have been found guilty of similar conduct":
That's my responsibilit[y] under 3553, to impose a sentence that's sufficient but not greater than necessary. I have to look at the nature of the circumstances of the offense, and I certainly have here. The offense before me itself is in all respects a relatively small offense compared to what I deal with on drug cases frequently. I'm not minimizing it, but I have to put it in some context.
But the history and characteristics of the defendant. It's clear, and I've said it already, he hasn't learned his lesson since he was 19 years old. He's had a number of different involvements and convictions for possession with intent to distribute, mostly in state court, mostly small amounts, glassine bags, marijuana, crack cocaine. Clearly, it shows that he's resorted to being a street dealer. Maybe not all the time, but when he's in need of money. And his biggest sentence up to now was 30 months to 4 years.
I don't know if he served — he had 4 years with a 30-month mandatory minimum. I don't know if he served more than 30 months or not. But he still didn't get it is my problem. And as I've said already, statutorily he's faced with at least 60 months [here]. I'm not sure that's going to be enough time for him to wake up and get it. Okay?
There are some — you know, look, he's unfortunately somewhat a product of his upbringing. You know, it appears that his upbringing — his parents were drug users and perhaps even violent; yet, there are some good things, at least hopeful things about him. He appears to have — considering his conduct in terms of being in the streets with drugs,
he's maintained a consensual union with his girlfriend for over 12 years. That's probably the strongest thing going in his life. He has one child from this relationship and one child from a previous one. He says he has a harmonious relationship with this woman, and he appears to at least attempt to want to take care of his kids. I don't see that all the time in these cases.
He says he's tried to gain employment but he's faced obstacles. And, of course, he's going to face more obstacles with the stronger criminal record that he has. I think a substantial variance is warranted. It's a serious crime, but I put it in context looking at his record. There's a need to deter others from this type of street conduct but, you know what? There's a need to deter him. And he hasn't gotten it yet. And there is some need to protect the public from people who feel that they can go out when they are struggling and sell drugs on the street. I also have a duty not to impose an unnecessarily harsh and punitive sentence for the conduct that's before me, even with the criminal record I have.
Based on this analysis, the Court sentenced Merced to 60 months in prison. This was the shortest prison term allowable by statute. It was also 128 months below the bottom of Merced's advisory Guidelines range, and 50 months below the bottom of what that range would have been even in the absence of § 4B1.1. Later, the Court issued a written Statement of Reasons ("SOR") explaining that this variance was based on
the nature and circumstances of the offense and the history and characteristics of the defendant pursuant to 18 U.S.C. § 3553(a)(1). The court finds that a sentence within the career offender guideline range[] is far greater than necessary to achieve just punishment and that a variance is appropriate because of the relatively small sale of crack cocaine[] in the instant offense and the relatively minor nature of the transactions in prior qualifying convictions.
The SOR did not mention the need to avoid unwarranted sentencing disparities, or the Court's personal policy of applying § 4B1.1 only in cases involving "violent, significant drug deals." The government appealed the sentence pursuant to 18 U.S.C. § 3742(b). We have jurisdiction under 28 U.S.C. § 1291.
II.
Before the Supreme Court's decision in Booker, the federal Sentencing Guidelines were binding on district courts; they had "the force and effect of laws." United States v. Booker, 543 U.S. 220, 234, 125 S.Ct. 738, 160 L.Ed.2d 621 (2005). Booker "held unconstitutional that portion of the Guidelines that made them mandatory," Rita v. United States, 551 U.S. 338, 354, 127 S.Ct. 2456, 168 L.Ed.2d 203 (2007), and replaced the mandatory regime with one in which the Guidelines are "effectively advisory." Booker, 543 U.S. at 245, 125 S.Ct. 738. After Booker, sentencing courts must "consider the Guidelines range" pursuant to § 3553(a)(4), but also "tailor the sentence in light of other statutory concerns" reflected in the sentencing factors of § 3553(a). Id. "[Appellate review of sentencing decisions is limited to determining whether they are reasonable." Gall v. United States, 552 U.S. 38, 46, 128 S.Ct. 586, 169 L.Ed.2d 445 (2007) (internal quotations omitted).
In Rita, the Supreme Court held that courts of appeals may (but need not) presume that a within-Guidelines sentence is reasonable. 551 U.S. at 347, 127 S.Ct. 2456. Our Court has declined this invitation; others have accepted it. Compare United States v. Cooper, 437 F.3d 324, 331 (3d Cir. 2006), with United States v. Dorcely, 454 F.3d 366, 376 (D.C. Cir. 2006). No court of appeals, however, may presume that an outside-the-Guidelines sentence is unreasonable. Rita, 551 U.S. at 354-55, 127 S.Ct. 2456.
"[B]oth the district court's crafting of an appropriate sentence and the appellate court's review of that sentence for reasonableness must be `guided by the factors set forth in 18 U.S.C. § 3553(a).'" United States v. Ausburn, 502 F.3d 313, 327-28 (3d Cir. 2007) (quoting United States v. Cooper, 437 F.3d 324, 327 (3d Cir. 2006)). Reasonableness review proceeds in two stages, and employs the "familiar abuse of discretion standard" at each stage. Gall, 552 U.S. at 46, 128 S.Ct. 586. First, we ensure that the district court committed no "significant procedural error," United States v. Tomko, 562 F.3d 558, 568 (3d Cir. 2009) (en banc), "such as failing to calculate (or improperly calculating) the Guidelines range, treating the Guidelines as mandatory, failing to consider the § 3553(a) factors, selecting a sentence based on clearly erroneous facts, or failing to adequately explain the chosen sentence[.]" Gall, 552 U.S. at 51, 128 S.Ct. 586. If the district court commits procedural error, our preferred course is to remand the case for re-sentencing, without going any further. Ausburn, 502 F.3d at 328. But if the district court's procedures are sound, we proceed to examine the substantive reasonableness of the sentence. Id. "The touchstone of reasonableness is whether the record as a whole reflects rational and meaningful consideration of the factors enumerated in § 3553(a)." United States v. Grier, 475 F.3d 556, 571 (3d Cir. 2007) (en banc) (internal quotations omitted).
But see United States v. Lychock, 578 F.3d 214, 219-20 (3d Cir. 2009) (finding procedural error yet proceeding to analyze substantive reasonableness).
In Tomko, we held that our substantive reasonableness inquiry must be highly deferential. 562 F.3d at 568. We recognized that the sentencing judge, not the court of appeals, "`is in a superior position to find facts and judge their import under § 3553(a) in the individual case. The sentencing judge sees and hears the evidence, makes credibility determinations, has full knowledge of the facts and gains insights not conveyed by the record.' This means that `the sentencing judge has access to, and greater familiarity with, the individual case and the individual defendant before him than . . . the appeals court.'" Id. at 566 (quoting Gall, 552 U.S. at 51-52, 128 S.Ct. 586) (internal citations omitted). We further recognized that the district court's superior vantage point compels us to "`give due deference to [its] determination that the § 3553(a) factors, on a whole,' justify the sentence." Id. at 568 (quoting Gall, 552 U.S. at 51, 128 S.Ct. 586). Accordingly, we held that "if the district court's sentence is procedurally sound, we will affirm it unless no reasonable sentencing court would have imposed the same sentence on that particular defendant for the reasons the district court provided." Tomko, 562 F.3d at 568.
