dissenting:
Respectfully, I dissent.
The majority opinion runs counter to Supreme Court precedent, Carachuri-Rosendo v. Holder, 560 U.S. 563, 130 S.Ct. 2577, 177 L.Ed.2d 68 (2010), and effectively guts our Circuit precedent, United States v. Simmons, 649 F.3d 237 (4th Cir. 2011) (en banc). It violates principles' of comity and federalism by directing federal district courts to' ignore the careful sentencing decisions of their state counterparts. And it goes to such lengths all to affirm a twenty-two-year sentence imposed on a fifty-one-year old mentally ill veteran who had previously never served more than ten months in prison, tagging him with the moniker “armed career criminal.” We can do much better than this.
The Armed Career Criminal Act requires a fifteen-year minimum prison term for a defendant convicted of being a felon in possession of a firearm if he had three previous convictions, “for a violent felony or a serious drug offense.” 18 U.S.C. §§ 924(e)(1), 922(g).' A “violent felony” is defined as, among other things, “any crime punishable by imprisonment for a term exceeding one year.” Id. § 924(e)(2)(B). The issue in this case is: When a North Carolina state judge has made a finding that mitigating factors are present and sufficient to outweigh any aggravating factors, and the defendant’s mitigated sentencing range for a North Carolina conviction therefore does not exceed one year, is the conviction for a crime “punishable by imprisonment for a term exceeding one year”?
The answer is and should be no, but the majority answers yes. It holds that federal courts should ignore the defendant’s mitigated range of imprisonment as determined by the North Carolina judge, N.C. Gen.Stat. § 15A-1340.16(b), and instead look to the sentence the state judge could have imposed had she decided, hypothetically, to ignore her own finding that mitigating circumstances justified a sentence in a range that cannot exceed one year, ante, at 13. See N.C. Gen.Stat. § 15A-1340.16(c).
The majority’s answer is fantasy. It inserts an Alice-in-Wonderland analysis into what should be a straightforward question of statutory construction. But most troubling about the majority opinion is that it resurrects a speculative mode of analysis that we — on the foundation of unmistakable Supreme Court authority — discarded in Simmons, 649 F.3d at 237. A bit of history illustrates why the majority opinion is déjá vu.
In United States v. Harp, 406 F.3d 242, 245-46 (4th Cir.2005), we held that a North Carolina conviction for possession with intent to distribute marijuana was, for purposes of the career offender Guideline, U.S. Sentencing Guideline § 4Bl.l(a), a predicate “controlled substance offense” “punishable by imprisonment for a term exceeding one year.” Even though the defendant’s maximum possible punishment was only twelve months, we held that the proper inquiry was whether a hypothetical defendant with the worst possible criminal history “could” have received a sentence of over twelve months for committing the *41same crime. Id. at 246. In other words, the key inquiry was: What was a possible outcome for a hypothetical defendant who committed the same crime?
We rejected this speculative approach in Simmons, holding that “the mere possibility that [the defendant’s] conduct, coupled with facts outside the record of conviction, could have authorized a conviction of crime punishable by more than one year’s imprisonment cannot and does not demonstrate that [the defendant] was actually convicted of such a crime.” 649 F.3d at 244-45 (quotations and citations omitted) (emphasis added). The district court in Simmons, relying on Harp, had applied the Controlled Substances Act sentencing enhancement; it found that the defendant’s prior North Carolina conviction for marijuana possession — for which he faced no possibility of imprisonment — qualified as a felony drug offense “punishable by imprisonment for more than one year,” 21 U.S.C. § 802(44). Id. at 240-41. We vacated the defendant’s sentence and put an end to Harp. Id. at 241.
The primary basis for our decision was Carachuñ-Rosendo, 130 S.Ct. at 2577, a Supreme Court precedent involving the Immigration and Nationality Act. Simmons, 649 F.3d at 244-45. The issue there was whether a Texas conviction for misdemeanor possession of Xanax without a prescription qualified as an “aggravated felony” punishable by imprisonment for more than one year. Carachuñ-Rosendo, 130 S.Ct. at 2580-81. The Fifth Circuit held that it did; the Supreme Court reversed, holding that “the mere possibility, no matter how remote,” that the defendant could have received a two-year sentence had he been charged in federal court for the same criminal conduct did not transform the Texas conviction into an aggravated felony. Id. at 2583.
