No. 23-6028

Appeal from the United States District Court for the Western District of Oklahoma (D.C. No. 5:22-CR-00328-PRW-1)

Steven W. Creager, Assistant United States Attorney (Robert J. Troester, United States Attorney, with him on the briefs), Oklahoma City, Oklahoma, for Plaintiff-Appellant.

Laura K. Deskin, Assistant Federal Public Defender, Oklahoma City, Oklahoma, for Defendant-Appellee.

Before ROSSMAN , KELLY, and MURPHY, Circuit Judges.

ROSSMAN , Circuit Judge.

Page 2Jared Michael Harrison was charged with violating 18 U.S.C. § 922(g)(3), which prohibits firearm possession by “any person . . . who is an unlawful user of or addicted to any controlled substance.' He moved to dismiss the indictment under New York State Rifle & Pistol Ass’n v. Bruen , 597 U.S. 1, 37 (2022), contending the charging statute violates the Second Amendment as applied to non-intoxicated marijuana users. The district court granted the motion. The government now appeals.

The district court comprehensively analyzed the constitutional question. But the district court ruled before the Supreme Court decided United States v. Rahimi , 602 U.S. 680 (2024), which issued while this appeal was pending and clarified our constitutional inquiry. Rahimi instructs “the appropriate analysis [in Second Amendment cases] involves considering whether the challenged regulation is consistent with the principles that underpin our regulatory tradition.' Id. at 692. With the benefit of Rahimi , we cannot fully endorse the district court’s understanding of the historical record.

According to the district court, our historical tradition of firearm regulation is limited to disarming those who have acted dangerously in the past. But we conclude, contrary to the district court, disarming those believed to pose a risk of future danger is consistent with a “principle[] that underpin[s] our regulatory tradition.” Id. at 692. Still, we cannot yet decide Page 3the ultimate constitutional question. To determine whether § 922(g)(3) as applied here is 'consistent with' the principle that the government has correctly identified, the government must show non-intoxicated marijuana users pose a risk of future danger. This inquiry, which may involve fact finding, is best suited for the district court. Exercising jurisdiction under 18 U.S.C. § 3731, we therefore reverse and remand for further proceedings consistent with this opinion.

On May 20, 2022, a law enforcement officer from the Lawton, Oklahoma Police Department stopped Mr. Harrison for running a red light. When Mr. Harrison rolled down his window, the officer smelled marijuana. Mr. Harrison told the officer that he worked at a medical marijuana dispensary but did not have a state-issued medical marijuana card. The officer then instructed Mr. Harrison to get out of his car, and he complied.² The officer did not conduct a field sobriety test or blood draw. Another officer then arrived, and the two searched Mr. Harrison’s car. They found:

Page 4a loaded revolver on the driver’s side floorboard; two prescription bottles in the driver’s side door, one empty and one containing partially smoked marijuana cigarettes; and a backpack in the passenger seat. The backpack contained marijuana, THC gummies, two THC vape cartridges, and a prerolled marijuana cigarette and marijuana stems in a tray.

App. at 88. Mr. Harrison was arrested, and Oklahoma charged him with possession of marijuana, possession of paraphernalia, and failure to obey a traffic signal.

On August 17, 2022, a federal grand jury returned an indictment charging Mr. Harrison with violating § 922(g)(3). Section 922(g)(3) provides:

It shall be unlawful for any person . . . who is an unlawful user of or addicted to any controlled substance . . . to ship or transport in interstate or foreign commerce, or possess in or affecting commerce, any firearm or ammunition; or to receive any firearm or ammunition which has been shipped or transported in interstate or foreign commerce.

The indictment alleged Mr. Harrison “knowingly possessed a firearm . . . which was in or affecting interstate commerce,” “with knowledge that he was an unlawful user of marijuana, a controlled substance.” App. at 8. The indictment stated no additional information about Mr. Harrison’s marijuana use or gun possession.³

Page 5Mr. Harrison moved to dismiss the indictment under Federal Rule of Criminal Procedure 12(b)(3)(B). He argued “Section 922(g)(3) unconstitutionally infringes on an individual’s right to bear arms under the Second Amendment.'⁴ App. at 25. Applying the framework of Bruen , Mr. Harrison maintained the Second Amendment's text covered his conduct, and 'the Government cannot meet its burden to show that § 922(g)(3)'s restrictions are ‘consistent with the Nation’s historical tradition of firearm

App. at 8–9.

Page 6regulation.'' App. at 27 (quoting Bruen , 597 U.S. at 24). “[E]ven if this Court finds that § 922(g)(3) does not violate the Second Amendment on its face,” Mr. Harrison contended, 'the statute should not apply to a person like [him], who was merely found with a controlled substance.” App. at 29. The government responded Mr. Harrison fell outside the Second Amendment’s ambit, which it claimed includes only “law-abiding citizens.” App. at 52–55. The government also argued legislatures have long disarmed those believed to “pose a danger to public safety if armed”—including felons, the mentally ill, intoxicated people, Catholics, and loyalists. App. at 62; see App. at 183– 87, 190, 205. And the government contended § 922(g)(3) is consistent with this tradition as applied to Mr. Harrison. The district court heard argument on the motion.

On February 2, 2023, in a thorough 54-page order, the district court granted Mr. Harrison’s motion to dismiss. The district court held the Second Amendment’s text covered Mr. Harrison’s conduct because he is among “the People” to whom the Amendment refers. The court further held “[a]pplying § 922(g)(3) to Harrison is inconsistent with the Nation’s historical tradition of firearm regulation.” App. at 96. The government timely appealed.

While the appeal was pending, the Supreme Court decided Rahimi , which held 18 U.S.C. § 922(g)(8) does not facially violate the Second Amendment and clarified the methodology for adjudicating Second Page 7Amendment challenges. See 602 U.S. at 700, 692; Rocky Mountain Gun Owners v. Polis , 121 F.4th 96, 113 (10th Cir. 2024) (explaining the “twopart burden-shifting framework first established in Bruen ' was 'later clarified in Rahimi '). We then ordered the parties to file supplemental briefs addressing Rahimi ’s impact on Mr. Harrison’s case.

Before addressing the arguments before us, we must explain the scope of Mr. Harrison’s Second Amendment challenge. This is a threshold issue because it determines what Mr. Harrison must 'establish' in order to “succe[ed].” Rahimi , 602 U.S. at 693 (quoting United States v. Salerno , 481 U.S. 739, 745 (1987)). The parties dispute whether Mr. Harrison brought a facial challenge, an as-applied challenge, or some mixture of the two. Compare Ans. Br. at 12 (Mr. Harrison contending § 922(g)(3) “violates the Second Amendment . . . as applied to Mr. Harrison as an alleged marijuana user”), with Gov’t Supp. Resp. Br. at 3 (the government contending “despite the label Mr. Harrison has used, he has not raised an as-applied challenge”), and Oral Argument 1:25-3:00 (government counsel suggesting this case “presents a partial facial challenge, a partial as-applied challenge” (citing United States v. Sup. Ct. of N.M. , 839 F.3d 888 (10th Cir. 2016)).

“A facial challenge is a head-on attack [of a] legislative judgment, an assertion that the challenged statute violates the Constitution in all, or Page 8virtually all, of its applications.” United States v. Carel , 668 F.3d 1211, 1217 (10th Cir. 2011) (alteration in original) (quoting United States v. Pruitt , 502 F.3d 1154, 1171 (10th Cir. 2007) (McConnell, J., concurring), vacated , 552 U.S. 1306 (2008)). To succeed on a facial challenge, Mr. Harrison would need to “establish that no set of circumstances exists under which the Act would be valid.” Rahimi , 602 U.S. at 693 (quoting Salerno , 481 U.S. at 745); see also Doe v. City of Albuquerque , 667 F.3d 1111, 1122-27 (10th Cir. 2012) (discussing different ways of understanding facial challenges). '[A]n asapplied challenge concedes that the statute may be constitutional in many of its applications, but contends that it is not so under the particular circumstances of the case.' Carel , 668 F.3d at 1217 (quoting Pruitt , 502 F.3d at 1171 (McConnell, J., concurring)). These different standards implicate § 922(g)(3) because the statute applies to all 'unlawful user[s] of . . . controlled substance[s],' not just marijuana users. And even among individuals who use marijuana, § 922(g)(3) applies to both those who are intoxicated while possessing a firearm and those who are not.

In the district court, Mr. Harrison seemed to challenge § 922(g)(3) both as applied and on its face. See App. at 29 (“Even if this Court finds that § 922(g)(3) does not violate the Second Amendment on its face, the statute should not apply to a person like Mr. Harrison . . . .”). The district court adjudicated only Mr. Harrison's as-applied challenge. See App. at 99Page 9 (discussing '§ 922(g)(3)'s application to Harrison'); App. at 125 ('The Second Amendment . . . requires that § 922(g)(3), as applied to Harrison, be consistent with history and tradition[] . . . .”); App. at 126 (“But the United States' own conception of the historical tradition demonstrates why § 922(g)(3) as applied to Harrison is not analogous to these traditions.”); App. at 210 (district court stating at motion hearing that 'this is an asapplied challenge”). And Mr. Harrison has not objected to the district court’s understanding of his argument or reprised any facial challenge on appeal. See Harrison Supp. Resp. Br. at 1 ('A facial ruling is not before this court . . . .'); see also Ans. Br. at 12 (arguing § 922(g)(3) “violates the Second Amendment . . . as applied to Mr. Harrison'); Harrison Supp. Br. at 1 (“Application of Rahimi demonstrates 18 U.S.C. § 922(g)(3) as applied to Mr. Harrison is unconstitutional.'). Accordingly, we conclude no facial challenge is before us, and we consider only Mr. Harrison's as-applied challenge. See Perry v. Woodward , 199 F.3d 1126, 1141 n.13 (10th Cir. 1999) (“This court . . . will not craft a party’s arguments for him.”).

Mr. Harrison's as-applied challenge concerns the application of § 922(g)(3) to non-intoxicated marijuana users. To be sure, Mr. Harrison’s indictment was silent about whether he was intoxicated when the offense conduct occurred. But in the district court, Mr. Harrison suggested he was not intoxicated during his charged possession, the government did not Page 10challenge that understanding, and the district court appeared to accept it. See App. at 29 (Mr. Harrison stating he 'was merely found with a controlled substance in his vehicle'); App. at 61-64 (the government describing historical laws that disarmed intoxicated people without suggesting Mr. Harrison was intoxicated during his gun possession); App. at 102 (the district court distinguishing § 922(g)(3) from historical laws disarming intoxicated people, because § 922(g)(3) applies “regardless of whether the person is actually intoxicated or under the influence of a controlled substance”). Mr. Harrison’s position on appeal proceeds from this factual premise, and we see no reason to reject it. See Ans. Br. at 31 n.39 (“Mr. Harrison is not alleged to have been intoxicated at the time he possessed a firearm.”).

For the first time in its supplemental briefs on appeal, the government raises two new arguments concerning the scope of Mr. Harrison’s challenge. These arguments come too late.