We were careful to note, however, that our adoption of a deferential standard of review for substantive reasonableness was "not an exercise in self-abnegation." Id. at 575. We retain "an important role . . . in reviewing district courts' sentencing decisions." Id. Chief among our duties in fulfilling this "important role" is ensuring that district courts follow proper sentencing procedures. Indeed, the broad substantive discretion afforded district courts under Tomko makes adherence to procedural sentencing requirements all the more important. These procedural requirements exist to "guide the [district court's] exercise of discretion," and failure to observe them may lead a court to impose a substantively unreasonable sentence. United States v. Goff, 501 F.3d 250, 256 (3d Cir. 2007) (describing substantively unreasonable sentence as "the product of the District Court's procedurally flawed approach"); United States v. Lychock, 578 F.3d 214, 220 (3d Cir. 2009) (concluding that "by ignoring relevant factors and failing to offer a reasoned explanation for its departure from the Guidelines, the District Court . . . put at risk the substantive reasonableness of any decision it reached" (internal quotations omitted)).
"[O]ur post- Booker precedent instructs district courts to follow a three-step sentencing process." United States v. Gunter, 462 F.3d 237, 247 (3d Cir. 2006). First, the court must correctly calculate the defendant's Guidelines range. Id. Second, it must rule on any motions for departures. Id. Finally, "after giving both parties an opportunity to argue for whatever sentence they deem appropriate," Gall, 552 U.S. at 49, 128 S.Ct. 586, the court must "exercise[] its discretion" through "meaningful consideration to the § 3553(a) factors" before deciding on a sentence. Cooper, 437 F.3d at 329. While there is "no mandatory script for sentencing," Goff, 501 F.3d at 256, the court's analysis of the relevant factors must adequately demonstrate its exercise of "independent judgment" and meaningful consideration of the relevant sentencing factors. United States v. Sevilla, 541 F.3d 226, 232 (3d Cir. 2008). The district court need not make explicit "findings as to each of the § 3553(a) factors if the record makes clear that the court took the factors into account in sentencing." Cooper, 437 F.3d at 329. It is "not incumbent on the District [Court] to raise every conceivable relevant issue on [its] own initiative" during sentencing, Gall, 552 U.S. at 54, 128 S.Ct. 586, nor must the court "discuss every argument made by a litigant if an argument is clearly without merit." Cooper, 437 F.3d at 329. However, if a party raises a colorable argument about the applicability of one of the § 3553(a) factors, the district court may not ignore it. The court should address that argument as part of its "meaningful consideration" of the sentencing factors. Ausburn, 502 F.3d at 329 (stating that "the court must acknowledge and respond to any properly presented sentencing argument which has colorable legal merit and a factual basis" in the record).
"After settling on the appropriate sentence, [the court] must adequately explain the chosen sentence to allow for meaningful appellate review." Gall, 552 U.S. at 50, 128 S.Ct. 586. In other words, it is not enough for the district court to carefully analyze the sentencing factors. A separate and equally important procedural requirement is demonstrating that it has done so. Because of the "fact-bound nature of each sentencing decision," there is no "uniform threshold" for determining whether a court has supplied a sufficient explanation for its sentence. Tomko, 562 F.3d at 567. In some cases, a "brief statement of reasons can be "legally sufficient." Rita, 551 U.S. at 358, 127 S.Ct. 2456. See also United States v. Lessner, 498 F.3d 185, 203 (3d Cir. 2007) (finding no procedural error notwithstanding the district court's "scant" discussion of the § 3553(a) factors). In others, a longer explanation may be appropriate. "[T]he record must be adequate for review, [but] it need not be perfect. . . . [R]eview in this area is necessarily flexible[.]" Ausburn, 502 F.3d at 328. In all cases, however, the district court must "set forth enough to satisfy the appellate court that he has considered the parties' arguments and has a reasoned basis for exercising his own legal decisionmaking authority." Rita, 551 U.S. at 356, 127 S.Ct. 2456. Stated another way, the district court must furnish an explanation "sufficient for us to see that the particular circumstances of the case have been given meaningful consideration within the parameters of § 3553(a)." United States v. Levinson, 543 F.3d 190, 196 (3d Cir. 2008). We require this explanation, not because we distrust district courts or seek to second-guess them, Tomko, 562 F.3d at 575, but because such an explanation is necessary and vital to us in performing meaningful substantive reasonableness review. See Cooper, 437 F.3d at 329 (explaining that "there is no way to review [a court's] exercise of discretion" if it "does not articulate the reasons underlying its decision" (quoting United States v. Johnson, 388 F.3d 96, 101 (3d Cir. 2004))).
The extent of the explanation we require of the district court may turn on whether the court has varied from the Guidelines range, and, if it has, on the magnitude of the variance. If the court imposes a sentence outside of the Guidelines, it "must consider the extent of the deviation and ensure that the justification is sufficiently compelling to support the degree of the variance." Gall, 552 U.S. at 50, 128 S.Ct. 586. While there is "no mathematical formula" for assessing the adequacy of a district court's explanation for a variance, Levinson, 543 F.3d at 196, "a major departure should be supported by a more significant justification than a minor one." Gall, 552 U.S. at 50, 128 S.Ct. 586. See also Levinson, 543 F.3d at 197 (stating that "we may look for a more complete explanation to support a sentence that varies from the Guidelines than we will look for when reviewing a sentence that falls within a properly calculated Guidelines range"); Ausburn, 502 F.3d at 331 n. 36 (observing that the "farther a sentence varies from the advisory guidelines range, the more compelling the judge's reasons must be").
With these principles in mind, we turn to the government's specific claims of error in this case.
III.
Significantly, the government does not challenge the substantive reasonableness of Merced's sentence. Its claims of error are entirely procedural. The government claims that the District Court essentially ignored the advisory Guidelines range, contrary to § 3553(a)(4); decided to grant a variance before the government had an opportunity to argue its position; employed a personal sentencing policy antithetical to congressional and Sentencing Commission policies, without adequate explanation; failed to adequately explain the massive downward variance it granted Merced; ignored the need to avoid unwarranted sentencing disparities, contrary to § 3553(a)(6); and failed to evaluate the seriousness of Merced's offense or his criminal history. Merced argues that the District Court committed no procedural error, and that the chosen sentence reflects the Court's careful analysis of all of the § 3553(a) sentencing factors.
Most of the government's claims of error are underwhelming. In particular, we reject the claim that the District Court repeatedly interrupted the government at the sentencing hearing, and would not allow it to present its case. That accusation is simply not supported by the record. We also reject the government's claim that the District Court ignored the applicable Guidelines range. The Court stated that it had read the PSR and the parties' sentencing memoranda, all of which referenced the correct Guidelines range. The Court acknowledged the Guidelines range three times during the sentencing hearing, and correctly adopted that range in its written SOR explaining the sentence. The record makes clear that the Court properly considered the Guidelines range as its "starting point," Smalley, 517 F.3d at 211, then varied downward because it viewed that recommendation as excessive for various reasons.
