Steven Skoien has two convictions for “misdemeanor crime[s] of domestic violence” and therefore is forbidden to carry firearms in or affecting interstate commerce. 18 U.S.C. § 922(g)(9). Wisconsin informed Skoien about this rule; he signed an acknowledgment of the firearms disability. While he was on probation from the second of his domestic-violence convictions, he was found in possession of three firearms: a pistol, a rifle, and a shotgun. He pleaded guilty to violating § 922(g)(9) by possessing the shotgun and was sentenced to two years’ imprisonment. His conditional guilty plea, see Fed.R.Crim.P. 11(a)(2), reserves the right to contend that § 922(g)(9) violates the Constitution’s Second Amendment. We heard this appeal en banc to decide whether § 922(g)(9) comports with that amendment, as interpreted in District of Columbia v. Heller, — U.S. -, 128 S.Ct. 2783, 171 L.Ed.2d 637 (2008). The eleventh circuit has held that it does. United States v. White, 593 F.3d 1199, 1205-06 (11th Cir. 2010). The fourth circuit has implied otherwise, though in a non-precedential order. United States v. Chester, 367 Fed.Appx. 392 (4th Cir. 2010).

Heller concludes that the Second Amendment “protects the right to keep and bear arms for the purpose of self-defense” and that a law “that banned the possession of handguns in the home” violates that right. McDonald v. Chicago, -U.S.-, 130 S.Ct. 3020, 3021, 177 L.Ed.2d 894 (2010). The United States submits that, before considering how the amendment applies to shotguns and hunting (which is how Skoien contends he used that weapon), we must decide whether Congress is entitled to adopt categorical disqualifications such as § 922(g)(9). The prosecutor relies on this passage from Heller:

Like most rights, the right secured by the Second Amendment is not unlimited.... Although we do not undertake an exhaustive historical analysis today of the full scope of the Second Amendment, nothing in our opinion should be taken to cast doubt on longstanding prohibitions on the possession of firearms by felons and the mentally ill, or laws forbidding the carrying of firearms in sensitive places such as schools and government buildings, or laws imposing conditions and qualifications on the commercial sale of arms.26

128 S.Ct. at 2816-17, reiterated by McDonald, at 3063 (plurality opinion). To this Skoien replies that his prior offenses were misdemeanors rather than felonies, 'and that § 922(g)(9) is not a “longstanding” prohibition, having been enacted in 1996. See United States v. Hayes, — U.S.-, 129 S.Ct. 1079, 172 L.Ed.2d 816 (2009) (discussing its genesis). The prosecutor rejoins by noting that the Court stated its holding this way:

[W]e hold that the District’s ban on handgun possession in the home violates the Second Amendment, as does its prohibition against rendering any lawful firearm in the home operable for the purpose of immediate self-defense. Assuming that Heller is not disqualified from the exercise of Second Amendment rights, the District must permit him to register his handgun and must issue him a license to carry it in the home.

128 S.Ct. at 2821-22. The reference to being “disqualified” relates to prior convictions and mental illness. Id. at 2819. Heller also observes that the Second Amendment “elevates above all other interests the right of law-abiding, responsible citizens to use arms in defense of hearth and home.” Id. at 2821. People *640convicted of domestic violence are neither law-abiding nor responsible, the prosecutor contends. ■

We do not think it profitable to parse these passages of Heller as if they contained an answer to the question whether § 922(g)(9) is valid. They are precautionary language. Instead of resolving questions such as the one we must confront, the Justices have told us that the matters have been left open. The language we have quoted warns readers not to treat Heller as containing broader holdings than the Court set out to establish: that the Second Amendment creates individual rights, one of which is keeping operable handguns at home for self-defense. What other entitlements the Second Amendment creates, and what regulations legislatures may establish, were left open. The opinion is not a comprehensive code; it is just an explanation for the Court’s disposition. Judicial opinions must not be confused with statutes, and general expressions must be read in light of the subject under consideration. See Zenith Radio Corp. v. United States, 437 U.S. 443, 462, 98 S.Ct. 2441, 57 L.Ed.2d 337 (1978).

Although the passages we have quoted are not dispositive, they are informative. They tell us that statutory prohibitions on the possession of weapons by some persons are proper — -and, importantly for current purposes, that the legislative role did not end in 1791. That some categorical limits are proper is part of the original meaning, leaving to the people’s elected representatives the filling in of details. Heller identified, 128 S.Ct. at 2804, as a “highly influential” “precursor” to the Second Amendment the Address and Reasons of Dissent of the Minority of the Convention of the State of Pennsylvania to Their Constituents. (This report is reprinted in Bernard Schwartz, 2 The Bill of Rights: A Documentary History 662, 665 (1971).) The report asserted that citizens have a personal right to bear arms “unless for crimes committed, or real danger of public injury”. Many of the states, whose own constitutions entitled their citizens to be armed, did not extend this right to persons convicted of crime. See Stephen P. Hal-brook, The Founders’ Second Amendment 273 (2008) (concluding that this limitation was understood in the eighteenth century even when not stated expressly in the constitutional text); C. Kevin Marshall, Why Can’t Martha Stewart Have a Gun?, 32 Harv. J.L. & Pub. Policy 695, 700-13 (2009) (surveying the history of state laws limiting convicts’ entitlement to possess firearms). See also United States v. McCane, 573 F.3d 1037, 1047-50 (10th Cir. 2009) (Tymkovich, J., concurring).