First , the government suggests Mr. Harrison cannot litigate an asapplied challenge at this procedural stage. According to the government, “pretrial as-applied challenges to the constitutionality of a statute are rarely permissible.' Gov't Supp. Br. at 5 n.2 (citing United States v. Pope , 613 F.3d 1255, 1260 (10th Cir. 2010)). In Pope , we explained courts typically should not consider pre-trial “motions to dismiss that require resort to facts outside the Page 11indictment and bearing on the general issue” of guilt. 613 F.3d at 1260. But a court may consider facts outside the indictment when “‘[1] the operative facts are undisputed and [2] the government fails to object to the district court’s consideration of those undisputed facts,' and [3] the district court can determine from them that, ‘ as a matter of law , the government is incapable of proving its case beyond a reasonable doubt.'' Id. (alterations in original) (quoting United States v. Hall , 20 F.3d 1084, 1088 (10th Cir. 1994)).

Also, when the government litigates a case in a way that implicitly endorses the district court’s resort to facts outside the indictment, and objects only after the case has proceeded on that basis, we have recognized 'the government’s procedural objection is untimely.” United States v. Brown , 925 F.2d 1301, 1304 (10th Cir. 1991); Hall , 20 F.3d at 1087-88 (recognizing the timeliness holding in Brown , and allowing resort to facts outside the indictment when “the government did not dispute the facts surrounding the . . . charge”); see also Pope , 613 F.3d at 1262 (affirming denial of a motion to dismiss when, prior to appeal, “the government did contest the district court's ability to resolve [the] motion on the basis of facts outside [the] indictment”); see also Lyons v. Jefferson Bank & Tr. , 994 F.2d 716, 721 (10th Cir. 1993) (explaining a party cannot “lose in the district court on one theory of the case, and then prevail on appeal on a different theory').

Page 12That is the situation here. In the district court, the government never took issue with Mr. Harrison’s ability to bring—or the district court’s ability to resolve—an as-applied challenge. See, e.g. , App. at 56–65 (government failing to raise this issue in its district court brief). Even on appeal, the government initially accepted that Mr. Harrison was raising an as-applied challenge to § 922(g)(3). See Op. Br. at 31 (government arguing certain evidence “has no impact in Mr. Harrison's as-applied challenge'). We thus have no trouble concluding “the government’s procedural objection is untimely,” meaning Mr. Harrison's as-applied challenge is before us. Brown , 925 F.2d at 1304.

Second , the government raises a new factual objection—that the record before the district court 'permit[ted] an inference that Mr. Harrison was intoxicated.” Gov’t Supp. Br. at 3; see App. at 55–65 (the government failing to suggest this was the case in its district court brief). Again, we cannot endorse the government’s effort to raise this new argument in its supplemental briefing on appeal. See id. ; Lyons , 994 F.2d at 721; State Farm Fire & Cas. Co. v. Mhoon , 31 F.3d 979, 984 n.7 (10th Cir. 1994) (“[A]ppellant failed to raise this issue in his opening brief and, hence, has waived the point.”).

We therefore conclude this case presents a challenge only to § 922(g)(3)’s application to non-intoxicated marijuana users.

Page 13We now explain the legal standards applicable to Second Amendment challenges.

We “review challenges to the constitutionality of a statute de novo.” United States v. Wilks , 58 F.3d 1518, 1519 (10th Cir. 1995). “Statutes are presumed constitutional.” United States v. Dorris , 236 F.3d 582, 584 (10th Cir. 2000). 'To ascertain the constitutionality of a law burdening an individual's exercise of the Second Amendment, we apply a two-part burden-shifting framework first established in Bruen and later clarified in Rahimi .” Rocky Mountain Gun Owners , 121 F.4th at 113.

In Bruen , the Supreme Court explained the “standard for applying the Second Amendment is as follows:

[1] When the Second Amendment's plain text covers an individual's conduct, the Constitution presumptively protects that conduct.

[2] The government must then justify its regulation by demonstrating that it is consistent with the Nation’s historical tradition of firearm regulation.

597 U.S. at 24. We refer to the two parts of this standard as “step one” and 'step two.' Rocky Mountain Gun Owners , 121 F.4th at 113-28; see also Rahimi , 602 U.S. at 744 (Jackson, J., concurring) (referring to Bruen ’s “twostep evaluation”).

Page 14“At step one, the plaintiff is tasked with establishing that the Second Amendment's explicit text, 'as informed by history,' encompasses the conduct they seek to engage in.' Rocky Mountain Gun Owners , 121 F.4th at 113 (quoting Bruen , 597 U.S. at 19). We must 'ask (1) whether the challenger is part of 'the people' whom the Second Amendment protects, (2) whether the item at issue is an ‘arm’ that is ‘in common use’ today for self-defense, and (3) whether the proposed course of conduct falls within the Second Amendment.” Id. at 114 (quotation omitted) (citing Bruen , 597 U.S. at 31–32). If the plaintiff succeeds, the burden shifts to the government.

At step two, to show a challenged regulation is consistent with our historical tradition, the government must engage in “analogical reasoning.” Bruen , 597 U.S. at 28, 30. That is, Bruen requires the government to point to historical regulations that are analogous to the challenged regulation. See id. at 28–29. Whether a historical regulation is analogous to a modern one for purposes of Bruen step two-or whether they are 'relevantly similar”—depends on “ how and why the regulations burden' the right to bear arms. Id. at 29 (emphasis added). We thus ask “whether modern and historical regulations impose a comparable burden on the right of armed self-defense and whether that burden is comparably justified .” Id. (emphasis added). Bruen held the government must 'identify a wellestablished and representative historical analogue , not a historical twin . So Page 15even if a modern-day regulation is not a dead ringer for historical precursors, it still may be analogous enough to pass constitutional muster.” Id. at 30.

To evaluate which way the history cuts, Bruen provided another piece of guidance. “[W]hen a challenged regulation addresses a general societal problem that has persisted since the 18th century,” the Court explained, 'the lack of a distinctly similar historical regulation addressing that problem is relevant evidence that the challenged regulation is inconsistent with the Second Amendment.' Id. at 26 (emphasis added). But 'cases implicating unprecedented societal concerns . . . may require a more nuanced approach. The regulatory challenges posed by firearms today are not always the same as those that preoccupied the Founders in 1791 or the Reconstruction generation in 1868.” Id . at 27.

In Rahimi , the Supreme Court reinforced and developed Bruen ’s methodology for analogical reasoning. When considering historical analogues, Rahimi instructed, “the appropriate analysis involves considering whether the challenged regulation is consistent with the principles that underpin our regulatory tradition.' 602 U.S. at 692 (emphasis added); Rocky Mountain Gun Owners , 121 F.4th at 113 (emphasizing Rahimi requires us to find ' principles ' in our historical tradition (quoting Rahimi , 602 U.S. at 692)). Justice Sotomayor’s Page 16concurrence in Rahimi further elaborated on how the analogical reasoning must proceed. “The Court’s opinion also clarifies an important methodological point that bears repeating: Rather than asking whether a present-day gun regulation has a precise historical analogue, courts applying Bruen should 'conside[r] whether the challenged regulation is consistent with the principles that underpin our regulatory tradition.'' Id. at 703-04 (Sotomayor, J., concurring) (quoting id. at 692 (majority opinion)). The point to remember, therefore, is '[h]istorical regulations reveal a principle, not a mold.” Id. at 740 (Barrett, J., concurring).

Finally, we must consider what history is relevant to the analogical inquiry. Bruen and Rahimi provide guidance on this front. Evidence from the founding era-or the years surrounding 1791, when the Second Amendment was first ratified-is most probative.⁵ See Bruen , 597 U.S. atPage 17 34. And this makes sense, because we know our objective is to determine what 'scope' the Second Amendment was 'understood to have when the people adopted [ it ].” Id. (quoting District of Columbia v. Heller , 554 U.S. 570, 634–35 (2008)).

Evidence that predates the Second Amendment also bears on the analogical inquiry. In both Bruen and its precursor Heller , the Supreme Court emphasized we must consider 'the historical background of the Second Amendment . . . because it has always been widely understood that the Second Amendment . . . codified a pre-existing right.' Heller , 554 U.S. at 592; Bruen , 597 U.S. at 20. 'The Amendment 'was not intended to lay down a novel principle but rather codified a right inherited from our English ancestors.’” Bruen , 597 U.S. at 20 (quoting Heller , 554 U.S. at 599); see id. (favorably citing Heller ’s reliance on “English history dating from the late 1600s, along with American colonial views leading up to the founding”). Of Page 18course, “[h]istorical evidence that long predates [1791] may not illuminate the scope of the right if linguistic or legal conventions changed in the intervening years.” Id. at 34; see id. at 35 ('English common-law practices and understandings at any given time in history cannot be indiscriminately attributed to the Framers of our own Constitution.”).

Postenactment history demands similar care. See id. ('[W]e must also guard against giving postenactment history more weight than it can rightly bear.”). On one hand, “‘a regular course of practice’ can ‘liquidate & settle the meaning of' disputed or indeterminate 'terms & phrases' in the Constitution,” and “where a governmental practice has been open, widespread, and unchallenged since the early days of the Republic, the practice should guide our interpretation of an ambiguous constitutional provision.' Id. at 35-36 (first quoting Chiafalo v. Washington , 591 U.S. 578, 593 (2020); and then quoting NLRB v. Noel Canning , 573 U.S. 513, 572 (2014) (Scalia, J., concurring in the judgment)); see also id . at 35 (“It is true that in Heller we reiterated that evidence of 'how the Second Amendment was interpreted from immediately after its ratification through the end of the 19th century’ represented a ‘critical tool of constitutional interpretation.'' (quoting Heller , 554 U.S. at 605)). On the other hand, history from many years after enactment provides mostly 'secondary'Page 19 evidence of original meaning. Id. at 37. And 'to the extent later history contradicts what the text says, the text controls.'⁶ Id. at 36.

Keeping in mind the background law, we now describe the district court’s order in detail. The district court concluded § 922(g)(3) violates the Second Amendment as applied to non-intoxicated marijuana users and granted Mr. Harrison’s motion to dismiss his indictment.

To start, the court correctly began with the Bruen framework. At Bruen step one, the district court held “[t]he Second Amendment’s plain text covers Harrison’s conduct.” App. at 93. Mr. Harrison’s conduct constitutedPage 20 “‘keeping’ a firearm,” the court reasoned, and he is among the “people” to whom the Second Amendment refers. App. at 93–94. The court supported this conclusion with language in Heller and Bruen that suggests the word 'people' means 'all Americans.' App. at 94 (quoting Heller, 554 U.S. at 580– 81; Bruen , 597 U.S. at 70). And, as the court also recognized, nobody disputes Mr. “Harrison is an American citizen.” App. at 94.

The district court next turned to Bruen step two, which requires the government to “justify its regulation by demonstrating that it is consistent with the Nation's historical tradition of firearm regulation.' Bruen , 597 U.S. at 24; App. at 96. Here, the district court first concluded § 922(g)(3) addresses a societal problem that has persisted since the founding, and “the United States has not identified a single historical law that is ‘distinctly similar’ to § 922(g)(3).” App. at 98.