Finally, we cannot agree that the District Court failed to adequately consider Merced's criminal history or the seriousness of his offense, at least as a procedural matter. As recounted above, the Court weighed both of those factors extensively at the sentencing hearing. What the government seems really to be arguing is that the Court's choice of sentence did not afford those factors enough weight. True or not, that is a substantive complaint, not a procedural one. The "district court's failure to give [certain] factors the weight [the government] contends they deserve" does not mean that those factors were not considered. United States v. Bungar, 478 F.3d 540, 546 (3d Cir. 2007). The government cannot circumvent Tomko by repackaging a substantive claim of error as a procedural one. If the government wanted to argue that no reasonable court could have sentenced Merced to five years in prison, it should have argued that no reasonable court could have sentenced Merced to five years in prison.
With all of that said, two of the government's claims have merit. First, we agree that the District Court failed to adequately explain its apparent policy disagreement with the career offender provision of § 4B1.1, and what role, if any, that disagreement played in determining Merced's sentence. Second, we agree that the Court failed to explain how the variance it granted to Merced would not contribute to unwarranted sentencing disparities. These procedural errors compel us to vacate Merced's sentence.
A. Policy Disagreement with § 4B1.1
In the Sentencing Reform Act of 1984, Congress directed the Sentencing Commission to "assure that the guidelines specify a sentence to a term of imprisonment at or near the maximum term authorized" for adult offenders who (1) are convicted of a felony controlled substance offense or a felony crime of violence, and (2) have two or more such prior felony convictions. 28 U.S.C. § 994(h)(1)-(2). See also United States v. Sanchez, 517 F.3d 651, 662-64 (2d Cir. 2008) (tracing the origins of § 4B1.1). The Commission implemented this directive in § 4B1.1 of the Guidelines, which prescribes increased offense levels — and thus, all else equal, harsher recommended sentences — for "career offenders." Consistent with 28 U.S.C. § 994(h), § 4B1.1 provides that a defendant is a career offender if "(1) the defendant was at least eighteen years old at the time the defendant committed the instant offense of conviction; (2) the instant offense of conviction is a felony that is either a crime of violence or a controlled substance offense; and (3) the defendant has at least two prior felony convictions of either a crime of violence or a controlled substance offense." U.S.S.G. § 4B1.1(a). It is undisputed that Merced qualifies as a career offender under this definition.
At sentencing, however, the District Court stated that it "kind of reserve[s] career offender status for violent, significant drug deals, that type of thing, even though the guidelines may advise that it's appropriate." We interpret this as an expression of a policy disagreement with the Guidelines over who should (or should not) be subject to the enhanced punishment reserved for recidivists. It appears that the Court, while recognizing that Merced technically qualified as a career offender, thought that § 4B1.1 prescribed unduly harsh punishment given the low-level nature of Merced's previous crimes. In other words, the Court seemed to believe, as a matter of policy, that § 4B1.1's definition of a career offender is overbroad, and that someone like Merced with a history of mere street crimes should not be subject to the same heightened penalties as a criminal engaged in "violent, significant drug deals."
If this interpretation is wrong, the District Court will have the opportunity to correct us on remand. As a matter of best practices, of course, a district judge who sentences a defendant pursuant to a policy disagreement with the Guidelines should clearly state that he is doing so. This will minimize the risk that he will be misunderstood by a reviewing court.
A threshold issue is whether district courts may properly vary from the career offender Guidelines range based on this sort of policy disagreement. In Spears, the Supreme Court clarified its decision in Kimbrough and held that "district courts are entitled to reject and vary categorically from the crack-cocaine Guidelines based on a policy disagreement with those Guidelines." 129 S.Ct. at 843-44. There has been disagreement as to whether district courts are likewise free to vary from the Guidelines based on policy disagreements with § 4B1.1. The First, Sixth and Eighth Circuits hold that they are. United States v. Boardman, 528 F.3d 86, 87 (1st Cir. 2008); United States v. Michael, 576 F.3d 323, 327-28 (6th Cir. 2009); United States v. Gray, 577 F.3d 947, 950 (8th Cir. 2009). Opinions of the Seventh and Eleventh Circuits have taken the opposite view. See United States v. Welton, 583 F.3d 494 (7th Cir. 2009), overruled by United States v. Corner, 598 F.3d 411 (7th Cir. 2010) (en banc); United States v. Vazquez, 558 F.3d 1224 (11th Cir. 2009), vacated by Vazquez v. United States, ___ U.S. ___, 130 S.Ct. 1135, ___ L.Ed.2d ___ (2010) (mem.). Both Welton and Vazquez held that the principles of Kimbrough are inapplicable to the career offender provision because, unlike the crack cocaine Guidelines at issue in Kimbrough, § 4B1.1 was promulgated pursuant to direct statutory command. See, e.g., Welton, 583 F.3d at 496-97 (citing 28 U.S.C. § 994(h)); Vazquez, 558 F.3d at 1228-29 (same). They reasoned that "the Sentencing Guidelines may be only advisory for district judges, [but] congressional legislation is not." Welton, 583 F.3d at 496-97.
But see Sanchez, 517 F.3d at 663 (concluding that there is simply "no statutory provision instructing the court to sentence a career offender at or near the statutory maximum" because § 994(h) instructs the Sentencing Commission, not sentencing courts); Michael, 576 F.3d at 327 ("[A] directive that the Commission specify a particular Guidelines range is not a mandate that sentencing courts stay within it.").
This reasoning seems to be falling out of favor. In light of Spears, the government subsequently confessed error in both Welton and Vazquez. Neither case is good law today. We need not weigh in on this issue, however, because the parties do not disagree about it. The government concedes that a sentencing court may vary downward from the Guidelines range generated by the career offender provision based solely on a policy disagreement with the scope of that provision. We will proceed on the assumption that the government's concession is well-grounded. That assumption, however, is only the beginning of the inquiry. For even if district courts may choose a sentence at variance with the range generated by § 4B1.1, they must do so while following the same procedures that apply in every other case.
The Seventh Circuit sitting en banc overruled Welton, see Comer, 598 F.3d at 415-16; the Supreme Court vacated Vazquez at the government's request. See Vazquez v. United States, ___ U.S. ___, 130 S.Ct. 1135, ___ L.Ed.2d ___ (2010) (mem.). The Eleventh Circuit has not yet considered the issue on remand.
We can safely do so, of course, because if district courts lacked the freedom to vary based on policy disagreements with § 4B1.1, we would decide this case the same way we decide it today — the sentence would have to be vacated.