The first federal statute disqualifying felons from possessing firearms was not enacted until 1938; it also disqualified misdemeanants who had been convicted of violent offenses. Federal Firearms Act, c. 850, § 2(f), 52 Stat. 1250, 1251. (Technically the crime was “receipt” of a gun that had crossed state lines; the statute treated possession as evidence of receipt.) A 1938 law may be “longstanding” from the perspective of 2008, when Heller was decided, but 1938 is 147 years after the states ratified the Second Amendment. The Federal Firearms Act covered only a few violent offenses; the ban on possession by all felons was not enacted until 1961. Pub.L. 87-342, 75 Stat. 757 (extending the disqualification to all persons convicted of any “crime punishable by imprisonment for a term exceeding one year”, the current federal definition of a “felony”). In 1968 Congress changed the “receipt” element of the 1938 law to “possession,” giving 18 U.S.C. § 922(g)(1) its current form. If such a recent extension of the disqualification to non-violent felons (embezzlers and tax evaders, for example) is presumptively constitutional, as Heller said in note 26, it is difficult to condemn § 922(g)(9), which like the 1938 Act is limited to violent *641crimes. It would be weird to say that § 922(g)(9) is unconstitutional in 2010 but will become constitutional by 2043, when it will be as “longstanding” as § 922(g)(1) was when the Court decided Heller. Moreover, legal limits on the possession of firearms by the mentally ill also are of 20th Century vintage; § 922(g)(4), which forbids possession by a person “who has been adjudicated as a mental defective or who has been committed to a mental institution”, was not enacted until 1968. Pub.L. 90-618, 82 Stat. 1213, 1220.

So although the Justices have not established that any particular statute is valid, we do take from Heller the message that exclusions need not mirror limits that were on the books in 1791. This is the sort of message that, whether or not technically dictum, a court of appeals must respect, given the Supreme Court’s entitlement to speak through its opinions as well as through its technical holdings. See United States v. Bloom, 149 F.3d 649, 653 (7th Cir. 1998). This means that some categorical disqualifications are permissible: Congress is not limited to case-by-case exclusions of persons who have been shown to be untrustworthy with weapons, nor need these limits be established by evidence presented in court. Heller did not suggest that disqualifications would be effective only if the statute’s benefits are first established by admissible evidence.

Categorical limits on the possession of firearms would not be a constitutional anomaly. Think of the First Amendment, which has long had categorical limits: obscenity, defamation, incitement to crime, and others. See United States v. Stevens, -U.S.-,-, 130 S.Ct. 1577, 1584, 176 L.Ed.2d 435 (2010). These categories are not restricted to those recognized in 1791, when the states approved the Bill of Rights. The Justices have held that legislatures may add child pornography to the list, even though the materials do not meet the historical definition of obscenity. New York v. Ferber, 458 U.S. 747, 102 S.Ct. 3348, 73 L.Ed.2d 1113 (1982). More recently, the Court held that speech as part of a public employee’s job is categorically outside the First Amendment. Garcetti v. Ceballos, 547 U.S. 410, 126 S.Ct. 1951, 164 L.Ed.2d 689 (2006). There are other categories, which we need not discuss. See generally John Hart Ely, Flag Desecration: A Case Study in the Roles of Categorization and Balancing in First Amendment Analysis, 88 Harv. L.Rev. 1482 (1975). Neither Ferber nor Garcetti conditioned the categorical limit on proof, satisfactory to a court, that the exclusion was vital to the public safety.

We do not mean that a categorical limit on the possession of firearms can be justified under the rational-basis test, which deems a law valid if any justification for it may be imagined. E.g., Vance v. Bradley, 440 U.S. 93, 99 S.Ct. 939, 59 L.Ed.2d 171 (1979). If a rational basis were enough, the Second Amendment would not do anything, see Heller, 128 S.Ct. at 2817-18 n. 27 — because a rational basis is essential for legislation in general. The Court avoided deciding in Stevens how great the public interest must be to adopt a new categorical limit on speech — the United States had argued for treating depictions of extreme animal cruelty the same as child pornography — but stated that the showing must be strong. The United States concedes that some form of strong showing (“intermediate scrutiny,” many opinions say) is essential, and that § 922(g)(9) is valid only if substantially related to an important governmental objective. See Buckley v. American Constitutional Law Foundation, Inc., 525 U.S. 182, 202-04, 119 S.Ct. 636,142 L.Ed.2d 599 (1999) (using this formula for some First Amendment questions); Heckler v. Mathews, 465 U.S. 728, 744-51, 104 S.Ct. 1387, 79 L.Ed.2d 646 (1984) (using this formula *642for statutes that affect marriage and childbearing). The concession is prudent, and we need not get more deeply into the “levels of scrutiny” quagmire, for no one doubts that the goal of § 922(g)(9), preventing armed mayhem, is an important governmental objective. Both logic and data establish a substantial relation between § 922(g)(9) and this objective.

“Misdemeanor crime of domestic violence” is a defined term.