The court then addressed the government’s argument that § 922(g)(3) fits within a historical tradition of disarming those believed to be dangerous. The court acknowledged, as the government had argued, that legislative 'limitations on the right to armed self-defense 'were tied to dangerousness.’” App. at 119 (quoting Folajtar v. Att’y Gen. of the U.S. , 980 F.3d 897, 913 (3d Cir. 2020) (Bibas, J., dissenting), abrogated by Bruen , 597 U.S. 1, as recognized in Range v. Att’y Gen. United States , 124 F.4th 218, 225 (3d Cir. 2024) (en banc)). But in the district court’s view, the historical record Page 21showed only ''that the legislature may disarm those who have demonstrated a proclivity for violence' through past violent, forceful, or threatening conduct.' App. at 117-18 (emphasis added) (quoting Kanter v. Barr , 919 F.3d 437, 454 (7th Cir. 2019) (Barrett, J., dissenting)); see also App. at 118 (“[T]he Constitution permitted the dispossession of persons who demonstrated that they would present a danger to the public if armed.' (emphasis added) (quoting Binderup v. Att’y Gen. U.S. of Am. , 836 F.3d 336, 369 (3d Cir. 2016) (en banc) (Hardiman, J., concurring in part), abrogated by Bruen , 597 U.S. 1, as recognized in Range , 124 F.4th at 225 (en banc))).⁷ The district court saw nothing in the history that permitted legislatures to disarm people based on 'assessments of future risk.' App. at 125. It reasoned 'stripping a person's fundamental rights based on projected crimes untethered from past dangerous actions is a risky game indeed.' App. at 124–25 (quoting Folajtar , 980 F.3d at 923 (Bibas, J., dissenting)).

Page 22As relevant to this appeal, the court specifically addressed the government’s proposed analogies to historical laws disarming the mentally ill, Catholics, and loyalists.⁸ The court reasoned 'history and tradition would limit disarmament to dangerous lunatics,' not all 'lunatics.' App. at 126. The court also concluded laws disarming the mentally ill could not be analogous to laws disarming non-intoxicated marijuana users. The government's argument-that historical laws disarming the mentally ill show legislatures can disarm those with “difficulty exercising selfcontrol”—would “appear[] to have no limit,” as it would also apply to those with “autism, attention deficit disorder, and nicotine dependence.” App. at 126 (quoting App. at 64). Concerning laws disarming Catholics and loyalists, the court held the Second Amendment “did not incorporate” such laws. App. at 131. The court likewise found laws disarming Catholics and loyalists “were justified on the fear that the covered groups were likely to wage active war,” not that they generally posed a risk of danger. App. at 136–37.

Having recognized a tradition of disarming those who were violent, forceful, or threatening in the past , but not disarming those likely to pose a Page 23risk of danger in the future , the district court held § 922(g)(3) violated the Second Amendment as applied to non-intoxicated marijuana users like Mr. Harrison. “The use of marijuana,” the district court concluded, “is not in and of itself a violent, forceful or threatening act.” App. at 124.

We endorse much of the district court’s analysis and reasoning. At Bruen step one, we fully agree with the district court that the Second Amendment applies to Mr. Harrison's conduct. At Bruen step two, we agree with the district court in part, as we will explain.

At Bruen step one, we ask whether 'the Second Amendment's plain text covers [Mr. Harrison's] conduct.' Bruen , 597 U.S. at 24. Like the district court, we conclude it does.

The Second Amendment reads, 'A well regulated Militia, being necessary to the security of a free State, the right of the people to keep and bear Arms shall not be infringed.' U.S. Const. amend. II. All agree Mr. Harrison’s conduct—carrying a revolver in his car—constituted “keep[ing]” or “bear[ing]” “Arms.” See Bruen , 597 U.S. at 21 (explaining 'the Second Amendment protects the possession and use of weapons that are ‘in common use at the time’” (quoting Heller , 554 U.S. at 627)); Rocky Mountain Gun Page 24Owners , 121 F.4th at 113-14, 116 (describing the relevant inquiries at Bruen step one).

The only question is whether Mr. Harrison is among the 'people' protected by the Second Amendment’s text. Urging reversal, the government insists the Supreme Court has qualified the Second Amendment as applying only to “law-abiding citizens.” Op. Br. at 55–60 (citing, e.g. , Heller , 554 U.S. at 634–35; Bruen , 597 U.S. at 9, 15, 29–31, 33 n.8, 38 & n.9, 60, 70; Voisine v. United States , 579 U.S. 686, 715 (2016) (Thomas, J., dissenting) (“To be constitutional, therefore, a law that broadly frustrates an individual's right to keep and bear arms must target individuals who are beyond the scope of the ‘People’ protected by the Second Amendment.”)). So “unlawful drug users,” the government says, “are not part of the ‘people.’” Op. Br. at 53. Mr. Harrison insists Heller and Bruen show he is among the “people.” Mr. Harrison adds that, across the Bill of Rights, the word “people” has a broad scope.

We agree with the district court and Mr. Harrison. In Rocky Mountain Gun Owners , we concluded law-abiding 18- to 20-year-olds are part of the “people” protected by the Second Amendment. 121 F.3d at 114–16. We also suggested those with prior felony convictions are within the “people.” Id. at 116. And in United States v. Jackson , we held those with prior misdemeanor convictions are among the “people.” 138 F.4th 1244, 1252 (10th Cir. 2025).

Page 25While these cases support our conclusion, they do not conclusively resolve whether Mr. Harrison is part of the 'people.' Neither case addressed individuals who currently violate the law (as opposed to those with prior convictions); in Jackson , the government conceded the Bruen step one issue, id. ; and in Rocky Mountain Gun Owners , we specifically left open 'the definite contours of who ‘the people’ encompasses,” 121 F.4th at 115.

Today, we hold the “people” for purposes of the Second Amendment include, at least, all Americans. With this holding, we join the weight of persuasive authority since Rahimi . See, e.g. , United States v. Diaz , 116 F.4th 458, 466 (5th Cir. 2024) (“[A]ll people have the right to keep and bear arms . . . .' (first alteration in original) (quoting Kanter , 919 F.3d at 452 (Barrett, J., dissenting))); United States v. Williams , 113 F.4th 637, 649 (6th Cir. 2024) (“The right thus belongs to ‘all Americans.’” (quoting Heller , 554 U.S. at 580)); see also United States v. Duarte , 137 F.4th 743, 752–53 (9th Cir. 2025) (en banc) (finding a defendant is part of “the people” because he “is undoubtedly a member of the national community”).

As the district court recognized, a broad reading of “people” follows the Supreme Court’s clear directive in its Second Amendment jurisprudence. Heller recognized a 'strong presumption that the Second Amendment right . . . belongs to all Americans.” 554 U.S. at 581; see also id . at 580 (suggesting the Second Amendment applies to 'all members of Page 26the political community'). Bruen stated without reservation that the Second Amendment covers “all Americans.” Bruen , 597 U.S. at 70 (quoting Heller , 554 U.S. at 581). And most recently, in Rahimi , the Court did not dwell on Bruen step one and proceeded straight to Bruen step two, essentially assuming Mr. Rahimi was among the “people”—regardless of whether he was “law-abiding.” See Rahimi , 602 U.S. at 686, 690–99; id . at 752 (Thomas, J., dissenting) (“It is . . . undisputed that the Second Amendment applies to Rahimi.”).

As the government stresses, the Court has frequently referenced the Second Amendment’s application to “law-abiding citizens.” E.g. , Heller , 554 U.S. at 625, 635; Bruen , 597 U.S. at 26, 29, 30, 31. But in Rahimi , the Court disagreed that its use of the word ''responsible' to describe the class of ordinary citizens who undoubtedly enjoy the Second Amendment right' necessarily excluded 'citizens who were not 'responsible.'' Rahimi , 602 U.S. at 701–02; see Rocky Mountain Gun Owners , 121 F.4th at 122 (making this point). That logic forecloses the government’s argument Mr. Harrison may be disarmed simply because he is not 'law-abiding.'⁹

Page 27The government also relies on Justice Thomas’s dissent in Voisine , which it calls the “clearest statement on the issue.” Op. Br. at 57. Justice Thomas wrote, “To be constitutional . . . a law that broadly frustrates an individual’s right to keep and bear arms must target individuals who are beyond the scope of the 'People' protected by the Second Amendment.' Voisine , 579 U.S. at 715 (Thomas, J., dissenting). Whatever this statement may have implied about the scope of the “people” before Rahimi , Justice Thomas clarified his approach in his Rahimi dissent. He wrote, '[T]he Second Amendment extends to ‘the people,’ and that ‘term unambiguously refers to all members of the political community, not an unspecified subset.’” 602 U.S. at 752 (Thomas, J., dissenting) (quoting Heller , 554 U.S. at 580). This is consistent with the conclusion we reach today.

A contrary conclusion would defy law and logic. The First and Fourth Amendments also refer to the 'people,' and nobody contends only 'lawabiding citizens” enjoy the rights protected by these constitutional guarantees. See U.S. Const. amend. I (establishing 'the right of the people peaceably to assemble”); U.S. Const. amend. IV (establishing “[t]he right of the people to be secure in their persons, houses, papers and effects”). We have specifically recognized “the meaning of the phrase ‘the people’ does not appear to vary across the Constitution.” Rocky Mountain Gun Owners , 121 F.4th at 116 . It would be 'implausible under ordinary principles of Page 28construction” to conclude the word “people” has different meanings across provisions. Williams , 113 F.4th at 649; see also United States v. VerdugoUrquidez , 494 U.S. 259, 265 (1990) (suggesting “‘the people’ seems to have been a term of art employed in select parts of the Constitution,” and “‘the people’ protected by the Fourth Amendment, and by the First and Second Amendments . . . refers to a class of persons who are part of a national community'). This provides additional reason to doubt the government's attempt to limit the Second Amendment’s text.

Next, restricting the Second Amendment to “law-abiding” citizens— as the government urges us to do—would make it harder to administer and would risk turning it into 'a second-class right.' Bruen , 597 U.S. at 70 (quoting McDonald v. City of Chicago , 561 U.S. 742, 780 (2010) (plurality opinion)). We do not even know precisely what “law-abiding” means. How many laws must someone break to fall outside the Second Amendment’s ambit? And how severe must those violations be? The government provides no answers. See App. at 95 n.21 (the district court asking “who among us, after all, isn’t a ‘lawbreaker’?”).

In reaching our conclusion, we find instructive how then-Judge Barrett explained the issue. “Some maintain that there are certain groups of people . . . who fall entirely outside the Second Amendment's scope. Others maintain that all people have the right to keep and bear arms but Page 29that history and tradition support Congress’s power to strip certain groups of that right.' Kanter , 919 F.3d at 451–52 (Barrett, J., dissenting) (citations omitted). The first approach “uses history and tradition to identify the scope of the right, and the [second] uses that same body of evidence to identify the scope of the legislature's power to take it away.' Id. at 452. Like then-Judge Barrett, we think 'the latter is the better way to approach the problem.' Id.

We thus discern no error in the district court’s conclusion at Bruen step one. Mr. Harrison is among the 'people' protected by the Second Amendment.

B

We now move to Bruen step two. Here, we engage in what the Supreme Court calls “analogical reasoning.” Bruen , 597 U.S. at 28–30. The government must marshal historical analogues to demonstrate § 922(g)(3)-in its application to non-intoxicated marijuana users-is consistent with the Nation’s historical tradition of firearm regulation.