Here, the District Court committed two procedural errors. First, it never explained its statement that it "kind of reserved career offender status for large-scale or violent drug dealers. It was unclear whether the Court was describing a categorical rule that it followed in all cases, or an informal rule of thumb that it applied only in certain cases (perhaps, but not necessarily, including Merced's). The Court mentioned its personal sentencing policy only once, almost in passing, and said nothing about it at the time it announced Merced's sentence. Thus, the extent to which its disagreement with the scope of § 4B1.1 ultimately affected Merced's sentence remains a mystery. The Court identified several unrelated mitigating factors that weighed in favor of a downward variance, such as Merced's difficult childhood, his drug problem, his strong relationship with his longtime girl-friend, and his resolve to take care of his children. Along the same lines, the Court wrote in the SOR that it was basing the variance on its assessment of § 3553(a)(1) (the nature and circumstances of the offense and the history and characteristics of the defendant). In theory, the entirety of the variance could have been based on those mitigating factors. The Court's stated reluctance to assign career offender status absent "violent" or "significant" drug offenses could have been a stray comment, or it could have been central to the choice of sentence. We cannot tell on this record. The District Court must do more to "plainly state the reasoning" behind its sentence. Levinson, 543 F.3d at 197. See also Grier, 475 F.3d at 571-72 (requiring "a sufficiently detailed explanation" so as to allow for "effective review"). For that reason alone, we must vacate the sentence and remand with instructions for the Court to explain its sentence more clearly.
Indeed, just after the Court revealed its personal sentencing policy concerning § 4B1.1, it hastened to add that "I'm not saying by any means the guidelines application here is inappropriate or not proper . . . . anything I do would be by way of a variance looking at" the low-level nature of Merced's crimes and the fact that Merced had a drug problem. Then again, the Court also warned Merced that he needed to change his ways because "another judge" might apply the career offender Guidelines, thereby implying that this judge had decided not to apply them. These statements only add to our confusion as to whether the Court's policy disagreement with § 4B1.1 informed Merced's sentence, or whether the variance was based on Merced's personal characteristics.
The Court's handling of the § 4B1.1 issue is further clouded by its conflicting statements about the strength of Merced's argument that § 4B1.1 was analogous to the crack cocaine Guidelines at issue in Kimbrough. At one point, the Court told Merced's lawyer that he need not reprise the § 4B1.1 argument made in his memorandum because "I fully understand your argument . . . . and I buy into it to a large extent." Later, however, when the government attempted to refute Merced's Kimbrough analogy, the Court said, in reference to Merced's argument, "no, no, I'm not buying into that."
Second, if the District Court intends to vary downward based on a policy disagreement with § 4B1.1, it must better explain and justify that decision. The freedom to vary from the career offender Guidelines, assuming it exists, is not free. Its price is a reasoned, coherent, and "sufficiently compelling" explanation of the basis for the court's disagreement. Lychock, 578 F.3d at 219. Examination of our precedents reveals that the District Court's explanation here fell short of what is required.
In Lychock, the defendant pleaded guilty to possessing dozens of images of child pornography. The Guidelines recommended a prison term of 30 to 37 months, but the district court refused to impose any jail time. The court's reasoning rested in part on its view that imprisoning Lychock would neither protect the public nor deter future child pornography possession by others. Id. at 216-17. The court explained:
The only benefit I could see [to imprisonment would be] as a deterrent to others, and that is a factor. . . . So other people would recognize that they cannot subscribe to these images with impunity. I am not persuaded that a jail term for this defendant warrants, or is to be equated with that value. The kind of psychological problem in persons who are drawn to this kind of material it seems to me is not going to be deterred by a jail term for an internet porno observer. There is no suggestion the public otherwise is threatened by his conduct.
Id. at 217 (alterations in original). We interpreted this explanation as reflecting a policy disagreement with the Guidelines, which embodied the reasoned judgment of Congress and the Sentencing Commission that the goals of sentencing would be served by imposing jail time on child pornography consumers like Ly-chock. Id. at 219. We allowed that such disagreement was permissible, but only if the court provided a "sufficiently compelling," "reasoned explanation for its apparent disagreement with the policy judgments" reflected by the Guidelines range. Id. We concluded that the district court's justification for its lenient sentence, a mere "conclusory statement of personal belief about the benefits vel non of imprisonment, was insufficient support for its refusal to follow the policies embodied in the Guidelines. Id. at 220. Accordingly, we vacated the sentence.
As discussed in Section III.B, this failure to explain its policy disagreement with the Guidelines was not the only error that led us to vacate the sentence. The district court in Lychock also failed to consider the need to avoid unwarranted sentencing disparities pursuant to § 3553(a)(6).
We followed similar reasoning in Levinson. There, the defendant was the owner of a failing water cooler business who engaged in an elaborate fraud to make his company appear profitable. He ordered employees to shred documents, destroy electronic records, and create phony sales reports. He also filed false tax returns which deprived the government of over $40,000 in revenue. Levinson, 543 F.3d at 192. In reliance on Levinson's misrepresentations, a parent company, Elkay, invested millions of dollars in his business. Id. at 191. Levinson eventually pleaded guilty to wire fraud and filing a false tax return. He also reached a civil settlement with Elkay. Id. at 192. The Guidelines recommended a prison term of 24 to 30 months, but the district court imposed probation only. Id. at 192, 194. It reasoned that
[Levinson] put the appearance of prosperity above his respect for the law. Balanced against this is the propriety of putting into jail at a substantial cost to the public a nonviolent offender who poses little or no threat to the public and whose crimes had little impact beyond his business partners and his family. . . .
When I look at the costs associated with putting someone like Mr. Levinson [in] jail in this day and age compared to the harm he has caused, which has been resolved amicably with his business and which certainly will impose even more harm on his family, I just can't see that it makes much sense. I just do not.
Id. at 194 (emphasis added, some alterations in original). We found this explanation for the court's substantial downward variance insufficient, and vacated the sentence. We interpreted the district court's refusal to imprison Levinson, based on its view that the cost of imprisonment outweighed its benefits, as a policy disagreement with the Guidlines. While reiterating that such "[p]olicy considerations are not off-limits in sentencing," id. at 200, we emphasized that sentences influenced by policy disagreements must be accompanied by a "thorough explanation" from the court. Id. at 201. The district court's bare, unsupported assertion that jail time was too costly to be worthwhile — simply because Levinson's fraud was "white-collar" and directed at a private entity who had been made whole through a civil settlement — was insufficient to justify ignoring the "very deliberate policy choices" reflected in the Guidelines sentence. Id. at 200.
