592 F.3d 621

MICHAEL, Circuit Judge,
concurring in part and dissenting in part:
I concur in part III.A of the majority opinion, which holds that Kenneth Manigan did not waive his right to appeal his sentence. I respectfully dissent, however, from the majority’s determination that the district court properly enhanced Manigan’s sentence under guideline § 2Dl.l(b)(l) for possession of a dangerous weapon. Manigan’s handguns were not connected to the drug crimes he committed, and the enhancement should not apply.
Section 2Dl.l(b)(i )’s weapon enhancement may be the most frequently applied enhancement in all the guidelines. See U.S. Sentencing Commission, Sourcebook of Federal Sentencing Statistics tbl. 3 (2008) (listing § 2D1.1 as the most frequently applied guideline); U.S. Sentencing Commission, Application of Weapon Specific Offense Characteristic in § 2D1.1 Cases for Each Drug Type (2008) (enhancement applied in 2,845 out of 23,637 cases). This frequency of use makes it imperative that the boundaries for application of the enhancement remain clear and meaningful. Here, the majority’s error is compounded by the fact that it blurs a previously clear boundary, leaving district courts to wonder whether any real limitation remains. To maintain clarity, we should hold to the purpose and scope of the enhancement as recognized until today.
Section 2Dl.l(b)(l) reads: “If a dangerous weapon (including a firearm) was possessed, increase by 2 levels.” U.S. Sentencing Guidelines Manual § 2D1.1(b)(1) (2007). Application note 3 to § 2D1.1 describes the enhancement’s purpose and reach:
The enhancement for weapon possession reflects the increased danger of violence when drug traffickers possess weapons. The adjustment should be applied if the weapon was present, unless it is clearly improbable that the weapon was connected with the offense. For example, the enhancement would not be applied if the defendant, arrested at his residence, had an unloaded hunting rifle in the closet.
The application note confirms what is obvious from common sense, but perhaps not from a literal reading of § 2Dl.l(b)(l): mere possession is not enough for application of the enhancement. The majority notes this limitation, citing United States v. McAllister’s rule that the government must prove that “the weapon was possessed in connection with drug activity that was part of the same course of conduct or common scheme as the offense of conviction.” 272 F.3d 228, 233-34 (4th Cir.2001) (emphasis added). Thus, for the enhancement to apply, the government must prove not only that the defendant possessed the weapon, but also that the weapon had a role in the offense of conviction. In the case of a gun, it need not have been discharged, brandished, or even kept on the defendant’s person during the offense, see United States v. Johnson, 943 F.2d 383, 386 (4th Cir.1991), but it must have served the defendant’s purpose in carrying out the drug crime for which he was convicted. It is sufficient, for example, that the weapon was “readily available to protect” the defendant or his drugs, even if the defendant never specifically intended to use it. United States v. Thongsophapom, 503 F.3d 51, 58 (1st Cir.2007).
Unless the defendant’s words or actions explained the weapon’s role, proving that the § 2Dl.l(b)(l) enhancement applies will require evidence that the weapon had a temporal and spatial nexus to the offense. United States v. Willie, 462 F.3d 892, 899 (8th Cir.2006); United States v. Clark, 415 F.3d 1234, 1241 (10th Cir.2005); United States v. Partida, 385 F.3d 546, 562 (5th Cir.2004). When the defendant’s possession of the weapon is not established until long after the offense, the temporal nexus is not satisfied. See Clark, 415 F.3d at 1242 (rejecting application of enhancement when defendant was found in possession of gun 15 months after offense of conviction). Similarly, when the weapon is not in close proximity to the drugs or a drug-related transaction, the spatial nexus is not satisfied. See United States v. Harris, 128 F.3d 850, 852 (4th Cir.1997). In United States v. Rusher we held that the § 2Dl.l(b)(l) enhancement applied when “one of the firearms was located in the same briefcase that contained the drugs, and the others, fully loaded, were found in the bed of the same truck.” 966 F.2d 868, 880 (4th Cir.1992). In United States v. Nelson we held that the enhancement applied when guns were found with defendants in “a place of manufacturing, storing and distributing [of] crack cocaine.” 6 F.3d 1049, 1056 (4th Cir.1993). Of course, even if a temporal and spatial nexus exists, other evidence might establish that it was clearly improbable that the gun had a role in the offense. For example, even if an unloaded hunting rifle was in a nearby closet in the defendant’s residence during a drug transaction, the enhancement might not apply. § 2D1.1 cmt. n. 3 (2007).