Carachuñ-Rosendo involved a different statutory scheme and an analysis with multiple layers of speculation, but we concluded in Simmons that the relative simplicity of the Controlled Substances 'Act analysis did “not render the Carachuñ-Rosendo holding inapplicable.” Simmons, 649 F.3d at 248. Instead, we construed Carachuñ-Rosendo as a prohibition on considering “facts not at issue in the crime of conviction” in determining whether a conviction qualifies as a predicate felony. Id. at 248 (quotations and citation omitted).
Which brings me to this case: It is difficult to see how the majority’s hypothetical modé of analysis does not run square into the teeth of what Carachuñ-Rosendo and Simmons seek to prohibit. As the majority opinion makes clear, the North Carolina state judge acted consistent with the provisions of the North Carolina Structured Sentencing Act: she made a finding that Kerr’s mitigating factors outweighed those in aggravation, N.C. Gen.Stat.. § 15A-1340.16(b); Kerr had to prove those findings by a preponderance of the evidence, id. § 15A-1340.16(a); her mitigation finding was — and had to be — in writing and part of the record, id. § 15A-1340.16(c); and she used the mitigation finding to arrive at a sentencing range of eight to ten months, as contemplated by North Carolina law.
Despite all of this, the majority concludes that Kerr’s conviction was for a “crime punishable by imprisonment for a term exceeding one year” because the North Carolina state judge “may,” ante, at 38, have nonetheless imprisoned Kerr for up to fourteen months. In other words, Harp redux: “[T]he mere possibility, no matter how remote,” Carachuñ-Rosendo, 130 S.Ct. at 2583, coupled with an analysis that ignores a fact in the record, is sufficient to conclude that Kerr could have *42received a sentence of over one year. In doing so, it transforms Kerr into an armed career criminal deserving of a minimum of fifteen years of “3 hots and a cot,” People v. Shulman, 6 N.Y.3d 1, 809 N.Y.S.2d 485, 843 N.E.2d 125, 134 (2005), on the tab of United States taxpayers.
The majority responds to all this with a paradox: first, it says that the issue in this case was not before us in Simmons, ante, at 38 (“Kerr’s appeal presents an issue that neither Simmons nor its progeny expressly addressed”); second, it says that a footnote in Simmons resolves this case, ante, at 39 (“Our holding remains faithful to our directive in Simmons requiring that sentencing courts examine ‘three pieces of evidence: the offense class, the offender’s prior record level, and the applicability of the aggravated sentencing range,’ 649 F.3d at 247 n. 9.”).
Let’s unscramble that egg: The majority’s position is that although Simmons does not mandate a particular outcome here, its “directive” is nonetheless consistent with the majority’s decision. Wrong on both counts. But, even if Simmons were deemed in some sense “consistent” with the majority’s decision, the decision I reach is equally “consistent” and achieves the more' sensible result.
The majority’s first argument, that Simmons does not mandate a particular outcome here, is incorrect because it is based on a misunderstanding of the issue in this case. The majority frames the issue as: “Must a district court, in determining whether a defendant has the requisite predicate felonies for sentencing as an armed career criminal, consider the fact that the defendant received a mitigated sentence of less than one year in prison under North Carolina law for those felonies?” Ante, at 38 (emphasis added). That framing subtly misconstrues the issue in this case: it is not the “fact” of the defendant’s actual sentence that is relevant or the focus of this analysis, but the defendant’s actual sentencing range. The North Carolina state court judge made an express finding — a real-world fact — that mitigating factors outweighed those in aggravation; indeed, she made that finding because it is a predicate to sentencing in the mitigated range. N.C. Gen.Stat. § 15A-1340.16(b).
The majority’s framing of the issue seems to be based on a peculiar understanding of the North Carolina Structured Sentencing Act: it thinks that a North Carolina judge is always in the presumptive range because she has the discretion to sentence in that range even after she’s made the mitigation finding. State v. Bivens, 155 N.C.App. 645, 573 S.E.2d 259, 262 (2002).
Although the majority is correct that a North Carolina judge retains the discretion to sentence in the presumptive range even when the judge finds several mitigating factors, the existence of that possibility alone is an insufficient basis for ignoring the range as determined according to the North Carolina Structured Sentencing Act. We said as much in Simmons: North Carolina judges always have the discretion to use the presumptive range even after the state has proven the existence of aggravating factors, N.C. Gen.Stat. § 15A-1340.16(a), but we stated that federal district courts should pay close attention to the actual presence and proof of those aggravating factors, rather than some hypothetical analysis, because it is a more coherent and just analysis that better informs whether a defendant is deserving of a major imprisonment term. Simmons, 649 F.3d at 244-45, 247 n. 9. In the face of this, the majority opinion is silent on why the judge’s retained discretion renders her decision to affirmatively make the mitigation finding and to impose a sentence *43based on that finding a legal nullity. If anything, we should honor the judge’s mitigation finding because, as the counterpart to the aggravation finding, it is a predicate to deviating from the presumptive range (in this case, to the mitigated range). N.C. Gen.Stat. § 15A-1340.16(b).