Again, we assume for present purposes that the freedom district courts enjoy under Kimbrough and Spears includes the freedom to vary from a career offender Guidelines range based on a policy disagreement. However, "such disagreement is permissible only if a District Court provides `sufficiently compelling reasons to justify it.'" Lychock, 578 F.3d at 219 (quoting Gall, 552 U.S. at 50, 128 S.Ct. 586). A "sufficiently compelling" explanation is one that is grounded in the § 3553(a) factors. The authors of the Guidelines, no less than district courts, have been tasked with ensuring that criminal sentences meet the goals of sentencing set forth in § 3553(a). Rita, 551 U.S. at 348, 127 S.Ct. 2456 (explaining that "both the sentencing judge and the Commission . . . [carry] out the same basic § 3553(a) objectives, the one, at retail, the other at wholesale."). Thus, the Guidelines reflect the Sentencing Commission's "rough approximation of sentences that might achieve § 3553(a)'s objectives." Id. at 350, 127 S.Ct. 2456. If a district court concludes that those objectives are not achieved by a sentence within the career offender Guideline range, and that belief is driven by a policy disagreement with the career offender provision, then the court must explain why its policy judgment would serve the § 3553(a) sentencing goals better than the Sentencing Commission's judgments. In doing so, he should take into account all of the sentencing factors, not just one or two of them in isolation. We require this explanation "so that, on appeal, we can determine whether the [court's] disagreement is valid in terms of the § 3553 factors, the Sentencing Guidelines, and the perception of fair sentencing." Lychock, 578 F.3d at 219 (internal quotations omitted).
For example, a district judge who distinguishes between kingpins and street level drug dealers for purposes of § 4B1.1 may be furthering the goal of imposing sentences that reflect "the history and characteristics of the defendant" and "the seriousness of the offense." 18 U.S.C. § 3553(a)(1), (2)(A). Even so, he should also consider whether such a distinction serves other sentencing goals, such as the need to "promote respect for the law" and "protect the public from further crimes of the defendant." 18 U.S.C. § 3553(a)(2)(A), (C).
Here, the District Court's explanation of its policy disagreement fell short even of the explanations found wanting in Lychock and Levinson. In fact, the Court offered no real explanation at all — only a suggestion as to what its personal sentencing practices are, in light of that disagreement. If the District Court has a policy disagreement about the scope of § 4B1.1 — about who should and should not be subject to the enhanced sentences reserved for recidivists — then it might, in light of Kimbrough and Spears, permissibly sentence Merced according to that disagreement. But first, it must provide a "thorough explanation" of its reasoning. Levinson, 543 F.3d at 201. So far, the District Court has offered little more than a "conclusory statement of personal belief that career offender status should be reserved for violent or large-scale drug dealers. Lychock, 578 F.3d at 220. This is inadequate, and constitutes procedural error.
B. Sentencing Disparities
As noted, a district court is not required to recite and make findings as to every one of the § 3553(a) factors, as long as the record makes clear that the factors have been considered in deciding the sentence. Cooper, 437 F.3d at 329. Where one party raises a colorable argument about the applicability of one of the factors, however, the court should respond to that argument as part of its "meaningful consideration of the relevant statutory factors and the exercise of independent judgment." Grier, 475 F.3d at 571-72. One factor the court must consider is the need to avoid "unwarranted sentencing disparities." 18 U.S.C. § 3553(a)(6). Its failure to do so in the face of a colorable argument that an outside-the-Guidelines sentence will create a risk of such disparities constitutes procedural error. See, e.g., Lychock, 578 F.3d at 220; Goff, 501 F.3d at 258; Ausburn, 502 F.3d at 330.
As explained above, in Lychock, the defendant pleaded guilty to possession of child pornography and his Guidelines range was 30 to 37 months. The district court sentenced Lychock to probation and a fine, but declined to impose any jail time. This decision was based in part on its policy view that imprisoning Lychock would neither protect the public nor deter child pornography crime, Lychock, 578 F.3d at 216-17, but also in part on the court's analysis of several other § 3553(a) factors. The court acknowledged that Ly-chock had committed "a serious offense," but described him as "basically [a] law-abiding . . . young man." Id. at 216. It further noted Lychock's "cooperation with law enforcement, his acknowledgment of wrongdoing, his supportive family, his decision to seek psychological help immediately, . . . the report of his psychologist that he was benefitting from their sessions," id. at 216, and the unlikelihood that imprisoning Lychock would deter criminal conduct or protect the public. Id. at 216-17; see also 18 U.S.C. § 3553(a)(2). Nevertheless, we held that the district court "failed to properly consider the § 3553 factors" and vacated the sentence. Ly-chock, 578 F.3d at 218. We found no evidence that the district court considered the need to avoid unwarranted sentencing disparities, even though the government had specifically cited that need in its sentencing memorandum and pointed out that several defendants caught in the same investigation that snared Lychock had received sentences within the Guidelines range. Id. at 219. We emphasized that courts need not specifically discuss each of the sentencing factors in every case, nor must they analyze every frivolous argument advanced by a party in a sentencing proceeding. Id. But we also reiterated that where "the sentence imposed is `far below the sentences given to similar offenders,'" the risk of disparities should be analyzed with "particular care." Id. (citing Goff, 501 F.3d at 256). Goff unfolded similarly. In that case, the defendant pleaded guilty to possession of child pornography and his advisory Guidelines range was 37 to 46 months. Goff 501 F.3d at 252. Nevertheless, the district court sentenced him to only four months in prison because it concluded that Goff was, in many respects, a model citizen. Id. at 253. He had no criminal history and, in the court's estimation, "had lived an exemplary life" before his arrest. Id. The court reasoned that Goff had committed a "victimless crime," and that his interest in child pornography did not necessarily mean that he was a danger to the community. Id. We vacated the sentence, faulting the district court for failing to analyze the "`need to avoid unwarranted sentence disparities among defendants with similar records who have been found guilty of similar conduct,' as required by § 3553(a)(6)." Id. at 256. We observed that a defendant caught in the same investigation that led to Goff's arrest received 28 months in prison based on a Guidelines range of 27 to 33 months. Id. at 261. In light of that fact, "discussion of [§ 3553(a)(6)] should have been undertaken with particular care," because Goff's sentence created "a potential disparity in sentence for those convicted of child pornography . . . based on little, if anything, more than [his] luck" in assignment of judge. Id.