(A) Except as provided in subparagraph (C), the term “misdemeanor crime of domestic violence” means an offense that—

(i) is a misdemeanor under Federal, State, or Tribal law; and

(ii) has, as an element, the use or attempted use of physical force, or the threatened use of a deadly weapon, committed by a current or former spouse, parent, or guardian of the victim, by a person with whom the victim shares a child in common, by a person who is cohabiting with or has cohabited with the victim as a spouse, parent, or guardian, or by a person similarly situated to a spouse, parent, or guardian of the victim.

(B) (i) A person shall not be considered to have been convicted of such an offense for purposes of this chapter, unless—

(I) the person was represented by counsel in the case, or knowingly and intelligently waived the right to counsel in the case; and

(II) in the case of a prosecution for an offense described in this paragraph for which a person was entitled to a jury trial in the jurisdiction in which the case was tried, either

(aa) the case was tried by a jury, or

(bb) the person knowingly and intelligently waived the right to have the case tried by a jury, by guilty plea or otherwise.

(ii) A person shall not be considered to have been convicted of such an offense for purposes of this chapter if the conviction has been expunged or set aside, or is an offense for which the person has been pardoned or has had civil rights restored (if the law of the applicable jurisdiction provides for the loss of civil rights under such an offense) unless the pardon, expungement, or restoration of civil rights expressly provides that the person may not ship, transport, possess, or receive firearms.

18 U.S.C. § 921(a)(33). A “misdemeanor crime of domestic violence” thus is one in which violence (actual or attempted) is an element of the offense; it is not enough if a risky act happens to cause injury. Cf. Begay v. United States, 553 U.S. 137, 128 S.Ct. 1581, 170 L.Ed.2d 490 (2008); Johnson v. United States, — U.S. -, 130 S.Ct. 1265, 176 L.Ed.2d 1 (2010); United States v. Howell, 531 F.3d 621 (8th Cir. 2008) (applying Begay to § 921(a)(33)).

The belief underpinning § 922(g)(9) is that people who have been convicted of violence once — toward a spouse, child, or domestic partner, no less — are likely to use violence again. That’s the justification for keeping firearms out of their hands, for guns are about five times more deadly than knives, given that an attack with some kind of weapon has occurred. See Franklin E. Zimring, Firearms, Violence, and the Potential Impact of Firearms Control, 32 J.L. Med. & Ethics 34 (2004) (collecting studies).

Hayes, which we mentioned above, held that whether a crime is one of “domestic violence” depends on the identity of the victim rather than the elements of the offense. When describing why § 922(g)(9) was enacted, the Court wrote (129 S.Ct. at 1087):

Existing felon-in-possession laws, Congress recognized, were not keeping fire*643arms out of the hands of domestic abusers, because “many people who engage in serious spousal or child abuse ultimately are not charged with or convicted of felonies.” 142 Cong. Ree. 22985 (1996) (statement of Sen. Lautenberg). By extending the federal firearm prohibition to persons convicted of “misdemeanor crime[s] of domestic violence,” proponents of § 922(g)(9) sought to “close this dangerous loophole.” Id., at 22986.

Construing § 922(g)(9) to exclude the domestic abuser convicted under a generic use-of-force statute (one that does not designate a domestic relationship as an element of the offense) would frustrate Congress’ manifest purpose. Firearms and domestic strife are a potentially deadly combination nationwide.

There are three propositions in this passage: first that domestic abusers often commit acts that would be charged as felonies if the victim were a stranger, but that are charged as misdemeanors because the victim is a relative (implying that the perpetrators are as dangerous as felons); second that firearms are deadly in domestic strife; and third that persons convicted of domestic violence are likely to offend again, so that keeping the most lethal weapon out of their hands is vital to the safety of their relatives. Data support all three of these propositions.

Start with prosecuting domestic violence as a misdemeanor when similar acts against a stranger would be a felony (a practice often called “undercharging”). Prosecutors face two major obstacles to obtaining felony convictions: some family members are willing to forgive the aggressors in order to restore harmonious relations, while others are so terrified that they doubt the ability of the police to protect their safety. Either way, victims of domestic violence are less willing to cooperate with prosecutors, who may need to reduce charges to obtain even limited cooperation and thus some convictions. See Eve S. Buzawa & Carl G. Buzawa, Domestic Violence: The Criminal Justice Response 177-89 (3d ed. 2002). Indeed, either forgiveness or fear induces many victims not to report the attack to begin with. The result is that many aggressors end up with no conviction, or a misdemeanor conviction, when similar violence against a stranger would produce a felony conviction. See, e.g., Report of the Florida Supreme Court Gender Bias Study Commission, reprinted in 42 Fla. L.Rev. 803, 859-60 (1990); Sarah Eaton & Ariella Hyman, The Domestic Violence Component of the New York Task Force Report on Women in the Courts: An Evaluation and Assessment of New York City Courts, 19 Ford-ham Urb. L.J. 391, 461-62 (1992); Patrick A. Langan & Christopher A. Innes, Preventing Domestic Violence Against Women 2 (Bureau of Justice Statistics 1986).