In Carachuri-Rosendo, the Court acknowledged that Texas law afforded prosecutors broad discretion in determining whether to seek a conviction whose punishment would satisfy an element of federal law (there, in the immigration context); nevertheless, that fact did not persuade the Court to base its analysis of federal law on the existence of such discretion. See 130 S.Ct. at 2588. It is unclear to me why, if the existence of state prosecutorial discretion is not an acceptable basis for an interpretation of federal immigration law that leads to draconian outcomes, the existence of a similar judicial discretion should permit an interpretation of federal sentencing law that, correspondingly, leads to draconian outcomes, as the majority concludes here. Paradox, indeed.
The majority’s approach builds an unwarranted asymmetry into our treatment of North Carolina convictions, where we pay special attention, if not conclusive attention, to a finding of aggravation, but purposely ignore a judge’s finding of mitigation. To build our jurisprudence on this asymmetry expresses the view that federal sentencing courts do not really care about a defendant’s criminal background — or at least, they only care when it offers up the opportunity to send him to prison for a longer period of time than common sense and simple justice demand.
The majority’s second argument is that its holding “remains faithful to our directive in Simmons requiring that sentencing courts examine ‘three pieces of evidence: the offense class, the offender’s prior record level, and the applicability of the aggravated sentencing range,’ 649 F.3d at 247 n. 9, ” ante, at 39. This protestation is unavailing. Indeed, what a member of today’s majority said in railing against the en banc majority in Simmons applies with genuine force here: “The majority opinion puts great weight on footnote [9] of [Simmons ] to undergird its position; but that weight is more than the footnote can bear.” Simmons, 649 F.3d at 255 (Agee, J., dissenting). The core of Simmons is (obviously) its holding, informed by Supreme Court precedent, of how the “mere possibility” of a defendant’s conduct, “coupled with facts outside the record of conviction,” was insufficient for concluding that a conviction was punishable by more than one year’s imprisonment. 649 F.3d at 244-45 (quotations and citations omitted) (emphasis added). The majority today makes that line of inquiry sufficient: Kerr’s criminal conduct, coupled with the “mere possibility” that a judge could have exercised her discretion to ignore her own finding of mitigation (a real-world fact in the record), transforms Kerr’s conviction from one for which he could only receive eight to ten months into a “crime punishable by imprisonment for a term exceeding one year.” Id. § 924(e)(2)(B).1 Because such an analysis is contingent on too many *44assumptions, it is quintessential^ “hypothetical” and therefore foreclosed by Simmons and Carachuri-Rosendo.
The majority’s constricted reading of Simmons stems from its refusal to acknowledge the principles that animated the core holding of Simmons: “well-established federalism principles” that prohibit federal courts from “rejectfing] North Carolina’s judgment as to the seriousness of a North Carolina crime, prosecuted in a North Carolina court and adjudicated by a North Carolina judge,” Simmons, 649 F.3d at 249; deference to the North Carolina legislature’s “carefully crafted sentencing scheme,” id. at 249-50; and an unwillingness to permit peculiarly harsh, “counter-intuitive and unorthodox” results in federal sentencing, id. (quoting Carachuri-Rosendo, 130 S.Ct. at 2585). The majority’s decision is violative of all three principles: it contravenes principles of federal respect for state judges and their assessment of offenders, the convictions those offenders incur, and the just sentences they deserve; it re-writes for purposes of federal law a carefully-crafted state statute; and it results in preposterous sentences.