The need to consider the risk of unwarranted disparities also motivated this Court's decision in Ausburn. Unlike Ly-chock and Goff which were government appeals of below-Guidelines sentences, Ausburn involved a defendant's appeal of an above-Guidelines sentence. The defendant in Ausburn was a police officer who engaged in a sexual relationship with a fourteen-year-old girl whom he had met while responding to a call at her home. Ausburn, 502 F.3d at 316. When the relationship was discovered, Ausburn was arrested and quickly pleaded guilty. His Guidelines range was 57 to 71 months. Id. at 318. He argued for a sentence comparable to that which had recently been handed down by the same judge in a case called Kenrick, which also involved a sexual offense against a minor. He contended that a comparable sentence was necessary to ensure "the appearance of fairness" and to avoid unnecessary disparities in sentencing. Id. at 317-18. The district court rejected his request for leniency. After a cursory recitation of the § 3553(a) factors, it handed down a sentence of 144 months. The court explained only that Ausburn had committed a "serious" offense and that his "position as a law enforcement officer [made his] violation of the law that much more unacceptable." Id. at 320. The court added, "your conduct was totally unacceptable and you deserve double the number that I set forth in the Sentencing Guidelines." Id. at 321. Ausburn's lawyer objected, arguing that "this case, as I see it, is very . . . close to the Kenrick case, where the Court did impose [forty-six] months." Id. (some alterations in original). The court responded that it considered four or even six years to be insufficient, in light of Ausburn's abuse of his position as a police officer. Id. He dismissed Ausburn's argument that his sentence should be comparable to the one imposed in Kenrick, saying that if he didn't like the sentence he could take it up on appeal. Id. We vacated the sentence, finding error because "the District Court did not address Ausburn's argument that two cases recently decided in the same district — both of which concerned sexual offenses involving minors — provided bench marks for determining a proper sentence, and that the court should hew close to the sentences in those cases" in order to avoid unwarranted sentencing disparities. Id. at 330. We found that the court's terse explanation that four to six years was "insufficient" punishment, in light of Ausburn's abuse of his position as a police officer, was "inadequate" to demonstrate that it had meaningfully considered Ausburn's sentencing disparities argument. Id. Lychock, Goff, and Ausburn demonstrate that a district court's failure to analyze § 3553(a)(6) may constitute reversible procedural error, even where (as here) the court engages in thorough and thoughtful analysis of several other sentencing factors. In other words, meaningful consideration of the nature of the offense, the characteristics of the defendant, the need to protect the public, the need to promote deterrence, etc., may not save a sentence if the sentence is imposed without considering the risk of creating unwarranted disparities, and the sentence in fact creates such a risk. See Lychock, 578 F.3d at 219 (stating that where "the sentence imposed is `far below the sentences given to similar offenders,' consideration of [sentencing disparities] deserves `particular care'" (quoting Goff 501 F.3d at 256)). This is especially true if the sentence falls outside of the Guidelines, or where, as in Lychock and Ausburn, a party specifically raises a concern about disparities with the district court and that argument is ignored.
Here, as in Lychock, the government voiced unmistakable concern that granting Merced a significant variance could create unwarranted sentencing disparities. In his sentencing memorandum, Merced requested a below-Guidelines sentence of less than 10 years; the government opposed that request in part because it argued that such a sentence would create unwarranted sentencing disparities between Merced and other recidivist crack cocaine dealers. This was, at a minimum, a "colorable legal argument" with a "factual basis" in the record. Cooper, 437 F.3d at 329. The District Court should have addressed it as part of its meaningful consideration of the sentencing factors. It never did so explicitly, and just as in Ly-chock, Ausburn, and Goff, there is no evidence that the District Court accounted for this factor at all, notwithstanding its thoughtful analysis of other factors. We reiterate that sentencing courts need not respond in detail and on the record to each and every argument presented by the parties. But they should respond to colorable arguments with a factual basis in the record. If there was some indication in the record that the District Court had considered the risk of unwarranted sentencing disparities, we might overlook its failure to explicitly analyze that factor at length, on the record, at the hearing. See Cooper, 437 F.3d at 329 (explaining that a court need not "discuss and make findings as to each of the § 3553(a) factors if the record makes clear the court took the factors into account in sentencing") (emphasis added). Nothing in this record, however, indicates that the District Court considered § 3553(a)(6) at all, despite the government's emphasis on that argument in its sentencing memorandum and the risk of disparities that Merced's sentence undoubtedly created. See Ausburn, 502 F.3d at 331 ("Where the record is inadequate, we do not fill in the gaps by searching the record for factors justifying the sentence.").
In particular, the Court might have considered the fact that if Merced had possessed just .9 grams more of cocaine, i.e., 50 grams, he would have been subject to a mandatory minimum sentence of 10 years in prison. See 21 U.S.C. § 841(b)(1)(A). See also Kimbrough, 552 U.S. at 108, 128 S.Ct. 558 (stating that as part of its § 3553(a)(6) analysis, a sentencing court "must take account of sentencing practices in other courts and the `cliffs' resulting from . . . statutory mandatory minimum sentences").
Merced's 60-month sentence was 128 months less than what a similarly situated recidivist crack cocaine dealer could expect to receive under the circumstances. See, e.g., United States v. Tupuola, 587 F.3d 1025, 1026 (9th Cir. 2009) (describing career offender who pleaded guilty to distribution of crack cocaine, had an advisory Guidelines range of 188-235 months, and received a 188-month sentence); Welton, 583 F.3d at 495-96 (same); United States v. Sharkey, 543 F.3d 1236, 1238 (10th Cir. 2008) (same); United States v. Torres, 541 F.3d 48, 50-51 (1st Cir. 2008) (describing defendant who sold 10 grams of crack cocaine, qualified as a career offender, fell within the same Guidelines range as Merced, and received a 195-month sentence). Before the District Court granted such a large variance, it should have explained why that variance would .not contribute to unwarranted sentencing disparities pursuant to § 3553(a)(6). Its failure to do so was procedural error.
IV.
In summary, we hold that the District Court committed two errors. First, it may have sentenced Merced pursuant to a personal policy disagreement with the Guidelines; specifically, disagreement with the scope of the career offender provision of U.S.S.G. § 4B1.1. While granting a variance on such grounds may be permissible, the District Court must, at the very least, explain itself more thoroughly than it has so far. On remand, the District Judge should clearly explain whether he is granting a variance based on a policy disagreement with § 4B1.1. If so, he must justify that decision to the extent required by our precedents. Second, the District Court failed to analyze a highly relevant sentencing factor, § 3553(a)(6). The Court's choice of sentence may have created a risk of unwarranted disparities between Merced and similarly situated recidivist crack cocaine dealers. The Court should have considered this issue, and addressed the government's argument that a Guidelines sentence was necessary to promote uniformity in sentencing.
Our insistence that sentencing courts follow the requisite procedures by no means diminishes the "superior position" of the district judge to make those determinations so critical to a just and reasonable sentence. See Tomko, 562 F.3d at 566. Rather, requiring a "reasoned and rational justification" on the record for the sentence chosen serves several critical purposes. Grier, 475 F.3d at 572. Most basically, it allows us to fulfill our "important role," Tomko, 562 F.3d at 575, of exercising "effective appellate over-sight" as required by Supreme Court precedent. Grier, 475 F.3d at 572. But it also serves an even higher purpose within our system of criminal justice. Requiring coherent explanations of sentence "instill[s] public confidence in the judicial process." Id. It "assure[s] the parties of the fairness of the proceedings," "demonstrat[es] the serious reflection and deliberation that underlies each criminal sentence," and "offers the defendant, the government, the victim, and the public a window into the decision-making process and an explanation of the purposes the sentence is intended to serve." Id. See also Rita, 551 U.S. at 356, 127 S.Ct. 2456 ("Confidence in a judge's use of reason underlies the public's trust in the judicial institution. A public statement of those reasons helps provide the public with the assurance that creates that trust.") We believe that each of these important interests is served by our decision today.
"We do not suggest that the original sentence reflects anything less than the sound judgment of the District Judge, or that the final sentence should necessarily differ from the one previously imposed." Grier, 475 F.3d at 572. We simply ask for a clearer, more complete explanation than the District Court has offered to date. We do not think that this asks too much. Demanding close adherence to procedural requirements — including the requirement that sentencing courts explain their reasoning with clarity — is, we think, more than fair in light of the deference we afford to district courts as a substantive matter under Tomko.