These controlling principles prohibit the application of the enhancement in this case because neither a temporal nor a spatial nexus existed. The majority sees a nexus between Manigan’s handguns and his offense because the confidential informant gave Manigan cash for previously received drugs outside 202 Charlesworth Avenue on April 5, 2007, and May 30, 2007. On June 20, 2007, the police executed a search warrant for Manigan’s car and the residence at that address. Although the police found cocaine and drug paraphernalia in Manigan’s car, nothing of that sort was found in the residence. The police did find two handguns belonging to Manigan in a shoe box inside the house.
There is no evidence that Manigan possessed the handguns prior to June 20, 2007, the day of the search, let alone 20 days prior on May 30, 2007, the second (and last) day the informant paid Manigan for drugs outside the residence. There is no suggestion that the informant ever reported that there was a gun in Manigan’s possession during the transactions. And the government does not claim that it had a suspicion prior to the search that Manigan possessed guns. (Neither the search warrant nor the application is in the record.) No evidence supports a finding that the guns Manigan possessed on June 20, 2007 — which he had received from a third party and intended to sell — were possessed on May 30, 2007.
Similarly, even assuming Manigan possessed the guns on May 30, 2007, they were not sufficiently close to the drug transaction to satisfy the spatial nexus requirement. Manigan trafficked in cocaine out of his car. Only his car is present at every face-to-face interaction with the informant, whether outside 202 Charles-worth Avenue or in the CVS Pharmacy parking lot. The police found drugs in Manigan’s car, not in the residence at 202 Charlesworth Avenue. Unlike the house in United States v. Nelson, 202 Charles-worth Avenue was not “a place of manufacturing, storing and distributing [of] crack cocaine.” 6 F.3d at 1056. The residence belonged to Manigan’s mother, and Manigan merely lived there. The only connection between the residence and Manigan’s offense is that Manigan’s car was parked outside. Moreover, despite the majority’s assertion to the contrary, it would amount to a legal fiction to conclude that the guns were “readily available” to, or “constructively possessed” by, Manigan during the drug transactions. Cf Arizona v. Gant, — U.S. -,---, 129 S.Ct. 1710, 1718-19, 173 L.Ed.2d 485 (2009) (rejecting the “fiction ... that the interior of a car is always within the immediate control of an arrestee who has recently been in the car”) (quoting New York v. Belton, 453 U.S. 454, 466, 101 S.Ct. 2860, 69 L.Ed.2d 768 (1981)) (emphasis in original). Had Manigan conducted a transaction outside 202 Charles-worth Avenue and been threatened with violence, keeping guns indoors and hidden in a shoe box would not have provided him any protection. Accordingly, the absence of both a temporal and spatial nexus precludes the enhancement.
The majority reaches a contrary conclusion after several missteps, starting with the standard of review it applies. Citing Rusher, the majority states that “[t]he question of whether a defendant, as a specific offense characteristic, possessed a dangerous weapon ... is a factual determination subject to clear error review.” Ante at 626. Rusher, however, says that the general rule for reviewing a guidelines determination is a “sliding scale” with primarily factual issues reviewed for clear error and primarily legal issues reviewed de novo. 966 F.2d at 880. It concludes that a finding of possession, that is, control or dominion over a weapon, is reviewed for clear error. Id. But a finding of weapon possession is not the final step. A determination of whether the weapon had a role in the offense of conviction is still required. Here, pure possession, along with the fact of the transactions themselves, is not disputed. We are left to review only whether the facts in the presentence report— adopted by the district court — require application of the enhancement. This question is primarily legal and should be reviewed de novo.