That firearms cause injury or death in domestic situations also has been established. Domestic assaults with firearms are approximately twelve times more likely to end in the victim’s death than are assaults by knives or fists. Linda E. Saltzman, James A. Mercy, Patrick W. O’Carroll, Mark L. Rosenberg & Philip H. Rhodes, Weapon Involvement and Injury Outcomes in Family and Intimate Assaults, 267 J. Am. Medical Ass’n 3043 (1992). Part of this effect stems from the fact that some would-be abusers go buy a gun, see Susan B. Sorenson & Douglas J. Wiebe, Weapons in the Lives of Battered Women, 94 Am. J. Pub. Health 1412 (2004), and much from the fact that guns are more lethal than knives and clubs once an attack begins. See Zimring, Firearms & Violence, supra. The presence of a gun in the home of a convicted domestic abuser is “strongly and independently associated with an increased risk of homicide.” Arthur L. Kellermann, et al., Gun Ownership as a Risk Factor for Homicide in the *644Home, 329 New England J. Medicine 1084, 1087 (1993). See also, e.g., Jacquelyn C. Campbell, et al., Risk Factors for Femicide in Abusive Relationships: Results from a Midtisite Case Control Study, 93 Am. J. Pub. Health 1089, 1090 (2003); James E. Bailey, et al, Risk Factors for Violent Death of Women in the Home, 157 Archives of Internal Medicine 777 (1997); Douglas J. Wiebe, Homicide and Suicide Risks Associated with Firearms in the Home: A National Case-Control Study, 41 Annals of Emergency Medicine 771 (2003). And for this purpose the victims include police as well as spouses, children, and intimate partners. Responding to a domestic-disturbance call is among an officer’s most risky duties. Approximately 8% of officers’ fatalities from illegal conduct during 1999 through 2008 arose from attempts to control domestic disturbances. FBI, Law Enforcement Officers Killed and Assaulted 2008 Table 19 (2009).

Finally, the recidivism rate is high, implying that there are substantial benefits in keeping the most deadly weapons out of the hands of domestic abusers. For example, a study of persons arrested for misdemeanor domestic violence in Cincinnati concluded that 17% of those who remained in the area were arrested again for domestic violence within three years. John Wooldredge & Amy Thistlethwaite, Reconsidering Domestic Violence Recidivism: Individual and Contextual Effects of Court Dispositions and Stake in Conformity vi (1999). The full recidivism rate includes violence that does not lead to an arrest. Estimates of this rate come from survey research and range from 40% to 80% “when victims are followed longitudinally and interviewed directly.” Carla Smith Stover, Domestic Violence Research, 20 J. Interpersonal Violence 448, 450 (2005). See also Julia C. Babcock, et al., Does Batterers’ Treatment Work? A Meta-Analytic Review of Domestic Violence Treatment, 23 Clinical Psychology Rev. 1023, 1039 (2004) (estimating a 35% recidivism rate based on partners’ reports). Skoien cites, as if it were favorable, a study showing that within three years of conviction 48% of domestic abusers “suspended” their abusive conduct— which means that the other 52% did not, and that even the 48% may have committed new crimes within three years after conviction. John H. Laub & Robert J. Sampson, Understanding Desistance from Crime, 28 Crime & Justice 1, 31 (2001). No matter how you slice these numbers, people convicted of domestic violence remain dangerous to their spouses and partners.

By the time this appeal reached oral argument en banc, Skoien’s principal argument had shifted. Instead of denying the logical and empirical basis of § 922(g)(9), he contended that Congress overreached by creating a “perpetual” disqualification for persons convicted of domestic violence. This goes too far, according to Skoien, because the propensity for violence declines with advancing age, and people who are not convicted of additional offenses have demonstrated that they no longer pose risks to other members of their households. Applying § 922(g)(9) to older persons who have not been in legal trouble for many years cannot be substantially related to an important governmental objective, the argument concludes.

Although the statute provides that ex-pungement, pardon, or restoration of civil rights means that a conviction no longer disqualifies a person from possessing firearms, see 18 U.S.C. § 921(a)(33)(B)(ii), Skoien maintains that, as a practical matter, these routes to restoration are unavailable to domestic-battery misdemeanants in Wisconsin. We have our doubts. As the Supreme Court observed in Logan v. United States, 552 U.S. 23, 128 S.Ct. 475, 169 L.Ed.2d 432 (2007), although Wisconsin *645does not deprive misdemeanants of the civil rights to vote, serve on a jury, or hold public office — so these rights cannot be “restored” by the passage of time, as felons’ rights often are — the state does give misdemeanants an opportunity to seek pardon or expungement. Some of the largest states make expungement available as of right to misdemeanants who have a clean record for a specified time. California, for example, has such a program. Cal.Penal Code § 1203.4a. See also Robert A. Mikos, Enforcing State Law in Congress’s Shadow, 90 Cornell L.Rev. 1411, 1463-64 & nn. 187, 188 (2005) (finding that expungement increased following the enactment of § 922(g)(9)). This means that § 922(g)(9) in its normal application does not create a perpetual and unjustified disqualification for a person who no longer is apt to attack other members of the household. True, the statute tolerates different outcomes for persons convicted in different states, but this is true of all situations in which a firearms disability (or any other adverse consequence) depends on state law. The Justices held in Logan that this variability does not call into question federal firearms limits based on state convictions that have been left in place under the states’ widely disparate approaches to restoring civil rights.