The judgment of sentence will be vacated, and the case remanded for re-sentencing.
I fully join my colleagues in concluding that the District Court erred procedurally by failing to explain its apparent policy disagreement with the career offender provision, U.S.S.G. § 4B1.1, and by failing to consider the need to avoid unwarranted sentencing disparities, 18 U.S.C. § 3553(a)(6). I write separately to explain why I do not doubt that district courts are free to vary from the career offender Guideline range based on a categorical policy disagreement with § 4B1.1. Because of the importance of this question, I believe we should resolve it in this case, despite the fact that the parties do not dispute it. Cf. Swift Co. v. Hocking Valley Ry. Co., 243 U.S. 281, 290, 37 S.Ct. 287, 61 L.Ed. 722 (1917) (noting that a "court cannot be controlled by agreement of counsel on a subsidiary question of law"). Indeed, I find it odd that we vacate the District Court's sentence with instructions to explain its policy disagreement without first concluding that it was authorized to vary on that ground in the first place.
In addition, though I agree with my colleagues that the District Court did not "ignore[]" the Guideline range, Maj. Op. at 216, I am unsure that it gave the Sentencing Guidelines "the consideration they are due" in reaching its sentence. United States v. Goff, 501 F.3d 250, 256 (3d Cir. 2007).
At sentencing, the Court quickly rejected the propriety of a sentence within both the career offender Guideline range of 188 to 235 months and the Guideline range that applied in the absence of the career offender enhancement — 110 to 137 months, calculated under the drug quantity table in U.S.S.G. § 2D1.1(c). Though the Court subsequently articulated a reason for rejecting the career offender Guideline range, and found that a sentence within that range was not "near a reasonable sentence," App. at 76, it did not explain why a sentence within the 110 to 137 month-range was also inappropriate. A sentence within this range would seemingly take into account the Court's predominant concern that the amount of crack cocaine possessed by Merced was "relatively small" when compared to the "kilograms and pounds of stuff" that it "usually sees," id. at 77, as the offense levels provided for in § 2D1.1(c) are tied specifically to the amount of crack cocaine involved in the offense. Compare U.S.S.G. § 2D 1.1 (c)(2) (providing for an offense level of 36 for an offense involving between 1.5 and 4.5 kilograms of crack cocaine), with § 2D1.1 (c)(6) (providing for an offense level of 28 for an offense involving between 35 and 50 grams of crack cocaine). In that light, I cannot conclude with conviction that the Court gave the Guidelines "the consideration they are due" as a factor under § 3553(a). Goff, 501 F.3d at 256; 18 U.S.C. § 3553(a)(4)(A).
I.
It is, of course, not always the case that, when a district court varies from the career offender Guideline range, it has done so based on a "policy disagreement" with § 4B1.1. Rather, a variance from that range often may reflect a sentencing court's determination that, "despite meeting the formal criteria for career offenders," the defendant's individual circumstances "fall outside the guideline's heartland or intended scope." United States v. Pruitt, 502 F.3d 1154, 1170 (10th Cir. 2007) (McConnell, J., concurring), vacated on other grounds, ___ U.S. ___, 128 S.Ct. 1869, 170 L.Ed.2d 741 (2008); see also United States v. Martin, 520 F.3d 87, 95-96 (1st Cir. 2008) (district court's variance from career offender Guideline range was not a "repudiation of the policies embodied in the [S]entencing [G]uidelines," as the court had "grounded the defendant's sentence in case-specific considerations," including the court's conclusion that the defendant was a "changed man" who would "not re-offend").
These types of justifications were also typical of pre- Booker downward departures in calculating the sentencing range under the Guidelines for defendants qualifying as career offenders. See Michael S. Gelacak et al., Departures Under the Federal Sentencing Guidelines: An Empirical and Jurisprudential Analysis, 81 Minn. L. Rev. 299, 356-57 n. 248 (1996) (reporting that a frequent justification for a sentence below the career offender Guideline range was that the defendant's predicate offenses were "minor or too remote in time to warrant consideration").
Turning to our case, this is how Merced seeks to characterize the District Court's variance: he argues that, rather than basing its sentencing decision "on a categorical policy about the career offender provision in general[,] . . . the District Court grounded its sentence firmly in the individual circumstances of Mr. Merced's case, as filtered through the § 3553(a) factors." (Appellee's Br. at 28.) The Court similarly characterized its sentence as solely based on its evaluation of "the nature and circumstances of the offense and the history and characteristics of Merced, 18 U.S.C. § 3553(a)(1), rather than a criticism of the career offender provision itself.
Yet the Court's stated policy of reserving a sentence within the career offender Guideline range for "significant, violent drug deals" is essentially a rejection of § 4B1.1's policy of treating repeat drug offenders — regardless of the quantity of drugs involved or whether the defendant's offense or prior offenses involved violence — as offenders whose Guideline ranges should be at or near the statutory maximum sentence. Stated another way, the Court imposed new, categorical factors (the quantity of drugs involved and whether the offense or predicate offenses involved violence) in determining the appropriateness of a sentence within the career offender Guideline range. Cf. United States v. Moreland, 437 F.3d 424, 436 (4th Cir. 2006) (noting that the career offender provision is "fraught with potential imprecision," and "covers a broad range of offenders, encompassing the street-level dealer who handles only small quantities of drugs and the drug kingpin or the recidivist with a history of violence") (internal quotation marks omitted); see also Pruitt, 502 F.3d at 1167 (McConnell, J., concurring) (noting that, under § 4B1.1, "it does not matter, for sentencing purposes, whether [the defendant's] prior drug felonies were large-scale or petty, violent or nonviolent"). In that light, I believe the Court's variance is best understood as motivated by a policy that "applies to a wide class of offenders or offenses," United States v. Cavera, 550 F.3d 180, 191 (2d Cir. 2008) ( en banc) — and one that disagrees with the policy underlying § 4B1.1 — rather than a straightforward application of the § 3553(a) factors to reach a reasonable sentence for a defendant whose circumstances fall outside the intended scope of the career offender provision.