It is the last point that resonates most. The majority opinion is profoundly wrong on the law. But what is most concerning is how completely untethered its analysis is from the task before federal district judges in these cases: to decide who should be treated as a repeat offender responsible for a disproportionately large percentage of violent crimes, i.e., an armed career criminal. See United States v. Hawkins, 811 F.2d 210, 216-17 (3d Cir. 1987) (discussing the legislative history of the precursor to the current form of the Armed Career Criminal Act). For example, suppose a petty criminal receives a mitigation finding because he proved to the North Carolina state judge by a preponderance of the evidence that he was under coercion when he committed the crime, N.C. Gen.Stat. § 15A-1340.16(e)(1), or was of limited mental capacity, id. § 15A-1340.16(e)(4), or was a minor, id. § 15A-1340.16(e)(6). Suppose also that the North Carolina judge, after determining that those mitigating factors outweigh those in aggravation, concludes that the defendant’s range is eight to ten months. The majority’s decision today tells federal district courts to repudiate that finding and treat that defendant as if he were the mastermind of the crime, a criminal genius, or an adult, and impose a fifteen-year minimum sentence (or something closer to the twenty-two year sentence imposed here). These results were not what Congress intended. Overreaching by federal prosecutors, who are sometimes anxious to wield their “armed career criminal” lancets willy-nilly, see United States v. Foster, 662 F.3d 291, 301 (4th Cir.2011) (Wynn, J., dissenting); United States v. Foster, 674 F.3d 391, 398 (4th Cir.2012) (Motz, J., dissenting from the den. of reh’g); id. at 403-04 (Davis, J., dissenting from the den. of reh’g), should not be rewarded in this manner.
And even if there were a colorable argument that Congress intended these types of results, the rule of lenity requires much better evidence before adoption of the majority’s construction of the Armed Career Criminal Act. At best, it is not at all clear what Congress wants federal courts to do with these peculiar wrinkles in North Carolina’s Structured Sentencing Act, and we should not render a construction that increases the penalty on a defendant when that construction is “based on no more than a guess as to what Congress intend*45ed.” Ladner v. United States, 358 U.S. 169, 178, 79 S.Ct. 209, 3 L.Ed.2d 199 (1958). The Armed Career Criminal Act has proved enormously complex in application; trying to decide how it should interact with equally complex state sentencing regimes is akin to proclaiming mastery at reading tea leaves. See United States v. Rodriquez, 553 U.S. 377, 405, 128 S.Ct. 1783, 170 L.Ed.2d 719 (2008) (Souter, J., dissenting) (arguing for application of the rule of lenity to the Armed Career Criminal Act).2 Rather than boldly continue down this dreary path, we should exercise judicial restraint by holding that the statute’s ambiguity coupled with the rule of lenity tips the scales in favor of the defendant, thereby allowing Congress the opportunity to provide us with additional guidance, which is so sorely needed.
Whether a prior state judge has made a written factual finding that the circumstances in a particular case are atypical and warrant application of a mitigated sentencing range that by definition does not exceed one year is a highly relevant data point in assessing whether an individual is an armed career criminal. The majority does not think so; it is wrong. But the true error it commits is justifying its decision by sticking its head in the sand of legal artifice.
I am willing to believe, and to act on the belief, that “no one — not even the prosecutors themselves — thinks [a twenty-two-year sentence on a fifty-one-year-old mentally ill veteran is] appropriate” under the circumstances of this case. Cf. United States v. Kupa, — F.Supp.2d-,-, 2013 WL 5550419, at *1 (E.D.N.Y. Oct. 9, 2013). The law affords us an opportunity to decide this case on that belief. I deeply regret the institutional ennui that precludes our doing so.
. The majority is determined to hang its hat on United States v. Edmonds, 679 F.3d 169, 176-77 (4th Cir.2012), for the proposition that "the actual sentence” a defendant receives is not the relevant point of analysis, ante, at 37-38. Edmonds is irrelevant because it did not involve a defendant who had received a mitigation finding. Edmonds, 679 F.3d at 176-77. And I am not arguing "myopically,” ante, at 38, for an assessment of the defendant’s actual sentence; I am arguing for an assessment of the defendant's actual sentencing range. The majority says this is a "distinction without a difference,” ante, at 39 n. 7. Of course, if the majority is comfortable contravening the finding of a state court judge *44in a state court conviction sustained pursuant to a state statute, then I suppose the distinction is probably one without a difference.
. As Justice Souter explained:
The rule [of lenity] is grounded in "the instinctive distaste against men languishing in prison unless the lawmaker has clearly said they should,” and we have used it to resolve questions both about metes and bounds of criminal conduct and about the severity of sentencing. This policy of lenity means that the Court will not interpret a federal criminal statute so as to increase the penalty that it places on an individual when such an interpretation can be based on no more than a guess as to what Congress intended.
United States v. Rodriquez, 553 U.S. 377, 405-06, 128 S.Ct. 1783, 170 L.Ed.2d 719 (2008) (Souter, J., dissenting) (citations omitted). See also United States v. Granderson, 511 U.S. 39, 54, 114 S.Ct. 1259, 127 L.Ed.2d 611 (1994) ("[W]here text, structure, and history fail to establish that the Government's position is unambiguously correct ... we apply the rule of lenity.”).