But let us assume that the effect of § 922(g)(9) should be assessed state by state, rather than for the nation as a whole. The fact remains that Skoien is poorly situated to contend that the statute creates a lifetime ban for someone who does not pose any risk of further offenses. First, Skoien is himself a recidivist, having been convicted twice of domestic battery. The first victim (in 2003) was his wife; after that marriage ended, the second victim (in 2006) was jhis new fiancée. And Skoien was arrested for possessing multiple guns just one year after that second conviction — while he was still on probation.

A person to whom a statute properly applies can’t obtain relief based on arguments that a differently situated person might present. See United States v. Salerno, 481 U.S. 739, 745, 107 S.Ct. 2095, 95 L.Ed.2d 697 (1987). Although the Salerno principle has been controversial, and the Justices have allowed “over-breadth” arguments when dealing with laws that restrict speech and reach substantially more conduct than the justifications advanced for the statute support, see Stevens, 130 S.Ct. at 1587, the Court has continued to cite Salerno favorably in other situations. See, e.g., Washington State Grange v. Washington State Republican Party, 552 U.S. 442, 449-50, 128 S.Ct. 1184, 170 L.Ed.2d 151 (2008); cf. Gonzales v. Carkart, 550 U.S. 124, 167-68, 127 S.Ct. 1610, 167 L.Ed.2d 480 (2007) (observing that “facial” challenges to statutes generally are restricted to litigation under the First Amendment). If convictions may be used to limit where sex offenders can live (and whether they must register), see Connecticut Department of Public Safety v. Doe, 538 U.S. 1, 123 S.Ct. 1160, 155 L.Ed.2d 98 (2003), a disqualification-on-conviction statute such as § 922(g)(9) also is generally proper. Whether a misdemeanant who has been law abiding for an extended period must be allowed to carry guns again, even if he cannot satisfy § 921(a)(33)(B)(ii), is a question not presented today. There will be time enough to consider that subject when it arises.

Affirmed

dissenting.

Steven Skoien was indicted under 18 U.S.C. § 922(g)(9) for possessing a hunting shotgun after he was convicted of a misdemeanor crime of domestic violence. He argued in the district court and reiterated here that applying the statute to his possession of a long gun for hunting violat*646ed his Second Amendment right to keep and bear arms as explained in District of Columbia v. Heller, — U.S. -, 128 S.Ct. 2783, 171 L.Ed.2d 637 (2008).1 The government invoked Heller’s anticipatory language about certain “presumptively lawful” firearms regulations — specifically, felon-dispossession laws — as a sort of “safe harbor” for analogous prohibitions. Heller, 128 S.Ct. at 2816-17 & n. 26; Brannon P. Denning & Glenn H. Reynolds, Heller, High Water(mark)? Lower Courts and the New Right to Keep and Bear Arms, 60 Hastings L.J. 1245, 1250 (2009) (noting that Heller’s dicta about “presumptively lawful” exceptions to the Second Amendment right may have opened a “safe harbor” for a wide swath of firearms regulation). This approach fell far short of the legal heavy lifting normally required to justify criminally punishing the exercise of an enumerated constitutional right.

The now-vacated panel opinion rejected the government’s argument and instead read Heller's holdings in light of its limiting language about exceptions, distilling a decision method focused first on a textual and historical inquiry into “the terms of the [Second Amendment] right as publicly understood when the Bill of Rights was ratified,” and then — if this inquiry didn’t resolve the case — an application of a degree of heightened judicial review appropriate to the nature of the challenged law’s burden on the right. United States v. Skoien, 587 F.3d 803, 809 (7th Cir. 2009). Because the government hadn’t argued that domestic-violence misdemeanants were excluded from the scope of the Second Amendment right as a textual-historical matter, we assumed Skoien’s Second Amendment rights were intact. Id. at 810. Nor had the government tried to establish a strong relationship between the important governmental objective of reducing firearm violence against domestic intimates and § 922(g)(9)’s permanent disarmament of domestic-violence misdemeanants like Skoien. Id. at 814. So we vacated Skoien’s conviction and remanded for application of intermediate scrutiny on an appropriately developed record. Id. at 815-16.

The en banc court now performs the analysis that would have occurred on remand had we not reheard this case. That’s understandable, I suppose, given the considerable shift in the government’s approach before the en banc court. My colleagues flag the recalibration in Skoien’s argument on rehearing, Majority Op. at 644-45, but it is just as important— even more so, I think — that the government has (belatedly) developed arguments about the original meaning of the Second Amendment right and the means-end justification for § 922(g)(9). The government argued on rehearing that domestic-violence misdemeanants are excluded from the scope of the Second Amendment right as it was originally understood, or if they are not, that § 922(g)(9) survives intermediate scrutiny, at least as applied to Skoien.