The majority’s next error is to place disproportionate weight — in what is a case-specific inquiry — on the general ground that handguns are associated with drugs and that there is a “settled connection between firearms and drug activities.” Ante at 629. Indeed, the majority goes so far as to say that “a sentencing court might reasonably infer, in proper circumstances, that a handgun seized from the residence of a drug trafficker was possessed in connection with his drug activities.” Ante at 630. Of all the cases cited by the majority in support of the association between handguns and drugs, only one cites any empirical evidence, and that evidence supports a strong (and obvious) correlation between handguns and firearm-related crimes, not drug-related crimes. See United States v. Lipford, 203 F.3d 259, 267 n. 7 (4th Cir.2000) (citing Department of Justice, Bureau of Justice Statistics, Selected Findings, Firearms, Crime, and Criminal Justice: Guns Used in Crime (July 1995)). In any event, like the “settled connection between firearms and drug activities,” the connection between handguns and drugs is a reason for the inclusion of the weapons enhancement potential in § 2D1.1 (b)(1), not a reason for relaxing the standard for determining whether a gun was connected to a particular drug offense. More specifically, giving sentencing courts what appears to be a green light to apply the § 2Dl.l(b)(l) enhancement to every drug dealer with a handgun in his residence will practically eliminate any limitation on the scope of the provision.
Finally, the majority errs when it concludes that the hand guns had a role in the offense in this case. Perhaps recognizing that the guns were not close enough to the drug transactions to support a spatial nexus, the majority attempts to link the 202 Charlesworth Avenue residence directly to Manigan’s crimes. Ante at 631-32 (“[T]he Charlesworth Avenue residence was ... central to the common scheme or plan relating to Manigan’s drug activities.”). In support of this link, the majority cites the two occasions when the informant gave Manigan money outside the residence and the fact that “the authorities observed Manigan travel directly from the residence to other locations where he engaged in drug-fronting transactions.” Id. Manigan had to live somewhere, and simply saying that he left his residence to go to other locations to do drug deals does not make his residence central to his drug trafficking. It is only Manigan’s choice to accept payment from the informant on two occasions outside his residence that supports the argument that the residence had a “central” role in Manigan’s drug dealings. This choice is not, however, enough to establish a role for the residence, and therefore for the gun, any more than Manigan’s choice to meet in a CVS Pharmacy parking lot is sufficient to establish a role for CVS. The majority is left only with the transactions’ relative proximity to the handgun which, as I have already noted, is not sufficient because the handguns were not close enough to be readily available for use by Manigan.
Because I conclude that the district court erred when it applied the enhancement, I must reach the question of whether the error was harmless. Manigan admits that if this court reversed the district court’s application of the enhancement, his sentence would remain the same. He nevertheless argues that application of the enhancement prejudices him because it disqualifies him from the Bureau of Prison’s (BOP) drug treatment program and therefore from early release. The government responds that this should not count as prejudice and that, in any event, the injury is hypothetical because there is no guarantee Manigan will be admitted to BOP’s treatment program.
The Tenth Circuit opinion in United States v. Torres persuades me that the district court’s error is not harmless. In Toms the court considered and rejected the exact argument advanced by the government here. 409 F.3d 1000, 1002-03 (10th Cir.2005). The court reasoned that if the enhancement barred the defendant from a shorter sentence, it was clearly prejudicial, and the only question was whether the defendant needed to wait until he was actually denied acceptance to the BOP’s program to challenge the enhancement. Id at 1002. Because the defendant’s challenge at that point would be through habeas corpus, and because a habeas petitioner cannot challenge the execution of a sentence, denying review on direct appeal would deny all review. Id. at 1003. I agree with the Tenth Circuit’s reasoning and conclude that this same prospect in Manigan’s case requires a finding of prejudice.
Because the district court erred and the error is prejudicial, I would vacate Manigan’s sentence and remand for him to be resentenced without the § 2D1.1(b)(1) weapon enhancement.

Passage view · GavelSight