OPINION

SHWARTZ, Circuit Judge.

Today we address whether one of New Jersey's responses to the rise in active and mass shooting incidents in the United States — a law that limits the amount of ammunition that may be held in a single firearm magazine to no more than ten rounds — violates the Second Amendment, the Fifth Amendment's Takings Clause, and the Fourteenth Amendment's Equal Protection Clause. We conclude that it does not. New Jersey's law reasonably fits the State's interest in public safety and does not unconstitutionally burden the Second Amendment's right to self-defense in the home. The law also does not violate the Fifth Amendment's Takings Clause because it does not require gun owners to surrender their magazines but instead allows them to retain modified magazines or register firearms that have magazines that cannot be modified. Finally, because retired law enforcement officers have training and experience that makes them different from ordinary citizens, the law's exemption that permits them to possess magazines that can hold more than ten rounds does not violate the Fourteenth Amendment's Equal Protection Clause. We will therefore affirm the District Court's order denying Plaintiffs' motion to preliminarily enjoin enforcement of the law.

I

A

Active shooting and mass shooting incidents have dramatically increased during recent years. Statistics from 2006 to 2015 reveal a 160% increase in mass shootings over the prior decade. App. 1042. Department of Justice and Federal Bureau of Investigation ("FBI") studies of active shooter incidents (where an individual is actively engaged in killing or attempting to kill people with a firearm in a confined, populated area) reveal an increase from an average of 6.4 incidents in 2000 to 16.4 incidents in 2013. App. 950, 953. These numbers have continued to climb, and in 2017, there were thirty incidents. App. 1149, 1133. In addition to becoming more frequent, these shootings have also become more lethal. App. 906-07 (citing 2018 article noting "it's the first time [in American history] we have ever experienced four gun massacres resulting in double-digit fatalities within a 12-month period").

In response to this trend, a number of states have acted. In June 2018, New Jersey became the ninth state to pass a new law restricting magazine capacity.1 New Jersey has made it illegal to possess a magazine capable of holding more than ten rounds of ammunition ("LCM").2 N.J. Stat. Ann. 2C:39-1(y), 2C:39-3(j) ("the Act").

Active law enforcement officers and active military members, who are "authorized to possess and carry a handgun," are excluded from the ban. N.J. Stat. Ann. 2C:39-3(g). Retired law enforcement officers are also exempt and may possess and carry semi-automatic handguns with magazines that hold up to fifteen rounds of ammunition.3 Id. at 2C:39-17.

The Act provides several ways for those who are not exempt from the law to comply. Specifically, the legislation gives LCM owners until December 10, 20184 to (1) modify their LCMs "to accept ten rounds or less," id. at 2C:39-19(b); (2) render firearms with LCMs or the LCM itself inoperable, id.; (3) register firearms with LCMs that cannot be "modified to accommodate ten or less rounds," id. at 2C:39-20(a); (4) transfer the firearm or LCM to an individual or entity entitled to own or possess it, id. at 2C:39-19(a); or (5) surrender the firearm or LCM to law enforcement, id. at 2C:39-19(c).

B

On the day the bill was signed, Plaintiffs Association of New Jersey Rifle and Pistol Clubs and members Blake Ellman and Alexander Dembrowski (collectively, "Plaintiffs")5 filed this action under 42 U.S.C. § 1983, alleging that the Act violates the Second Amendment, the Fifth Amendment's Takings Clause, and the Fourteenth Amendment's Equal Protection Clause. App. 46-64. Plaintiffs also sought a preliminary injunction to enjoin Defendants Attorney General of New Jersey, Superintendent of New Jersey State Police, and the Chiefs of Police of the Chester and Lyndhurst Police Departments from enforcing the law.

The District Court held a three-day evidentiary hearing on the preliminary injunction request. The Court considered declarations from witnesses, which served as their direct testimony, and then these witnesses were thoroughly examined.6 The parties also submitted various documents, including declarations presented in other cases addressing LCM bans, books and journal articles on firearm regulations, reports on the efficacy of the 1994 federal assault weapons ban, statistics about gun ownership and use, news articles about shooting incidents, FBI reports on active shooter incidents, historical materials on LCMs, and police academy training materials.7 The evidence disclosed the purpose of LCMs, how they are used, and who uses them.