My colleagues discuss but do not decide the scope question and avoid the standard-of-review “quagmire” by simply accepting the government’s “concession” that “some form of strong showing (‘intermediate scrutiny,’ many opinions say) is essential, and that § 922(g)(9) is valid only if substantially related to an important governmental objective.” Majority Op. at 641. When it comes to applying this standard, they give the government a decisive assist; most of the empirical data cited to sustain § 922(g)(9) has been supplied by the court. *647This is an odd way to put the government to its burden of justifying a law that prohibits the exercise of a constitutional right. With respect, I cannot join the en banc opinion. The court declines to be explicit about its decision method, sends doctrinal signals that confuse rather than clarify, and develops its own record to support the government’s application of § 922(g)(9) to this defendant.

My colleagues start with an incomplete reading of the Supreme Court’s opinion in Heller. They say the Court held only that “the Second Amendment creates individual rights, one of which is keeping operable handguns at home for self-defense. What other entitlements the Second Amendment creates, and what regulations legislatures may establish, were left open. The opinion is not a comprehensive code; it is just an explanation for the Court’s disposition.” Majority Op. at 640.

I appreciate the minimalist impulse, but this characterization of Heller is hardly fair. It ignores the Court’s extensive analysis of the original public meaning of the Second Amendment and understates the opinion’s central holdings: that the Amendment secures (not “creates”) an individual natural right of armed defense not limited to militia service, Heller, 128 S.Ct. at 2801, and at the core of this guarantee is the right to keep and bear arms for defense of self, family, and home, id. at 2797-99; see also id. at 2817 (invalidating the District of Columbia’s ban on handgun possession because “the inherent right of self-defense has been central to the Second Amendment right” and the D.C. handgun ban “extends ... to the home, where the need for defense of self, family, and property is most acute”); McDonald v. City of Chicago, — U.S. -, 130 S.Ct. 3020, 3023, 177 L.Ed.2d 894 (2010) (plurality opinion) (The “central holding in Heller ” is “that the Second Amendment protects a personal right to keep and bear arms for lawful purposes, most notably for self-defense within the home.”). Heller was “the biggest Second Amendment ease ever decided,” 2 a “landmark ruling [that] merits our attention for its method as well as its result,”3 “the most extensive consideration of the Second Amendment by the Supreme Court in its history,”4 and “the most explicitly and self-consciously originalist opinion in the history of the Supreme Court.”5 It is true that Heller left many issues open, but that is not an invitation to marginalize the Court’s holdings or disregard its decision method.

The en banc court reads Hellers reference to exceptions as a warning not to apply the opinion too broadly. Fair enough. This “precautionary language”— especially the inclusion of felon-disqualification laws on the list of “presumptively lawful” firearms regulations — is “informative” but not “dispositive,” and conveys a message that “whether or not technically dictum, a court of appeals must respect.” Majority Op. at 639-40, 641. I agree, and all the more so after McDonald.6 130 S.Ct. at 3047 (plurality opinion) (reiterating the presumptive validity of certain “longstanding regulatory measures”). But *648my colleagues are not clear about how this limiting dicta should inform the constitutional analysis. The court thinks it “not ... profitable to parse these passages of Heller as if they contained an answer to the question whether § 922(g)(9) is valid,” Majority Op. at 640, but proceeds to parse the passages anyway. My colleagues read Heller’s dicta to mean that “statutory prohibitions on the possession of weapons by some persons are proper — and, importantly for current purposes, that the legislative role did not end in 1791. That some categorical limits are proper is part of the original meaning, leaving to the people’s elected representatives the filling in of details.” Id.

There are several problems with this analysis. First, no one has suggested that the legislative role ended in 1791; the pertinent question is how contemporary gun laws should be evaluated to determine whether they infringe the Second Amendment right. More significantly, that “categorical” disarmament is “proper” as “part of the original meaning” of the Second Amendment has not been established. Heller certainly did not say this; its reference to exceptions was — and remains— unexplained. See Heller, 128 S.Ct. at 2821 (“[Sjince this case represents this Court’s first in-depth examination of the Second Amendment, one should not expect it to clarify the entire field.... [TJhere will be time enough to expound upon ... the exceptions we have mentioned....”); McDonald, at 3047 (plurality opinion) (repeating, without more, Heller’s “assurances” about exceptions).

My colleagues imply that the original meaning of the Second Amendment right excluded persons convicted of a crime, citing the Minority Report of the Pennsylvania ratifying convention, in which Pennsylvania’s dissenting Anti-Federalists proposed amendments to the new Constitution for the protection of individual rights, including the right to bear arms. Majority Op. at 640; see also Stephen P. Halbrook, The Founders’ Second Amendment: Origins of the Right to Bear Arms 195 (2008). The Pennsylvania dissenters proposed an amendment guaranteeing that “the people have a right to bear arms for defence of themselves and their own State or the United States, or for the purpose of killing game,” and providing that “no law shall be passed for disarming the people or any of them unless for crimes committed, or real danger of public injury from individuals.” Bernard Schwartz, 2 The Bill of Rights: A Documentary History 662, 665 (1971). It is true that Heller identified this report as “highly influential” in the run-up to the Second Amendment, but it did so in the context of concluding that the Amendment codified an individual right not limited to militia service. 128 S.Ct. at 2804. There is no reference in Heller to the “unless” clause in the Pennsylvania dissenters’ proposal, and needless to say, this limiting language did not find its way into the Second Amendment.