We first ask whether the statute 'addresses a general societal problem that has persisted since the 18th century.” Id. at 26. Recall, if it does, then the lack of a 'distinctly similar historical regulation . . . is relevant evidence' § 922(g)(3) violates the Second Amendment. Id. The district court reasoned § 922(g)(3) addresses a persistent societal problem, and we agree. And the parties seem to agree there is no historical regulationPage 30 “distinctly similar” to § 922(g)(3) as applied to non-intoxicated marijuana users.

We then move to the heart of Bruen step two. We ask: Has the government shown § 922(g)(3), as applied here, 'is consistent with the principles that underpin our regulatory tradition?' Rahimi , 602 U.S. at 692. The district court answered no, but as we will explain, we ultimately conclude otherwise.

Under Bruen step two, our threshold inquiry focuses on whether § 922(g)(3)—as applied to non-intoxicated marijuana users—“addresses a general societal problem that has persisted since the 18th century.” Bruen , 597 U.S. at 26. If it does, then the lack of a historical regulation “distinctly similar” to § 922(g)(3)—again, as applied here—is “relevant evidence” the statute violates the Second Amendment. Id.

What is the problem addressed by § 922(g)(3)?¹⁰ In the district court's view, § 922(g)(3) concerns 'possession of firearms by users of substances Page 31with the potential for abuse.” App. at 98. Mr. Harrison offers an understanding similar to the district court's. See Ans. Br. at 31. The government, meanwhile, tells us § 922(g)(3) addresses the “risk of dangerousness posed by the possession of firearms by individuals who unlawfully use or are addicted to controlled substances.” Op. Br. at 7. The government's description of this problem is narrower than the district court’s in two important respects. According to the government, § 922(g)(3) addresses a narrow problem of those who use 'controlled substances'which the law defines to exclude alcohol—and specifically those who use those substances “unlawfully.” See Op. Br. at 7–10. With this understanding, the government suggests the 'problem addressed by § 922(g)(3) could not have existed in the 18th century because there were no drug laws.” Op. Br. at 7.

We agree with the district court that § 922(g)(3) addresses the dangerous mixture of guns and intoxicants. Recall, Bruen asks us to determine whether modern and historical laws face the same ' general societal problem,” which counsels against allowing the government to define the problems addressed by modern laws too narrowly. 597 U.S. at 26 (emphasis added). Bruen itself evaluated the purpose of a modern law with a high level of generality-finding a New York regulation addressed the overall problem of 'handgun violence' in 'urban area[s].' Id. at 27Page 32 (alteration in original) (quoting Heller , 554 U.S. at 634). The government, meanwhile, attempts to define the problem addressed by § 922(g)(3) so narrowly as to precisely match the statute's scope. But if that were permissible, almost no modern gun regulations would address general societal problems that have persisted since the founding.¹¹

Having determined § 922(g)(3) generally addresses the dangers of mixing guns and intoxicants, we ask whether the Founders were concerned with a similar 'general' problem. They undisputedly were. The district court aptly observed, “[H]istorical prohibitions on the carrying of firearms by intoxicated persons . . . proves” this concern is longstanding. App. at 98; see 1 William Waller Hening, The Statutes at Large: Being a Collection of all the Laws of Virginia 401-02, Act XII (1823) (1655 law); 5 The Colonial Laws of New York from the Year 1664 to the Revolution 244-46, ch. 1501 (1894) (1771 law).

Page 33Consider how colonial New York and Virginia explained laws disarming intoxicated people. New York was concerned about “damages . . . frequently done on . . . New Year’s days, by persons . . . with guns and other fire arms and being often intoxicated with liquor, [who] have not only put the inhabitants in great terror, but committed many mischiefs.” 5 Colonial Laws of New York, supra , at 244-45, ch. 1501 (1771 law) (capitalization modernized). Virginia expressed concern that alarms could not warn against Indian invasion, and the public would waste gunpowder, if men could use firearms while drinking alcohol. See 1 Hening, supra , at 401–02, Act XII (1655 law). Likewise, as the government itself stresses, the Founders understood drug use to cause “temporary insanity.” Op. Br. at 45 (quoting Thomas Cooley , A Treatise on the Constitutional Limitations Which Rest upon the Legislative Power of the States of the American Union 599 n.1 (1868)).

This historical evidence concerns regulations disarming intoxicated people. Here, Mr. Harrison challenges § 922(g)(3) as applied to nonintoxicated marijuana users. But the general problem § 922(g)(3) addresses does not vary based on how the statute is applied. No party contends otherwise. And in Bruen , the Court accepted the challenged regulation addressed a persistent societal problem without citing significant foundingera evidence. 597 U.S. at 27. We are therefore satisfied, on the record before Page 34us, that § 922(g)(3) addresses a societal problem that has persisted since the founding, including in its application to Mr. Harrison—the dangerous mixture of firearms and intoxicants. See United States v. Veasley , 98 F.4th 906, 912 (8th Cir. 2024) (“The lesson here is that disarmament is a modern solution to a centuries-old problem.”).

The parties seem to agree there is no historical regulation “distinctly similar' to § 922(g)(3) as applied to non-intoxicated marijuana users. Bruen , 597 U.S. at 26. Instead, when the Founders addressed the dangerous mixture of firearms and intoxicants, they seemed to disarm only intoxicated people. Bruen instructs us to treat this as “relevant evidence” that, in the context litigated here, § 922(g)(3) violates the Second Amendment. Id. The parties dispute what it means for the lack of a “distinctly similar” historical regulation to be “relevant evidence” § 922(g)(3) is unconstitutional.

The government contends the lack of a “distinctly similar” historical regulation does not alone determine whether § 922(g)(3) “is consistent with this Nation’s historical tradition of firearm regulation”—the question posed by Bruen step two. Id. at 17. 'Regardless of whether § 922(g)(3) addresses a general societal problem that existed in the 18th century,” the government explains, “analogical reasoning still applies.” Op. Br. at 11. Mr. Harrison insists 'the lack of historical laws distinctly similar to § 922(g)(3) is Page 35dispositive” and “ establishes that § 922(g)(3) is inconsistent with the Second Amendment.' Ans. Br. at 31 (emphasis added).¹² We agree with the government.

The Supreme Court has never said the lack of a distinctly similar historical regulation is dispositive of the constitutional question. Rather, Bruen described this absence as “relevant evidence,” no more and no less. 597 U.S. at 26. And if there were any doubt after Bruen , the Court clarified its approach in Rahimi . There, the Court engaged in analogical reasoning without even discussing whether the challenged regulation addressed a persistent societal problem, or whether there was a distinctly similar historical regulation, even though the dissent invoked these concepts. See Rahimi , 602 U.S. at 693-700; id. at 750-51 (Thomas, J., dissenting); see also Diaz , 116 F.4th at 471 (stating Bruen and Rahimi 'rejected' the argument that the lack of a distinctly similar law can be dispositive).

'[T]he absence of a distinctly similar historical regulation in the presented record, though undoubtedly relevant, can only prove so much.” Antonyuk v. James , 120 F.4th 941, 969 (2d Cir. 2024). When a regulation is absent from the historical record, perhaps the Founders believed it was Page 36unconstitutional. But this is a limited heuristic. Holding it too closely would improperly 'assume[] that founding-era legislatures maximally exercised their power to regulate” and “adopt[] a ‘use it or lose it’ view of legislative authority,” which “originalism does not require.” Rahimi , 602 U.S. at 739– 40 (Barrett, J., concurring); Pitsilides v. Barr , 128 F.4th 203, 209 (3d Cir. 2025) (“[W]e remain vigilant not to ‘assume[ ] that founding-era legislatures maximally exercised their power to regulate' conduct.' (quoting Rahimi , 602 U.S. at 740 (Barrett, J., concurring)). There are countless reasons the Founders may not have used a distinctly similar regulation to one used today: they were less concerned with the specific problem than we are, they prioritized legislation in other domains, logistics of enforcement were more tenuous, or non-constitutional policy concerns favored different paths. None of those reasons implicate the Second Amendment. We therefore use the lack of a historical regulation distinctly similar to § 922(g)(3) as a data point in our analogical inquiry.

We now move to the heart of Bruen step two. By examining historical analogues, we ask whether the “principles” in “our regulatory tradition” are “consistent with” Congress’s decision to disarm non-intoxicated marijuana users under § 922(g)(3). Rahimi , 602 U.S. at 692. Recall, the district court concluded § 922(g)(3) did not fall within a historical tradition of either Page 37disarming the mentally ill or disarming those believed to pose a risk of danger. The government argues these holdings were erroneous. According to the government, § 922(g)(3) 'is relevantly similar to' both 'laws disarming the mentally ill' and 'laws disarming those believed to be dangerous.'¹³ Op. Br. at 43, 14. We address each argument in turn.

a

Defending the constitutionality of § 922(g)(3), the government first insists our Nation has a broad tradition of disarming the mentally ill. It maintains the Supreme Court acknowledged this tradition in Heller , when it stated 'nothing in our opinion should be taken to cast doubt on longstanding prohibitions on the possession of firearms by . . . the mentally ill.” 554 U.S. at 626. The government also relies on a Supreme Court case from 1849 discussing an 'acknowledged right'-outside the firearm context—to “exclude . . . lunatics.” Op. Br. at 44 (quoting Smith v. Turner , 48 U.S. 283, 463 (1849)); a 1788 New York law allowing the state to seize property from a “lunatick” until “he comes to his right mind,” Op. Br. at 44Page 38 (quoting 2 Laws of the State of New York 617, ch. 12 (1886) (1788 law)); and a variety of secondary sources, Op. Br. at 43–44 (citing, e.g. , Robert Dowlut, Comment, The Right to Arms: Does the Constitution or Predilection of Judges Reign? , 36 Okla. L. Rev. 65, 96 (1983); Cooley , supra , at 28-29 (1868)).

The government then argues the mentally ill are “relevantly similar” to non-intoxicated marijuana users, such that laws disarming the former support laws disarming the latter. Op. Br. at 45-47. The government defends this analogy by insisting the Founders understood intoxication to be akin to temporary insanity, and “scientific studies have noted the link between mental illness and illicit drug use.” Op. Br. at 45 (citing sources). Ultimately, the government argues, § 922(g)(3) matches the “how” and the “why” of a founding-era practice: it disarms a “limited, narrowly tailored” group (the how), because that group is not “responsible,” “ordinary,” or able to exercise 'self-control' (the why). Op. Br. at 46 (quotations omitted). Urging affirmance, Mr. Harrison asks us to endorse the district court's conclusion that laws disarming the mentally ill are not instructive here.¹⁴

Page 39We agree with the district court and Mr. Harrison. On the record before us, the government has not shown laws disarming the mentally ill are relevant historical analogues. The government suggests laws disarming the mentally ill reveal a principle that legislatures may disarm those who are not “responsible,” “ordinary,” or able to exercise “self-control.” Op. Br. at 46 (quotations omitted). This analysis relies on constructs the Supreme Court has explicitly refused to endorse. In Rahimi , the Court “reject[ed] the Government’s contention that Rahimi may be disarmed simply because he is not 'responsible.'' 602 U.S. at 701. Rahimi teaches that capacious and subjective factors—like whether a group is “responsible,” “ordinary,” or able to exercise 'self-control'-are unacceptable grounds for stripping the Second Amendment right.¹⁵

Page 40Further, the government's arguments linking the mentally ill and non-intoxicated marijuana users are unhelpful. For instance, even if the founders viewed intoxication as temporary insanity, as the government contends, we do not see how that would apply to non-intoxicated marijuana users. See United States v. Connelly , 117 F.4th 269, 276 (5th Cir. 2024) (“Repeat marijuana users . . . are of sound mind upon regaining sobriety . . . .').¹⁶ No matter what our historical tradition might reveal about disarming the mentally ill, that principle is not broad enough to extend to non-intoxicated marijuana users under § 922(g)(3).