A magazine is an implement that increases the ammunition capacity of a firearm. App. 128. An LCM refers to a particular size of magazine. App. 159. LCMs allow a shooter to fire multiple shots in a matter of seconds without reloading. App. 225, 865. Millions of LCMs have been sold since 1994, App. 1266, and 63% of gun owners reported using LCMs in their modern sporting rifles, App. 516, 753. LCMs often come factory standard with semi-automatic weapons. App. 656, 994-95.

Gun owners use LCMs for hunting and pest control. App. 655. LCMs have also been used for self-defense. App. 225, 844-51, 915-16, 1024. The record does not include a reliable estimate of the number of incidents where more than ten shots were used in self-defense,8 but it does show that LCMs "are not necessary or appropriate for self-defense," App. 861, and that use of LCMs in self-defense can result in "indiscriminate firing," App. 863, and "severe adverse consequences for innocent bystanders," App. 1024.

There is also substantial evidence that LCMs have been used in numerous mass shootings,9 App. 851-53, 909-10, 914, 967-88, 1024, 1042, 1057, 1118-26, 1165-71, and that the use of LCMs results in increased fatalities and injuries, App. 562. "[W]hen you have a high capacity magazine it allows you to fire off a large number of bullets in a short amount of time, and that gives individuals much less opportunity to either escape or to try to fight back or for police to intervene; and that is very valuable for mass shooters." App. 225, 865. The record demonstrates that when there are pauses in shooting to reload or for other reasons, opportunities arise for victims to flee, as evidenced by the 2017 Las Vegas and 2013 D.C. Navy Yard shootings, App. 114, 914, 1045, or for bystanders to intervene, as in the 2018 Tennessee Waffle House shooting and 2011 Arizona shooting involving Representative Gabrielle Giffords, App. 830, 1113.

While a trained marksman or professional speed shooter operating in controlled conditions can change a magazine in two to four seconds, App. 109, 263-67, 656, 1027, an inexperienced shooter may need eight to ten seconds to do so, App. 114. Therefore, while a ban on LCMs does not restrict the amount of ammunition or number of magazines an individual may purchase, App. 231, without access to LCMs, a shooter must reload more frequently.

"[S]hooters in at least 71% of mass shootings in the past 35 years obtained their guns legally," App. 853, or from a family member or friend (as was the case with the Newtown shooter who took his mother's lawfully-owned guns), App. 190, 195, 486, and gun owners in lawful possession of firearms are a key source of arming criminals through loss and theft of their firearms, App. 221-22, 800-01, 924-25.

New Jersey law enforcement officers regularly carry LCMs, App. 116, 1102, and along with their retired counterparts, are trained and certified in the use of firearms, App. 143-46, 1101-02. Law enforcement officers use certain firearms not regularly used by members of the military and use them in a civilian, non-combat environment.10 App. 137, 140, 1103.

After carefully considering all of the evidence and the parties' arguments, the District Court denied the motion to preliminarily enjoin the Act. The Court found the expert witnesses were credible but concluded that the testimony of certain experts was "of little help in its analysis .... [because] their testimony failed to clearly convey the effect this law will have on reducing mass shootings in New Jersey or the extent to which the law will impede gun owners from defending themselves." Ass'n of N.J. Rifle & Pistol Clubs, Inc. v. Grewal, No. 3:17-cv-10507, 2018 WL 4688345, at *8 (D.N.J. Sept. 28, 2018). Specifically, the Court stated that although it found both Kleck and Allen credible, their testimony "relied upon questionable data and conflicting studies," suggesting that both of the experts' methodologies and conclusions were flawed.11 Id.