The court also asserts that “[m]any of the states, whose own constitutions entitled their citizens to be armed, did not extend this right to persons convicted of crime.” Majority Op. at 640. This is a considerable overstatement. Only four state constitutions had what might be considered Second Amendment analogues in 1791 — Massachusetts, North Carolina, Pennsylvania, and Vermont — and none of these provisions excluded persons convicted of a crime. See Heller, 128 S.Ct. at 2802-03; McDonald, at 3037; see also Eugene Volokh, State Constitutional Rights to Keep and Bear Arms, 11 Tex. Rev. L. & Pol. 191, 197-204 (2006); Amar, supra note 2, at 172-73 & nn. 101-02. The sources cited by the court for this very broad proposition simply do not bear it out. To the contrary, two of the cited *649works specifically emphasize the lack of founding-era evidence that persons convicted of a crime were categorically excluded from possessing firearms. See C. Kevin Marshall, Why Can’t Martha Stewart Have a Gun?, 32 Harv. J.L. & Pub. Pol’y 695, 728-35 (2009); United States v. McCane, 573 F.3d 1037, 1047-50 (10th Cir. 2009) (Tymkovich, J., concurring). The third, Stephen Halbrook’s The Founders’ Second Amendment, does not support the court’s suggestion that “many” states during the founding period imposed a general firearms disability on anyone convicted of a crime.7

Regardless, the court hazards these historical observations but ultimately leaves the matter unresolved, moving on to compare categorical limits on firearms possession to categorical limits on the freedom of speech: “obscenity, defamation, incitement to crime, and others.”8 Majority Op. at 641. Adapting First Amendment doctrine to the Second Amendment context is sensible in some cases; indeed, Heller expressly approved the comparison of the Second Amendment to the First. See 128 S.Ct. at 2799, 2821; see also Eugene Volokh, Implementing the Right To Keep and Bear Anns for Self-Defense: An Analytical Framework and a Research Agenda, 56 UCLA L.Rev. 1443, 1449, 1452, 1454-55 (2009); Nelson Lund, The Second Amendment, Heller, and Originalist Jurisprudence, 56 UCLA L.Rev. 1343, 1375-76 (2009) . But this particular First Amendment analogy doesn’t work here. Obscenity, defamation, incitement, and so on are among the few “ ‘well-defined and narrowly limited classes of speech, the prevention and punishment of which have never been thought to raise any Constitutional problem.’ ” United States v. Stevens, — U.S. -, 130 S.Ct. 1577, 1584, 176 L.Ed.2d 435 (2010) (quoting Chaplinsky v. New Hampshire, 315 U.S. 568, 571-72, 62 S.Ct. 766, 86 L.Ed. 1031 (1942)). These “historic and traditional categories [of speech] long familiar to the bar” are “outside the reach of that Amendment altogether — [] they fall into a ‘First Amendment Free Zone.’ ” Id. at 1584-85 (quotation marks omitted).

But my colleagues elide the historical-scope question; they do not decide whether persons convicted of a domestic-violence misdemeanor are completely “outside the reach” of the Second Amendment as a matter of founding-era history and background legal tradition. For this analogy to hold up, the court would have to make a judgment on the matter. Absent that, it’s hard to make sense of the court’s reliance *650on this strain of First Amendment doctrine.9

Moreover, it is one thing to say that certain narrowly limited categories of speech have long been understood to fall outside the boundaries of the free-speech right and are thus unprotected by the First Amendment. It is quite another to say that a certain category of persons has long been understood to fall outside the boundaries of the Second Amendment and thus may be excluded from ever exercising the right. The relevant question for our purposes is whether domestic-violence misdemeanants are wholly unprotected by the Second Amendment. By invoking this line of First Amendment caselaw, my colleagues imply an affirmative answer, but do not see the analysis through.10

The better approach is to acknowledge the limits of the scope inquiry in a more straightforward way: The historical evidence is inconclusive at best. As noted in the panel opinion, scholars disagree about the extent to which felons — let alone misdemeanants — were considered excluded from the right to bear arms during the founding era. Compare, e.g., Marshall, supra, at 714-28 (the founding-era understanding of the right to keep and bear arms for self-defense did not categorically exclude persons convicted of a crime), with Don B. Kates & Clayton E. Cramer, Second Amendment Limitations and Crimi*651nological Considerations, 60 Hastings L.J. 1339, 1359-64 (2009) (the founding generation understood that persons convicted of common-law felonies could be disarmed), and Don B. Kates, Jr., Handgun Prohibition and the Original Meaning of the Second Amendment, 82 Mich. L.Rev. 204, 266 (1983) (same); see also Adam Winkler, Heller’s Catch-22, 56 UCLA L.Rev. 1551, 1562-66 (2009); Lund, supra, at 1356-57; Glenn Harlan Reynolds, A Critical Guide to the Second Amendment, 62 Tenn. L.Rev. 461, 480 (1995) (collecting originalist scholarship). We simply cannot say with any certainty that persons convicted of a domestic-violence misdemeanor are wholly excluded from the Second Amendment right as originally understood.11 Because Skoien is not categorically unprotected, the government’s use of § 922(g)(9) against him must survive Second Amendment scrutiny.12