Though the line between a variance based on a "policy disagreement" with § 4B1.1 and one based on an "individualized determination" of a particular career offender's circumstances is less than precise, it is vital that we give district courts explicit guidance regarding their authority to vary from the career offender Guideline range on policy grounds. First, doing so will discourage courts from "masking" their policy disagreements as "individualized determinations," an "[unacceptable sentencing practice" that the Supreme Court has described as "institutionalized subterfuge." Spears v. United States, ___ U.S. ___, ___, 129 S.Ct. 840, 844, 172 L.Ed.2d 596 (2009). Second, the distinction between these types of variances may have important implications for the scope of our appellate review. When a district court finds that a defendant's circumstances place him outside the "heartland" of defendants to whom § 4B1.1 was intended to apply, its decision to vary from the career offender Guideline range presumably will be entitled to the "greatest respect." Kimbrough v. United States, 552 U.S. 85, 109, 128 S.Ct. 558, 169 L.Ed.2d 481 (2007) (a district court's decision to vary from the Guidelines "may attract greatest respect when the sentencing judge finds a particular case outside the `heartland' to which the Commission intends individual Guidelines to apply") (internal quotation marks omitted). By contrast, a variance for a career offender whose circumstances place him within the "heartland" of defendants to whom § 4B1.1 was intended to apply — and thus "is necessarily based on a policy disagreement with the Guidelines," Spears, 129 S.Ct. at 843 — may be subject to "closer review" and entitled to less deference. Kimbrough, 552 U.S. at 109, 128 S.Ct. 558; Spears, 129 S.Ct. at 843 (the "implication" of Kimbrough is that "an `inside the heart-land' departure . . . may be entitled to less respect").
Though we are not required in this case to determine whether the District Court's seeming policy disagreement survives such "closer review" (as the Court did not explain its policy in light of the § 3553(a) factors), I nonetheless believe that we should determine (before remanding for resentencing) whether a sufficiently explained policy disagreement with § 4B1.1 is a permissible ground on which to vary from the Guidelines. I address this second question below.
II.
Until recently, the Government had taken the position that district courts were not free to vary from the career offender Guideline range on policy grounds. The Circuit Courts for the Seventh and Eleventh Circuits previously agreed, and concluded that district courts are not authorized under Kimbrough to vary based on policy disagreements with the career offender provision because § 4B1.1 was promulgated by the Sentencing Commission in direct response to a statutory directive in 28 U.S.C. § 994(h). See United States v. Welton, 583 F.3d 494 (7th Cir. 2009), overruled by United States v. Corner, 598 F.3d 411 (7th Cir. 2010) ( en banc); United States v. Vazquez, 558 F.3d 1224 (11th Cir. 2009), vacated, ___ U.S. ___, 130 S.Ct. 1135, ___ L.Ed.2d ___ (2010).
As the majority notes, this position is "falling out of favor." Maj. Op. at 218. The Government has now abandoned it, see Appellant's Br. at 24 (conceding that policy disagreements "may be the basis for varying from the career offender [G]uideline"), Welton was overruled by the en banc Seventh Circuit Court, and Vazquez was vacated by the Supreme Court. Moreover, each of the First, Sixth, and Eighth Circuit Courts have concluded that, after Kimbrough, district courts may vary from the career offender Guideline range based on a policy disagreement (just as they may for any provision of the Guidelines). See United States v. Gray, 577 F.3d 947, 950 (8th Cir. 2009); United States v. Michael, 576 F.3d 323, 327-28 (6th Cir. 2009); United States v. Boardman, 528 F.3d 86, 87 (1st Cir. 2008) ("[W]e do not see why disagreement with the Commission's policy judgment (as expressed in [§ 4614]) would be any less permissible a reason to deviate than disagreement with the guideline policy judgment at issue in Kimbrough.").
Though the Second Circuit Court has not precisely held that Kimbrough authorizes district courts to vary from the career offender Guideline range on policy grounds, it similarly has rejected the argument that § 994(h) restricts district courts' sentencing authority. See United States v. Sanchez, 517 F.3d 651, 663-65 (2d Cir. 2008).
I have no hesitancy reaching the same conclusion. Section 994(h) is directed to the Sentencing Commission, not sentencing courts, and does not purport to limit their sentencing discretion. See Michael, 576 F.3d at 328 ("By its terms, [§ 994(h)] tells the Sentencing Commission, not the courts, what to do."); Sanchez, 517 F.3d at 663 (same). And though Kimbrough cited § 994(h) as an example of "Congress . . . direct[ing] sentencing practices in express terms," 552 U.S. at 103, 128 S.Ct. 558, it did not thereby suggest that the policies reflected in § 994(h) are binding on sentencing courts. Rather, Kimbrough cited § 994(h) in the context of explaining why a different statutory provision, 21 U.S.C. § 841(b), was not binding on the Sentencing Commission. Kimbrough, 552 U.S. at 102-03, 128 S.Ct. 558 (rejecting the Government's argument that § 841(b) required the Sentencing Commission to establish offense levels for crack cocaine offenses reflecting the crack/powder cocaine sentencing disparity imbedded in that statute). See Corner, 598 F.3d at 415-16 ( Kimbrough referred to § 994(h) solely in the context of explaining why "the crack/powder ratio in the Guidelines was the choice of the Commission rather than [mandated by] Congress [in 21 U.S.C. § 841(b)]"). In sum, though Congress required the Commission to follow certain policies in crafting the career offender provision, neither Kimbrough nor § 994(h) supports the conclusion that a district court is prohibited from considering a policy disagreement with § 461.1 in sentencing a career offender.
Section 994(h) provides that
[t]he Commission shall 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 — (1) has been convicted of a felony that is . . . (B) an offense described in . . . 21 U.S.C. § 841 . . . and (2) has previously been convicted of two or more prior felonies, each of which is . . . (B) an offense described in . . . 21 U.S.C. § 841[.]
28 U.S.C. § 994(h) (emphasis added).
Moreover, our Court recently rejected the basic premise underlying Welton and Vazquez in the context of sentencing disparities resulting from "fast-track" programs, which apply in certain judicial districts and authorize a downward departure if a qualifying illegal immigrant defendant pleads guilty and waives his or her appellate rights. See United States v. Arrelucea-Zamudio, 581 F.3d 142, 145-46 (3d Cir. 2009) ( citing U.S.S.G. § 5K3.1); see also Vazquez, 558 F.3d at 1229 (concluding that district courts may not vary based on a policy disagreement with the career offender provision because the enhanced sentencing ranges provided for in § 4B1.1, like the disparities resulting from fast-track programs, are "the result of `direct congressional expression'"). There we rejected the Fifth, Ninth, and Eleventh Circuit Courts' conclusion that district courts may not consider disparities resulting from the limited availability of the "fast-track" program because the Guideline provision authorizing the downward departure was the result of a congressional directive in the 2003 PROTECT Act. See id. at 149-53. We criticized these Courts' "attempt to distinguish fast-track programs from the sentencing guidance provided in Kimbrough [] and [to] constrain a district court's sentencing discretion solely on the basis of a congressional policy argument," which we characterized as "an attempt to manipulate the advisory character of the Guidelines." Id. at 151. That observation applies equally here; though both § 994(h) and the PROTECT Act direct the Sentencing Commission to promulgate Guideline provisions reflecting certain policies, neither restrains a district court's sentencing discretion under Booker.
We did so even though the Government had declined to argue that "congressional policy concerning fast-track programs prohibited the exercise of a district court's discretion." Id. at 150 n. 8.
* * * * * *
In sum, I believe, and would hold, that the Supreme Court's reasoning in Kimbrough extends to § 4B1.1 and that district courts are authorized to vary from the career offender Guideline range on policy grounds. Save these statements and my supplemental comment in the first note of this concurrence, I join Judge Smith's excellent opinion in full.