We turn now to the second historical principle the government identifies to justify § 922(g)(3). According to the government, historical regulations reveal the principle that a legislature may disarm those “believed to be dangerous.” Op. Br. at 14. The district court agreed with the Page 41government insofar as “the Nation’s broader historical tradition demonstrate[s] that limitations on the right to armed self-defense ‘were tied to dangerousness.’” App. at 119 (quoting Folajtar , 980 F.3d at 913 (Bibas, J., dissenting)). But the district court’s understanding of “dangerousness” was exclusively backward looking: “‘[T]he legislature may disarm those who have demonstrated a proclivity for violence' through past violent, forceful, or threatening conduct . . . .' App. at 117-18 (emphasis added) (quoting Kanter , 919 F.3d at 454 (Barrett, J., dissenting)). The government urges reversal, insisting legislatures throughout our historical tradition have disarmed groups based on a concern about future danger. To support this theory, the government points to pre-founding English laws and colonial laws disarming Catholics and loyalists-categories of historical evidence the district court did not find probative. In urging affirmance, Mr. Harrison reprises the district court’s reasoning.

We agree with the district court, but only in part. The district court correctly recognized that our Nation's tradition of firearm regulation supports disarming people who are dangerous. The district court was also right that past actions can serve as the basis to conclude an individual is dangerous. See Jackson , 138 F.4th at 1254 (explaining the defendant’s “acts of domestic violence and his subsequent convictions demonstrate his 'propensity for the use of physical violence against others,'' so he 'was Page 42disarmed for reasons consistent with this Nation's tradition of firearm regulation” (quoting United States v. Rogers , 371 F.3d 1225, 1229 (10th Cir. 2004))). But, as we will explain, the government has identified historical analogues that reveal another key principle underpinning our Nation's history of firearm regulation: legislatures may disarm those believed to pose a risk of future danger. We thus agree with the district court about one reason 'why' our predecessors disarmed certain individuals-to prevent dangerous people from possessing arms-but disagree about 'how' they made those determinations—they looked both backward and forward. See Bruen , 597 U.S. at 29 (instructing courts to focus on 'how and why . . . regulations” disarm).

In recognizing a historical principle that legislatures may disarm those believed to pose a risk of future danger, we join many other circuit courts that have understood the history the same way. See Williams , 113 F.4th at 657 (“This historical study reveals that governments in England and colonial America long disarmed groups that they deemed to be dangerous. . . . [G]overnments labeled whole classes as presumptively dangerous.” (emphasis added)); Duarte , 137 F.4th at 761 (en banc) (“[F]our centuries of unbroken Anglo-American history shows that legislatures consistently disarmed entire categories of people who were presumed to pose a special risk of misusing firearms.” (alteration in original) (emphasis Page 43added) (quoting Range , 124 F.4th at 273 (en banc) (Krause, J., concurring in the judgment))); United States v. Perez-Garcia , 96 F.4th 1166, 1189 (9th Cir. 2024) ('[F]ounding-era legislatures categorically disarmed groups whom they judged to be a threat to the public safety.' (emphasis added) (quoting Kanter , 919 F.3d at 458) (Barrett, J., dissenting)); United States v. Jackson , 110 F.4th 1120, 1128 (8th Cir. 2024) ('Legislatures historically prohibited possession by categories of persons based on a conclusion that the category as a whole presented an unacceptable risk of danger if armed.” (emphasis added)); United States v. Hunt , 123 F.4th 697, 707 (4th Cir. 2024) (“A determination of dangerousness was sometimes made by status . . . and sometimes by conduct . . . . Those historical restrictions swept broadly, disarming all people belonging to groups that were, in the judgment of those early legislatures, potentially violent or dangerous.' (emphasis added)); 17Page 44 Zherka v . Bondi, No. 22-1108-CV, 2025 WL 1618440, at *16 (2d Cir. June 9, 2025) ('[B]efore, during, and shortly after the Founding, legislative bodies regulated firearms by prohibiting their possession by categories of persons perceived to be dangerous.' (emphasis added)); see also Kanter , 919 F.3d at 454 (Barrett, J., dissenting) (“[T]he legislature may disarm those who have demonstrated a proclivity for violence or whose possession of guns would otherwise threaten the public safety. This is a category . . . [that] includes dangerous people who have not been convicted of felonies . . . .' (emphasis added)); Folajtar , 980 F.3d at 913 (Bibas, J., dissenting) ('Violence was one ground for fearing danger, as were disloyalty and rebellion.' (emphasis added)); Binderup , 836 F.3d at 368 (en banc) (Hardiman, J., concurring in part) (''[F]rom time immemorial, various jurisdictions recognizing a right to arms have . . . taken the step of forbidding suspect groups from having arms,' and 'American legislators at the time of the Bill of Rights seem to have been aware of this tradition.’” (second alteration in original) (emphasis added) (quoting Don B. Kates & Clayton E. Cramer, Second Amendment Limitations and Criminological Considerations , 60 Hastings L.J. 1339, 1360 (2009)))¹⁸ ; cf. United States v.

Page 45Harris , No. 21-3031, 2025 WL 1922605 (3d Cir. July 14, 2025) (concluding, in a challenge to the constitutionality of § 922(g)(3), the history shows '[d]rug users can be disarmed based on the likelihood that they will physically harm others if armed” (emphasis added)). Although most of these cases did not involve § 922(g)(3), they interpreted our Nation’s history of gun regulation using the methodology of Bruen and Rahimi , so they are highly instructive.

i

At the outset, Mr. Harrison raises an argument rooted in Rahimi that we must address. According to Mr. Harrison, Rahimi already forbids disarmament based on legislative judgments about risk. He says Rahimi limited disarmament to situations involving 'a judicial finding' that an individual poses a threat. Harrison Supp. Resp. Br. at 6 (emphasis added).

We disagree with Mr. Harrison’s reading of Rahimi . In Rahimi , the Supreme Court recognized historical laws 'confirm what common sense suggests: When an individual poses a clear threat of physical violence to another, the threatening individual may be disarmed.” Rahimi , 602 U.S. at 698. To be sure, Rahimi concerned § 922(g)(8)(C)(i), a law that disarms Page 46someone only if a judge finds they pose a risk of danger, while this case involves § 922(g)(3), a law that disarms those whom a legislature found pose a risk of danger. Id. at 688-89, 698-99.¹⁹ But Rahimi expressly declined to address whether legislative judgments of danger can justify disarmament. The Supreme Court instead clarified that it did “not suggest that the Second Amendment prohibits the enactment of laws banning the possession of guns by categories of persons thought by a legislature to present a special danger of misuse.' Rahimi , 602 U.S. at 698; see Gov. Supp. Resp. Br. at 4 (highlighting this excerpt from Rahimi ).

Following Rahimi , we must answer in the first instance whether it is “consistent with the principles that underpin our regulatory tradition” for a legislature to disarm those it believes to pose a risk of future danger. Id. at 692. Turning to that question, we agree with the government that the answer is yes.

Turning to the history, we begin in England.²⁰ See Bruen , 597 U.S. at 20 (“The Amendment ‘was not intended to lay down a novel principle but Page 47rather codified a right inherited from our English ancestors.'' (quoting Heller , 554 U.S. at 599)). The Militia Act of 1662 permitted government officials “to search for and seize all arms in the custody or possession of any person or persons whom [the official] shall judge dangerous to the peace of Page 48the kingdom.” Militia Act of 1662, 13 & 14 Car. 2, c.3, § 13, 5 Statutes of the Realm 360 (capitalization and spelling modernized). This provides early evidence that individuals believed to pose a risk of danger may be disarmed. Because the Militia Act did not limit disarmament to those who had acted dangerously in the past, it outlines a more expansive principle of disarmament than the district court acknowledged.

Mr. Harrison argues the 1689 English Bill of Rights casts some doubt on the Militia Act’s relevance. The English Bill of Rights—which contains the 'predecessor to our Second Amendment,' Heller , 554 U.S. at 593guaranteed 'that the subjects which are Protestants may have arms for their defence suitable to their conditions, and as allowed by law.” 1 W. & M., ch. 2, § 7; see Bruen , 597 U.S. at 44 (calling the English Bill of Rights “a watershed”). Assuming the English Bill of Rights cabined the Militia Act, it does not seem to have abrogated the Militia Act. Compare Rahimi , 602 U.S. at 694 (suggesting the English Bill of Rights was a “rebuke” of some of the conduct enabled by the Militia Act), and Connelly , 117 F.4th at 278 (stating the English Bill of Rights 'qualified' the Militia Act), with Joyce Lee Malcolm, To Keep and Bear Arms 123 (1994) ('As it turned out, the Militia Act of 1662 was to remain in force with only insignificant changes for many years to come.”), and Perez-Garcia , 96 F.4th at 1187 (“Use of the Militia Act provisions allowing search and seizure of weapons from disaffected personsPage 49 'continued unabated' after the adoption of the 1688-89 English Bill of Rights.” (quoting Diarmuid F. O’Scannlain, Glorious Revolution to American Revolution: The English Origin of the Right to Keep and Bear Arms , 95 Notre Dame L. Rev. 397, 405 (2019))); see also Calendar of State Papers, Domestic Series, of the Reign of William III, 1 April, 1700–8 March, 1702 , at 234 (1937) (showing, in 1701, William III instructed government officers to “search . . . for arms in the possession of any persons whom they judge dangerous, and seize such arms according to law”). We are in good company extracting relevant historical evidence from the Militia Act. See Williams , 113 F.4th at 651; Perez-Garcia , 96 F.4th at 1187; Jackson , 110 F.4th at 1126; see also Kanter , 929 F.3d at 456 (Barrett, J., dissenting); Folajtar , 980 F.3d at 914 (Bibas, J., dissenting).

Under the English Bill of Rights itself, legislatures could disarm those believed to pose a risk of future danger. The document was palpably limited. A 'principle that arms-bearing was constrained 'by Law' remained,' Rahimi , 602 U.S. at 694, and the document's terms mentioned only Protestants. This was no accident. Parliament explicitly disarmed Catholics unless they swore a loyalty oath.¹ W. & M., ch. 15, § 3, 6 Statutes of the Realm 71-72 (1688); see Malcolm, supra , at 122 ('About a month after Parliament passed the Declaration of Rights, it took up the problem of disarming Catholics.”).