The District Court, however, considered other evidence in the record to reach its conclusion, see, e.g., id. at *6, *6 n.7, *12, that the Act was constitutional. The District Court held that a "ban on magazines capable of holding more than ten rounds implicates Second Amendment protections," id. at *11, but that it does not violate the Second Amendment. Specifically, the District Court held that the Act (1) should be examined under intermediate scrutiny because it "places a minimal burden on lawful gun owners," id. at *13, and (2) "is reasonably tailored to achieve [New Jersey's] goal of reducing the number of casualties and fatalities in a mass shooting," id., based in part on evidence showing that "there is some delay associated with reloading, which may provide an opportunity for potential victims to escape or for a bystander to intercede," id. at *12.

The District Court also held that the Fifth Amendment Takings and Fourteenth Amendment Equal Protection claims lacked merit. The Court concluded that the Takings claim failed because the modification and registration options "provided property owners with ... avenue[s] to comply with the law without forfeiting their property." Id. at *16. The Court also determined that the Act's exemption for retired law enforcement officers did not violate Plaintiffs' right to equal protection because law enforcement officers, in light of their "extensive and stringent training" and experience "confronting unique circumstances that come with being a police officer," are different from, and hence not similarly situated to, other residents. Id. at *14.

After concluding that Plaintiffs failed to demonstrate a likelihood of success on their claims, the District Court stated that Plaintiffs did not satisfy the other requirements for a preliminary injunction, id. at *16, and denied their motion. Plaintiffs appeal.

Plaintiffs do not advocate an absolutist view of the Second Amendment but believe that the State's ability to impose any restriction on magazine capacity is severely limited. Plaintiffs argue that the Act is categorically unconstitutional because it bans an entire class of arms protected by the Second Amendment, there is no empirical evidence supporting the State ban, and the rights of law abiding citizens are infringed and their ability to defend themselves in the home is reduced.

On the other hand, the State asserts that it is imperative to the safety of its citizens to take focused steps to reduce the devastating impact of mass shootings. The State argues that the Act does not hamper or infringe the rights of law abiding citizens who legally possess weapons.

II12

The decision to grant or deny a preliminary injunction is within the sound discretion of the district court. Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 24, 33, 129 S.Ct. 365, 172 L.Ed.2d 249 (2008). "We employ a tripartite standard of review for ... preliminary injunctions. We review the District Court's findings of fact for clear error. Legal conclusions are assessed de novo. The ultimate decision to grant or deny the injunction is reviewed for abuse of discretion." K.A. ex rel. Ayers v. Pocono Mountain Sch. Dist., 710 F.3d 99, 105 (3d Cir.2013) (internal quotation marks and citations omitted).13

Id. (alteration in original) (quoting Tenafly Eruv Ass'n v. Borough of Tenafly, 309 F.3d 144, 157 (3d Cir.2002)); Fed. R. Civ. P. 65. A plaintiff's failure to establish a likelihood of success on the merits "necessarily result[s] in the denial of a preliminary injunction." Am. Express Travel Related Servs., Inc. v. Sidamon-Eristoff, 669 F.3d 359, 366 (3d Cir.2012) (internal quotation marks and citation omitted). On this factor, "a sufficient degree of success for a strong showing exists if there is `a reasonable chance or probability, of winning.'" In re Revel AC, Inc., 802 F.3d 558, 568 (3d Cir.2015) (quoting Singer Mgmt. Consultants, Inc. v. Milgram, 650 F.3d 223, 229 (3d Cir.2011) (en banc)). Here, we must decide whether Plaintiffs have a reasonable probability of showing that the Act violates the Second Amendment, the Fifth Amendment's Takings Clause, and the Fourteenth Amendment's Equal Protection Clause. We consider each claim in turn.