My colleagues evidently agree; they move on to discuss the standard for determining whether the disarmament of domestic-violence misdemeanants is constitutionally permissible. This inquiry is necessary only if Skoien’s Second Amendment rights are intact notwithstanding his domestic-violence conviction. The court properly concludes that some form of heightened judicial scrutiny is required; rational-basis review has been ruled out. Heller, 128 S.Ct. at 2818 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.”); Majority Op. at 641-42. The court assumes without deciding that intermediate scrutiny applies, and holds that data establish a substantial relationship between § 922(g)(9) and the important governmental objective of “preventing armed mayhem.” Id. What follows is a discussion of the Supreme Court’s decision in United States v. Hayes, — U.S. -, 129 S.Ct. 1079, 172 L.Ed.2d 816 (2009) — in particular, its reference to a statement in the congressional record by the principal Senate sponsor of § 922(g)(9) — and several pages of social-science research on the criminal-justice system’s treatment of domestic-violence cases, firearm violence in the home, and recidivism by domestic-violence offenders. Most of this data, as I have noted, has been supplied by the court.

The court thus accepts that it is the government’s burden to make a “strong *652showing” of the danger-reduction justification for stripping domestic-violence misdemeanants of their Second Amendment rights but in the end makes the case for itself. This relieves the government of its burden and deprives Skoien of the opportunity to review the outcome-determinative evidence, let alone subject it to normal adversarial testing. One obvious peril in this approach: The court’s understanding of the research on domestic violence might be mistaken. That is certainly true of my colleagues’ conclusion that “domestic abusers often commit acts that would be charged as felonies if the victim were a stranger, but that are charged as misdemeanors because the victim is a relative.” Majority Op. at 643. The court has misread the materials it cites for this conclusion, which document the well-recognized difficulty of prosecuting domestic violence because of victim fear or noneooperation but do not establish that acts of domestic violence are “often” chargeable as felonies but for the domestic dynamic. Perhaps the government can discharge its burden in this case, but the place for it to do so in the first instance is in the district court, not the court of appeals.13

The court also dismisses Skoien’s contention that § 922(g)(9) is impermissibly overinclusive because it is a permanent disqualification and provides no effective way for an offender to reacquire his Second Amendment rights. It is true, as the court notes, that a pardon, expungement, or restoration of civil rights will lift the federal firearms ban. See 18 U.S.C. § 921(a)(33)(B)(ii) (excluding domestic-violence convictions that have been pardoned, expunged, or for which civil rights have been restored unless the pardon, expungement, or restoration of rights provides that the person may not possess firearms). But as my colleagues acknowledge, in Wisconsin misdemeanants do not lose their civil rights, and rights not lost cannot be “restored” for purposes of the statutory exception. See Logan v. United States, 552 U.S. 23, 36, 128 S.Ct. 475, 169 L.Ed.2d 432 (2007). The court nonetheless maintains that “the state does give misdemeanants an opportunity to seek pardon or expungement.” Majority Op. at 645. Pardon, yes; expungement, no — at least not in the typical case. In Wisconsin the ex-pungement remedy is extremely narrow; it applies only to misdemeanants under the *653age of 21 and must be ordered at the time of sentencing.14 Wis. Stat. § 973.015(l)(a). There is no after-the-fact or generally available opportunity to seek expungement. It is true that the pardon power is very broad, but I doubt that governors — in Wisconsin or elsewhere — pardon domestic-violence misdemeanants with any regularity. So Skoien is right that the § 922(g)(9) ban is effectively permanent, at least as to him.15

This brings me to the court’s final point: that Skoien is “poorly situated” to complain about the perpetual nature of the § 922(g)(9) ban because he is a recidivist who was caught with “multiple guns just one year after [his] second conviction— while he was still on probation.” 16 Majority Op. at 645. Maybe so. Skoien’s status as a recent domestic-violence recidivist certainly diminishes the force of his argument about the permanent feature of § 922(g)(9) as the statute has been applied to him. The court properly reserves the question whether application of § 922(g)(9) would survive a Second Amendment challenge by “a misdemeanant who has been law abiding for an extended period.” Majority Op. at 645. Still, I think it highly inappropriate for the court to resolve this challenge to the application of the statute without requiring the government to shoulder its burden — and giving Skoien the opportunity to respond — on remand in the district court. The sort of empirical inquiry normally required by intermediate scrutiny should not be performed by the court of appeals in the first instance.17

*654The court thus short-circuits the usual process -and resolves this case on a record of its own creation, prematurely ending Skoien’s challenge and leaving markers for the future that will immunize most applications of § 922(g)(9) from serious Second Amendment scrutiny. This approach is difficult to reconcile with either the reasoning or the result in Heller, though it might be thought consistent with an aggressive reading of the Court’s reference to presumptively lawful firearms regulations. Of course there are several ways to understand the Court’s analysis in Heller in light of its limiting dicta about exceptions.18 But we cannot read Hellers dicta in a way that swallows its holdings. The government normally has the burden of justifying the application of laws that criminalize the exercise of enumerated constitutional rights. We should follow that norm, not pay lip service to it. I would remand for the government to make its own case for imprisoning Steven Skoien for exercising his Second Amendment rights.