Page 50It appears Parliament disarmed Catholics because of a concern that Catholics-given their status as Catholics-posed a risk of danger. See Malcolm, supra , at 122 (“Prevention of a Catholic counter-revolution was of paramount concern.”); 4 William Blackstone, Commentaries on the Laws of England 54 (16th ed. 1825) (1769) (suggesting the English disarmed Catholics because they threatened 'subversion of the civil government'); see also C. Kevin Marshall, Why Can’t Martha Stewart Have A Gun? , 32 Harv. J.L. & Pub. Pol’y 695, 723 (2009) (“In short, the stated principle supporting the disability was cause to fear that a person, although technically an English subject, was because of his beliefs effectively a resident enemy alien liable to violence against the king.”). Of course, as hardly requires mention, “not all . . . Catholics in England . . . were violent.” Jackson , 110 F.4th at 1128.

Again, we are not alone in using England’s disarmament of Catholics to inform the original scope of the Second Amendment. See id. at 1126; Williams , 113 F.4th at 651; Duarte , 137 F.4th at 759 (en banc); Zherka , 2025 WL 1618440, at *13; Kanter , 919 F.3d at 457 (Barrett, J., dissenting) ('And-perhaps unsurprisingly because they were presumptively thought to pose a [] threat or terror—Parliament also disarmed Catholics.” (emphasis added)). “To be sure, the American experience does not map on exactly to the English one. . . . Still, the American version [of the right to arms] was Page 51derived from its English predecessor, which makes English practice instructive.” Kanter , 919 F.3d at 457 n.5 (Barrett, J., dissenting) (citation omitted).

In any event, “[w]hile many practices didn’t survive the odyssey from the Old World to the New, the desire to promote peace by disarming dangerous groups arrived intact.' Williams , 113 F.4th at 652. First are three colonial laws disarming Catholics, continuing the English tradition apace. See 5 Statutes at Large of Pennsylvania from 1682 to 1801 , at 627 (1898) (1759 law) (declaring '[t]hat all arms . . . any papist or reputed papist within this province hath or shall have in his house . . . shall be taken from such papist or reputed papist”); 7 William Waller Hening, The Statutes at Large: Being a Collection of all the Laws of Virginia 35–38 (1820) (1756 law); 52 Archives of Maryland 454 (1935) (1756 law).²¹ These colonies disarmed Catholics because they concluded “it is dangerous at this time to permit Papists to be armed.” 7 Hening, supra , at 35 (1756 law); see also Williams , 113 F.4th at 653 (concluding the historical record showsPage 52 “Pennsylvanians believed that such measures were absolutely necessary to protect against Catholic-led violence” (quotation omitted)). As in England, 'not all Catholics' in the colonies 'were violent.' Jackson , 110 F.4th at 1128. Other circuits have, again, found these laws instructive. Id. at 1126; Williams , 113 F.4th at 653; Perez-Garcia , 96 F.4th at 1187; Duarte , 137 F.4th at 759 (en banc); Zherka , 2025 WL 1618440, at *14.

Turning next to the revolution, the government highlights a tradition of disarming loyalists because they were believed to pose a risk of danger. The Continental Congress “recommended to the several [colonies], immediately to cause all persons to be disarmed within their respective colonies, who are notoriously disaffected to the cause of America.'⁴ Journals of the Continental Congress 205 (1776). Many young states complied.⁹ Statutes at Large of Pennsylvania from 1682 to 1801 , at 348 (1903) (1779 law) (empowering officials “to disarm any person or persons who shall not have taken any oath or affirmation of allegiance”); 9 William Waller Hening, The Statutes at Large: Being a Collection of all the Laws of Virginia 281-82 (1821) (1777 law); 24 The State Records of North Carolina 89 (1905) (1777 law); Acts of the General Assembly of the State of New Jersey 90 (1777 law); 5 The Acts and Resolves, Public and Private, of the Province of Massachusetts Bay 479, ch. 21 (1886) (1776 law). Pennsylvania's law is especially probative as to the original scope of the Second Amendment, Page 53because Pennsylvania disarmed loyalists soon after it enshrined a “right to bear arms.” Pennsylvania Declaration of Rights, XIII (1776); see Williams , 113 F.4th at 654 n.11 (making this point). “[C]onfiscation of guns from those who refused to swear an oath of allegiance was meant to 'deal with [a] potential threat . . . .’” Kanter , 919 F.3d at 457 (Barrett, J., dissenting) (emphasis added) (quoting Saul Cornell & Nathan DeDino, A Well Regulated Right: The Early American Origins of Gun Control , 73 Fordham L. Rev. 487, 506 (2004)); see also Joseph G.S. Greenlee, The Historical Justification for Prohibiting Dangerous Persons from Possessing Arms , 20 Wyo. L. Rev. 249, 265 (2020) (concluding “the justification” for these laws “was always that those being disarmed were dangerous'). But 'not all early Americans who declined to swear an oath of loyalty . . . were violent.' Jackson , 110 F.4th at 1128.

We join others in using colonial laws disarming loyalists to discern our Nation's historical tradition of gun regulation. Id. at 1126-28; Williams , 113 F.4th at 653–54; Perez-Garcia , 96 F.4th at 1187; Duarte , 137 F.4th at 759 (en banc); Zherka , 2025 WL 1618440, at *15; see also Folajtar , 980 F.3d at 914 (Bibas, J., dissenting) ('Loyalists were potential rebels who were dangerous before they erupted into violence. . . . To ensure peace and safety, the colonies had to disarm them.' (emphasis added)); Binderup , 836 F.3d at 368 (en banc) (Hardiman, J., concurring in part) ('[A]lthough these Page 54Loyalists were neither criminals nor traitors , American legislators had determined that permitting these persons to keep and bear arms posed a potential danger.' (emphasis added) (quoting NRA v. ATF , 700 F.3d 185, 200 (5th Cir. 2012))).

Mr. Harrison does not dispute the existence of these English laws and colonial laws disarming Catholics and loyalists. At most, relying on a sentence in Rahimi , he contends disarming 'political opponents' was 'largely eliminated' in the colonies. Harrison Supp. Br. at 7 (quoting Rahimi , 602 U.S. at 694). We are not persuaded this gives us good reason to disregard the significant historical evidence marshaled by the government—particularly given the methodology Bruen and Rahimi prescribe.

In sum, “Colonial era laws thus demonstrate that England’s history and tradition of disarming dangerous individuals continued across the Atlantic Ocean. Colonial governments frequently deemed entire groups too dangerous to possess weapons.' Williams , 113 F.4th at 654. And 'every categorical disarmament law was overbroad-sweeping in law-abiding people who were not dangerous, violent, untrustworthy, or unstable.' Duarte , 137 F.4th at 761 (en banc) (quoting Range , 124 F.4th at 267 (en Page 55banc) (Krause, J., concurring)).²² Recall, Bruen requires courts to assess “why” and “how” historical regulations burdened the right to bear arms. 597Page 56 U.S. at 29. The colonial laws manifest a 'why'-preventing dangerous people from possessing firearms-that aligns with the district court's reasoning. But these laws also reflect a “how”—legislatures disarming those believed to pose a risk of future danger—which is broader than what the district court allowed.

As we have explained, the district court’s conclusion—that legislatures may only disarm those who have acted dangerously in the past —did not fully reflect the historical laws in the record, which govern our inquiry at Bruen step two. See Bruen , 597 U.S. at 24, 28–29. One key reason the district court rejected a broader principle of disarmament is that it dismissed the relevance of laws disarming Catholics and loyalists. We Page 57now explain why these laws are probative historical analogues, and the district court’s contrary view was mistaken.

First , the district court thought the Founders intended to repudiate the principle behind disarming Catholics and loyalists. According to the district court, James Madison critiqued “the European monarchical practice of being ‘afraid to trust the people with arms.’” App. at 133 (quoting The Federalist No. 46 (James Madison)). That Madison critiqued European monarchies for restricting the right to bear arms, however, does not show the Founders sought to disclaim numerous laws passed in their own colonies.

Second , the district court reasoned, if laws disarming Catholics and loyalists could serve as historical analogues, “then the Second Amendment would provide virtually no limit on Congress’s discretion.” App. at 134. The court doubted the Framers 'incorporated such a trojan horse into the Second Amendment.” App. at 133; see Ans. Br. at 43 (similar).

Here, we must conclude the district court’s reasoning departed from the methodology required by Bruen and affirmed in Rahimi . The Supreme Court has instructed, when it comes to interpreting the Second Amendment, historical evidence is our primary tool. The Court has directed us to examine 'the historical background of the Second Amendment,' because '[t]he Amendment 'was not intended to lay down a novel principle but Page 58rather codified a right inherited from our English ancestors.’” Bruen , 597 U.S. at 20 (quoting Heller , 554 U.S. at 592, 599). And the Court has stated we must 'rel[y] on history to inform the meaning of [the] constitutional text”—including whether the Second Amendment repudiated some historical practice. Id. at 25; see also id. (concluding this historical methodology is 'more legitimate, and more administrable, than asking judges to 'make difficult empirical judgments'' (quoting McDonald , 561 U.S. at 791 (plurality opinion))).

'To be sure, the historical understanding that legislatures have discretion to prohibit possession of firearms by a category of persons . . . who pose an unacceptable risk of dangerousness may allow greater regulation than would an approach that employs means-end scrutiny with respect to each individual person who is regulated.” Jackson , 110 F.4th at 1129. But we agree with our sister circuits that this “result is a product of the method of constitutional interpretation endorsed by Bruen .” Id. ; see Duarte , 137 F.4th at 762 (en banc) (same). Thus, without historical evidence that the Founders repudiated the principle behind disarming Catholics and loyalists, the district court could not assume they did. See Zherka , 2025 WLPage 59 1618440, at *20 (“The test that Bruen requires us to apply uses history as its guide, not policy concerns.').²³

Third , the district court determined 'the disarmament of Catholics and loyalists ‘involved wholesale deprivation’ of other civil liberties as well.” App. at 134 (quoting Marshall, supra , at 726). The court asked, “If such laws cannot create a historical tradition in the First Amendment context, how Page 60could they do so in the Second?” App. at 134–35. We decline to define the First Amendment entirely by reference to English and colonial law because the Supreme Court has not required it. For instance, in the First Amendment context, the Supreme Court has told us “[c]ontent-based laws— those that target speech based on its communicative content—are presumptively unconstitutional and may be justified only if the government proves that they are narrowly tailored to serve compelling state interests.” Reed v . Town of Gilbert , 576 U.S. 155, 163 (2015). That is simply a different question than whether a regulation 'is consistent with the Nation's historical tradition of firearm regulation.” Bruen , 597 U.S. at 24.

Fourth , the district court observed 'these laws were largely passed either during the Seven Years War or . . . the Revolutionary War,' but “[t]imes of war tend to bring out the worst in governments, at least when it comes to civil liberties.” App. at 135. The court stated “we do not look to Korematsu to determine when the government may discriminate based on race.” App. at 135. We respectfully reject this reasoning. As the government persuasively argues, “[R]ights do not mysteriously disappear during times of war.” Op. Br. at 36 (citing Ex Parte Milligan , 71 U.S. 2, 120–21 (1866)); see Milligan , 71 U.S. at 120–21 (“The Constitution of the United States is a law for rulers and people, equally in war and in peace, and covers with the shield of its protection all classes of men, at all times, and under all Page 61circumstances.”). We denounce Korematsu because it “was gravely wrong the day it was decided,” not because it was decided during war. Trump v. Hawaii , 585 U.S. 667, 710 (2018). The Supreme Court has not called for an exception to this rule-that wartime law is still law-in the Second Amendment context.

Fifth , as to laws disarming Catholics, the district court added 'it appears that only two American colonies” ever enacted them, making any purported tradition questionable. App. at 135-36 (citing Bruen , 597 U.S. at 46 ('[W]e doubt that three colonial regulations could suffice to show a tradition of public-carry regulation.')). The court also noted 'at the time these provisions were enacted, the operative rights-protecting document was the English Bill of Rights,” which meant Catholics were “not protected by the right to armed self-defense” in the first place. App. at 136.

We need not decide whether 'three colonial regulations' can ever establish a historical tradition. Bruen , 597 U.S. at 46; see Rahimi , 602 U.S. at 745–46 (Jackson, J., concurring) (identifying the “unresolved question[]” of 'how many analogues add up to a tradition'). This is because the few colonial laws disarming Catholics were not isolated. They joined a larger body of laws revealing a singular principle: the founding generation and their English ancestors permitted legislatures to disarm those believed to pose a risk of future danger. If we refused to group such similar laws Page 62together, then we could never find a tradition of anything. See Rahimi , 602 U.S. at 704 (Sotomayor, J., concurring) (suggesting courts should not “pick[] off the Government's historical sources one by one, viewing any basis for distinction as fatal'); Perez-Garcia , 96 F.4th at 1191 ('[The defendants'] divide-and-conquer approach to the historical evidence misses the forest for the trees.'). As for the district court's statement that the English Bill of Rights excluded Catholics, we are not sure why that makes the colonial disarmament of Catholics less probative. Rather, it appears to only deepen the historical tradition of disarming those believed to pose a risk of danger. Bruen , 597 U.S. at 34.

Sixth , and finally, the district court reasoned laws disarming Catholics and loyalists “were justified on the fear that the covered groups were likely to wage active war . . . . This is a radically different justification than the justification for § 922(g)(3).' App. at 136-37. On this approach, laws disarming Catholics and loyalists could only justify modern laws disarming those who are dangerous in a similar way. Regarding Bruen ’s 'how' and 'why' inquiries, the district court understood laws disarming Catholics and loyalists to involve a narrower “why” than § 922(g)(3) ever could—the focus of the historical laws was to disarm those who might wage war. Bruen , 597 U.S. at 29. The government, alternatively, reads these laws Page 63to show that legislatures may generally 'disarm those believed to pose a greater risk of danger.” Op. Br. at 42.

With the guidance of Rahimi , we land closer to the government. Determining the level of generality at which to define historical principles is no easy task. Concurring in Rahimi , Justice Jackson observed the Court has not 'adequately clarified' 'the level of generality at which a court evaluates [historical] sources.' 602 U.S. at 745 (Jackson, J., concurring). Justice Barrett concluded 'harder level-of-generality problems can await another day.” Id. at 740 (Barrett, J., concurring). And Justice Kavanaugh similarly acknowledged the “important question[]” of “the level of generality at which to define a historical practice.' Id. at 724 n.4 (Kavanaugh, J., concurring) (discussing post-ratification history); see also United States v. Morton , 123 F.4th 492, 498 n.2 (6th Cir. 2024) (asking the Supreme Court to clarify the level of generality at which to understand “laws that forbade Catholics, seditious libelers, Native Americans, loyalists, individuals who had engaged in 'actual rebellion,' and Black people from possessing firearms”).

But Rahimi suggests the government’s proposed level of generality is correct. For one, Rahimi repeated multiple times that historical laws need not be 'twin[s]' of modern laws. Rahimi , 602 U.S. at 692, 701. That Rahimi directed us toward “principles” is also important. Pitsilides , 128 F.4th atPage 64 210 (“[O]ur inquiry into principles that underlie our regulatory tradition does not reduce historical analogizing to an exercise in matching elements of modern laws to those of their historical predecessors.'); Rahimi , 602 U.S. at 703 (Sotomayor, J., concurring) (emphasizing Bruen does not require the government to identify “a precise historical analogue”). The word “principle” suggests a higher level of generality than the district court invoked.

The specific mode of reasoning used in Rahimi —the way the Court generalized from historical laws in that case-is instructive and further supports the government's position. Rahimi upheld a law that disarms individuals based on their risk of committing family violence. See Rahimi , 602 U.S. at 684–85 (citing 18 U.S.C. § 922(g)(8)). Yet the Supreme Court justified its decision in part by reference to historical “going armed” laws— which disarmed those who threatened the 'public order,' not those who threatened family violence.²⁴ Id. at 697–99 (quoting State v. Huntly , 25 N.C.

Page 65418, 421 (1843)). Rahimi understood historical laws that disarmed a specific kind of dangerous person—one who threatened public disorder—to illustrate a principle about disarming dangerous people more generally. Id. at 698 ('Section 922(g)(8) is by no means identical to these founding era regimes, but it does not need to be. Its prohibition on the possession of firearms by those found by a court to present a threat to others fits neatly within the tradition the . . . going armed laws represent.” (citation omitted)); see also id. at 704 (Sotomayor, J., concurring) (“Here, for example, the Government has not identified a founding-era or Reconstruction-era law that specifically disarmed domestic abusers, but it did not need to do so.” (citation omitted)).

The government asks us to do something similar. Laws disarming Catholics and loyalists were focused on war-related violence, like going armed laws were focused on public disorder. We are comfortable generalizing from the former, just as Rahimi generalized from the latter. We thus conclude laws disarming Catholics and loyalists reveal a tradition of disarming those believed to pose a risk of danger, regardless of whether the perceived danger is related to waging war.

Page 66We acknowledge the Fifth Circuit, in holding § 922(g)(3) violated the Second Amendment as applied to non-intoxicated marijuana users, looked at the history differently.²⁵ See Connelly , 117 F.4th at 278-82; see also United States v. Hemani , No. 24-40137, 2025 WL 354982, at *1 (5th Cir. Jan. 31, 2025) (per curiam) (summarily affirming under Connelly the dismissal of an indictment charging a violation of § 922(g)(3)), petition for cert. filed (U.S. June 2, 2025) (No. 24-1234). When considering whether laws disarming Catholics and loyalists provided historical analogues to § 922(g)(3), the Fifth Circuit stated, 'Laws disarming dissidents were passed during wartime or periods of unprecedented societal upheaval. . . . So too with laws disarming religious minorities.” Id. at 278. It continued, “Marijuana users are not a class of political traitors, as English Loyalists were perceived to be. Nor are they like Catholics and other religious dissenters who were seen as potential insurrectionists.' Id. The court Page 67ultimately found laws disarming Catholics and loyalists did not reveal a general principle that legislatures may disarm those believed to be dangerous. See id. The dissent finds the Fifth Circuit's view persuasive. See Dissent at 6 (“I think the Fifth Circuit [in Connelly ] got it right in holding that § 922(g)(3), as applied to a non-intoxicated marijuana user, goes much further than our history and tradition allow.”).

But in our view, the Fifth Circuit’s narrow reading failed to heed the Supreme Court's instruction that 'a 'historical twin ' is not required.' Rahimi , 602 U.S. at 701 (quoting Bruen , 597 U.S. at 30); see also id. (“For its part, the Fifth Circuit . . . read Bruen to require a 'historical twin' rather than a 'historical analogue.'' (quoting Bruen , 597 U.S. at 30)). Likewise, we conclude the Fifth Circuit did not interpret laws disarming Catholics and loyalists with the same level of generality that Rahimi used to interpret going armed laws. See id.

Historical laws disarming Catholics and loyalists may be imperfect sources, but we find they are instructive under the analysis Bruen and Rahimi require.

We now explain why remand is necessary. Pursuant to Rahimi , the government has articulated a “principle[] that underpin[s] our regulatory tradition” that may be “consistent with” the challenged regulation. Id. atPage 68 692. History shows legislatures can disarm those believed to pose a risk of future danger. We respectfully part ways with the district court’s contrary view. That said, we still cannot answer the ultimate constitutional question before us. Rather, with the history correctly understood, the district court must inquire into whether non-intoxicated marijuana users pose a risk of future danger.

The record suggests the parties attempted to engage in this inquiry in the district court. The government asserted it is “dangerous” for “habitual drug users' like Mr. Harrison 'to possess deadly firearms.' App. at 64 (quoting Yancey , 621 F.3d at 685). In support, the government highlighted opinions from other courts collecting “academic research confirm[ing] the connection between drug use and violent crime.” Yancey , 621 F.3d at 686; see App. at 63-64 (citing sources). Though he did not point to contrary evidence, Mr. Harrison insisted the government's assertions about drug users were “highly debatable.” App. at 85. The district court recognized the government was attempting to defend the constitutionality of the statute as applied by showing drug users pose a risk of danger. App. at 124 (acknowledging “the United States points out” its cited cases “are backed up by social science, statistics, and predictions about future crime”). But relying on its understanding of the history, the district court determined it Page 69was legally irrelevant whether non-intoxicated marijuana users pose a risk of future danger.²⁶ See App. at 124–25.

Contrary to the district court’s view, this inquiry is legally relevantand critical to the methodology announced in Bruen and clarified in Rahimi . The Supreme Court has emphasized the Second Amendment 'is not 'a second-class right’” or a matter of complete “legislative interest balancing.” Bruen , 597 U.S. at 70, 26 (quoting McDonald , 561 U.S. at 780); see also Heller , 554 U.S. at 628 n.27 (“If all that was required to overcome the right to keep and bear arms was a rational basis, the Second Amendment would be redundant with the separate constitutional prohibitions on irrational laws, and would have no effect.'); Williams , 113 F.4th at 660 ('If courts uncritically deferred to Congress’s class-wide dangerousness determinations, disarmament laws would most often be subject to rationalbasis review . . . that [would] run[] headlong into Heller . . . [and] allow Page 70legislatures to define away a fundamental right.”); Kanter , 919 F.3d at 465 (Barrett, J., dissenting) ('The legislature must be able to justify its designation . . . .”). We see no evidence the Founders understood legislatures to have complete discretion in selecting whom to disarm. That the Founders disarmed Catholics and loyalists—groups the broader community understood to be dangerous-does not show legislatures can disarm anyone at all. Indeed, historical legislatures often did not categorically disarm those who 'demonstrate[d] that their particular possession of a weapon posed no danger to peace.” Williams , 113 F.4th at 657; see Range , 124 F.4th at 275 (Krause, J., concurring) (“Under categorical disarmament laws, where an individual was presumed to pose a special risk to society by virtue of his membership in a particular group and thus was lawfully disarmed as an initial matter, there was typically a mechanism for him to petition and attempt to rebut that presumption . . . .”); see, e.g. , 9 Statutes at Large of Pennsylvania , supra , at 348 (1779 law) (permitting those who swore oaths to the union to retain their arms); 9 Hening, supra , at 282 (1777 law) (same); 7 Hening, supra , at 35 (1756 law) (permitting Catholics to retain their arms if they took an oath of allegiance); 1 W. & M., ch. 15, § 3, 6 Statutes of the Realm 72 (1688) (same). In short, the Supreme Court’s Second Amendment jurisprudence and the historical record itself show complete deference to the legislature would be inappropriate.

Page 71Accordingly, in this case, the district court should have inquired into whether the government could justify its assertion that non-intoxicated marijuana users pose a risk of danger. See Bruen , 597 U.S. at 24 (placing burden on government at Bruen step two). We find instructive that the Eighth Circuit has called for a similar inquiry in a similar context. In Worth v. Jacobson , the court considered a Second Amendment challenge to Minnesota’s ban on 18 to 20-year olds from carrying firearms. 108 F.4th 677, 683–84 (8th Cir. 2024). Minnesota defended the law by arguing “if the state deems a group of people to pose a risk of danger, it may ban the group’s gun ownership.” Id. at 693. But the Eighth Circuit required Minnesota to 'support its claim with . . . evidence.' Id. at 694. The court of appeals continued, “A legislature’s ability to deem a category of people dangerous based only on belief would subjugate the right to bear arms.” Id . We agree.

The parties ask us to engage in this inquiry in the first instance on appeal. The government maintains “for those who unlawfully use marijuana, like Mr. Harrison, research ‘amply demonstrate[s] a connection between marijuana use specifically and violence.’” Op. Br. at 21 (alteration in original) (quoting United States v. Carter , 750 F.3d 462, 467 (4th Cir. 2014)) (also citing studies). Mr. Harrison contends “marijuana users are not in a class of dangerous people.” Ans. Br. at 40 (citing additional studies).

Page 72But the prudent course is to remand for the district court to make the determination under the correct view of the law, particularly since factfinding may be required.²⁷ United States v. Hasan , 609 F.3d 1121, 1129 (10th Cir. 2010) (“When the court of appeals notices a legal error, it is not Page 73ordinarily entitled to weigh the facts itself and reach a new conclusion; instead, it must remand to the district court for it to make a new determination under the correct law.'); Brammer-Hoelter v. Twin Peaks Charter Acad. , 602 F.3d 1175, 1180 n.1 (10th Cir. 2010) (“It is not within the province of this court to find facts on appeal . . . .”). The Third Circuit recently ordered a similar remand in a Second Amendment challenge to § 922(g)(3). See Harris , 2025 WL 1922605, at *8 (remanding to the district court to consider, as relevant here, “[w]hether use of [a given] drug affects a person's judgment, decision-making, attention, inhibition, or impulse control,” and “[t]he long-term physical and mental effects of the use of that drug”); see also Cooper , 127 F.4th at 1097 ('The district court's task on remand is to figure out which side of the Second Amendment line Cooper’s case falls on.').²⁸

Page 74Ultimately, we agree with much of the district court’s analysis: Mr. Harrison is among “the People” protected by the Second Amendment, so the government must justify § 922(g)(3) by showing it is consistent with our tradition of firearm regulation; § 922(g)(3) lacks a distinctly similar regulation despite addressing a persistent societal problem, which is relevant evidence it violates the Second Amendment as applied; and the government cannot justify § 922(g)(3) by analogizing to laws disarming the mentally ill. We break from the district court in a narrow way. We hold the historical tradition supports a principle that legislatures may disarm those believed to pose a risk of future danger. And we further hold the district court must inquire into the government's assertion that non-intoxicated marijuana users pose a risk of danger.

We REVERSE and REMAND for further proceedings consistent with this opinion.

Page 75No. 23-6028, United States v. Jared Michael Harrison

KELLY , Circuit Judge, concurring in part and dissenting in part.

I concur in the court’s holdings that Mr. Harrison is properly mounting an asapplied challenge to § 922(g)(3), that he is among “the People” protected by the Second Amendment, and that the government cannot justify § 922(g)(3) by analogy to laws disarming the mentally ill. However, I disagree with the court’s conclusion that § 922(g)(3)’s disarmament of Mr. Harrison as a non-intoxicated marijuana user could be consistent with the principles underpinning this Nation’s history and tradition of firearms regulation. I also disagree with the court’s decision to remand the case in a manner which allows the government a fresh start in its prosecution of Mr. Harrison.

Beginning with the historical record, the societal problem posed by “the dangerous mixture of guns and intoxicants” is not a novel one. Ct. Op. at 31–34. The court recognizes this and thoroughly recounts the history of intoxication laws from the Founding.¹ See id. Suffice it to say that these intoxication laws reveal a clear principle: the Founders addressed the dangerous mixture of guns and intoxicants by disarming only actively intoxicated persons. United States v. Cooper, 127 F.4th 1092, 1096 (8th Cir. 2025) (“[I]ntoxication has been prevalent throughout our nation’s history, but earlier Page 76generations addressed that societal problem by restricting when and how firearms could be used, not by taking them away.” (quotations omitted)); United States v. Connelly, 117 F.4th 269, 280–82 (5th Cir. 2024) (explaining that historical intoxication laws support only “banning the carry of firearms while actively intoxicated”).

By disarming those who may use drugs from time to time regardless of their present intoxication, § 922(g)(3) “goes much further” than the historical intoxication laws that existed at the Founding. Connelly, 117 F.4th at 281–82. The Supreme Court has warned that “[e]ven when a law regulates arms-bearing for a permissible reason, [] it may not be compatible with the [Second Amendment] if it does so to an extent beyond what was done at the founding.” United States v. Rahimi, 602 U.S. 680, 692 (2024); see also New York State Rifle & Pistol Ass’n v. Bruen, 597 U.S. 1, 26–27 (2022). Yet the court describes this tension as a mere “data point” in its analogical inquiry. Ct. Op. at 36. It explains that the absence of a distinctly similar historical intoxication law could also be evidence that the Founders simply did not “‘maximally exercise their power to regulate.’” Id. (quoting Rahimi, 602 U.S. at 739–40 (Barrett, J., concurring)). One can of course imagine “countless reasons” why a legislature might not regulate in a specific manner. Id. But, in my view, speculation cannot outweigh evidence that when the Founders did act, they regulated in a distinct manner.

Rather than attempt to square § 922(g)(3) with the above principle, the court extracts from historical laws disarming Catholics and loyalists in times of war the Page 77principle that Congress may disarm those it believes to pose a risk of future danger. Id. at 46–56. But Catholics and loyalists were disarmed because they were “seen as potential insurrectionists” in times of war — a classification which we cannot attribute to nonintoxicated marijuana users. Connelly, 117 F.4th at 278; App. at 136–37. Unlike the court, I am not “comfortable” with this level of generality. Ct. Op. at 65. Rather, I agree with the Fifth Circuit that “we must ask: why were the groups disarmed at the Founding considered to be dangerous and therefore disarmed, and is that ‘why’ ‘relevantly similar’ to § 922(g)(3)?” Connelly, 117 F.4th at 278. Because that “why” is not “relevantly similar,” analogizing to these laws fails at Bruen step two. Id.

To be sure, dangerousness is a “touchstone” of disarmament under the Second Amendment. See Pitsilides v. Barr, 128 F.4th 203, 210 (3d Cir. 2025) (quotations omitted). The Supreme Court has not precisely instructed us as to the level of generality at which we may draw principles from historical laws. Ct. Op. at 63–64. And this seems to be a case where “reasonable minds [might] disagree about how broad or narrow the controlling principle should be.” Rahimi, 602 U.S. at 740 (Barrett, J., concurring). But we should resist invoking the principle of dangerousness “at such a high level of generality that it waters down the right” to bear arms. Id.

The elephant in the room here is that marijuana, despite being a Schedule I controlled substance under federal law, is “legal” to some extent in most states. See State Medical Cannabis Laws, Nat’l Conf. of State Legislatures (June 27, 2025), https://www.ncsl.org/health/state-medical-cannabis-laws. The regulatory landscape in this circuit alone is a patchwork. See id. The district court aptly observed that “[t]here Page 78are likely nearly 400,000 Oklahomans who use marijuana under state-law authorization.” App. at 126. Mr. Harrison himself even told the officer who pulled him over that he was on his way to work at a medical marijuana dispensary. Id. at 87. I do not read Bruen to endorse analogical reasoning which effectively writes Congress a “blank check” to disarm so many Americans, many of whom may be under the assumption that marijuana laws have been reformed. 597 U.S. at 30; see also United States v. Harris, 144 F.4th 154, 177 (3d Cir. 2025) (Ambro, J., concurring in part and dissenting in part) (warning against reasoning that “authorizes legislatures to suspend the constitutional rights of so many for such common behavior”).

III. The court’s remand allows the government to assert untimely arguments.

Finally, the scope of the court’s direction to the district court on remand is problematic. The court remands for the district court to “inquire into whether nonintoxicated marijuana users pose a risk of future danger.” Ct. Op. at 68. How the government will make this showing on remand remains to be seen. See id. at 73 n.27. Regardless, this case has never been about the future dangers posed by Mr. Harrison’s marijuana use — but the court’s remand makes it so.

This opens the door to factfinding on a forfeited issue. Mr. Harrison does not dispute that he is a marijuana user.² Oral Arg. at 18:10-18:23. And the government has Page 79prosecuted Mr. Harrison from the position that, to obtain a conviction under § 922(g)(3), it needed only show that he was a “user” of marijuana contemporaneous with his firearm possession. See App. 123; Aplt. Supp. Br. at 9. At the district court, the government never sought to introduce evidence that Mr. Harrison’s marijuana use makes him a danger to others, it never argued that such evidence was necessary to obtain a conviction, and it never so much as requested an evidentiary hearing on the issue. App. 55–65; Oral Arg. at 10:50–11:15. Thus, any evidence about Mr. Harrison’s actual marijuana use beyond the fact that he is a “user” comes too late in this as-applied challenge. The government should be left with the record that it made (or rather neglected to make) and it should not be given a fresh start in its prosecution of Mr. Harrison. See United States v. Malone, 937 F.3d 1325, 1327 (10th Cir. 2019).

The same is true about evidence which the government might introduce to show that non-intoxicated marijuana users categorically pose a risk of future danger. At best, the government analogized “habitual” drug users to the mentally ill and pointed to pre-Bruen cases which reasoned that drug users are more likely than non-users to engage in gun violence. App. 63–64, 124. I believe that the district court was correct to reject the government’s invitation to strip away Mr. Harrison’s Second Amendment rights based only on abstract statistics and “projected” dangerousness. Id. at 124 (quotations omitted). Remanding for factfinding on these issues puts a tremendous burden on Mr. Harrison, one which is inappropriate given that it is the government’s burden to justify § 922(g)(3)’s application in response to his Second Amendment challenge. Rahimi, 602 U.S. at 691. To do so — especially given the government’s failure to introduce Page 80evidence regarding Mr. Harrison’s actual marijuana use — more closely resembles the prior means-end scrutiny than it does the Bruen framework.

In sum, though I commend the scholarship of the court’s opinion, I cannot agree with its reading of the historical record. I think the Fifth Circuit got it right in holding that § 922(g)(3), as applied to a non-intoxicated marijuana user, goes much further than our history and tradition allow. And the scope of the court’s remand allows the government to assert untimely arguments. Because I would affirm the district court’s dismissal of the indictment against Mr. Harrison, I respectfully dissent.