III

The Second Amendment provides: "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. In District of Columbia v. Heller, 554 U.S. 570, 128 S.Ct. 2783, 171 L.Ed.2d 637 (2008), the Supreme Court held that the Second Amendment protects the right of individuals to possess firearms and recognized that the "core" of the Second Amendment is to allow "law-abiding, responsible citizens to use arms in defense of hearth and home." Id. at 628-30, 635, 128 S.Ct. 2783 (invalidating a statute banning the possession of handguns in the home).14

We therefore must first determine whether the regulated item is an arm under the Second Amendment. The law challenged here regulates magazines, and so the question is whether a magazine is an arm under the Second Amendment. The answer is yes. A magazine is a device that holds cartridges or ammunition. "Magazine," Merriam-Webster Dictionary, https://www.merriam-webster.com/dictionary/magazine (last visited Nov. 21, 2018); App. 128 (describing a magazine as "an implement that goes into the weapon to increase the capacity of the weapon itself"). Regulations that eliminate "a person's ability to obtain or use ammunition could thereby make it impossible to use firearms for their core purpose." Jackson v. City & Cty. of San Francisco, 746 F.3d 953, 967 (9th Cir.2014). Because magazines feed ammunition into certain guns, and ammunition is necessary for such a gun to function as intended, magazines are "arms" within the meaning of the Second Amendment. Id.; see also United States v. Miller, 307 U.S. 174, 180, 59 S.Ct. 816, 83 S.Ct. 1206 (1939) (citing 17th century commentary on gun use in America that "[t]he possession of arms also implied the possession of ammunition.").

Having determined that magazines are arms, we next apply a two-step framework to resolve the Second Amendment challenge to a law regulating them. United States v. Marzzarella, 614 F.3d 85, 89 (3d Cir.2010). First, we consider whether the regulation of a specific type of magazine, namely an LCM, "imposes a burden on conduct falling within the scope of the Second Amendment's guarantee." Id. Second, if the law burdens conduct that is protected by the Second Amendment, "we evaluate the law under some form of means-end scrutiny." Id. "If the law passes muster under that standard, it is constitutional. If it fails, it is invalid." Id.

A

Under step one, we consider whether the type of arm at issue is commonly owned,15 Marzzarella, 614 F.3d at 90-91, and "typically possessed by law-abiding citizens for lawful purposes,"16 Heller, 554 U.S. at 625, 128 S.Ct. 2783. The record shows that millions of magazines are owned, App. 516, 753, often come factory standard with semi-automatic weapons, App. 656, are typically possessed by law-abiding citizens for hunting, pest-control, and occasionally self-defense, App. 655, 554-55,17 and there is no longstanding history of LCM regulation.18 We will nonetheless assume without deciding that LCMs are typically possessed by law-abiding citizens for lawful purposes and that they are entitled to Second Amendment protection. See N.Y. State Rifle & Pistol Ass'n, Inc. v. Cuomo, 804 F.3d 242, 257 (2d Cir.2015); Heller v. District of Columbia, 670 F.3d 1244, 1261 (D.C. Cir.2011) [hereinafter Heller II].

B

Assuming that the Act implicates an arm subject to Second Amendment protection, we next address the level of means-end scrutiny that must be applied. Marzzarella, 614 F.3d at 89. The applicable level of scrutiny is dictated by whether the challenged regulation burdens the core Second Amendment right. If the core Second Amendment right is burdened, then strict scrutiny applies; otherwise, intermediate scrutiny applies.19 See Drake v. Filko, 724 F.3d 426, 436 (3d Cir.2013). "At its core, the Second Amendment protects the right of law-abiding citizens to possess non-dangerous weapons for self-defense in the home." Marzzarella, 614 F.3d at 92 (citing Heller, 554 U.S. at 635, 128 S.Ct. 2783); see Drake, 724 F.3d at 431 (declining to definitively hold that Second Amendment core "extends beyond the home"). Thus, laws that severely burden the core Second Amendment right to self-defense in the home are subject to strict scrutiny. Drake, 724 F.3d at 436; Marzzarella, 614 F.3d at 97; see also Kolbe v. Hogan, 849 F.3d 114, 138 (4th Cir.2017) (en banc) (applying intermediate scrutiny where the law "does not severely burden the core protection of the Second Amendment"); N.Y. State Rifle & Pistol Ass'n, 804 F.3d at 260 (applying intermediate scrutiny where "[t]he burden imposed by the challenged legislation is real, but it is not `severe'" (citation omitted)); Fyock v. City of Sunnyvale, 779 F.3d 991, 998-99 (9th Cir.2015) (determining appropriate level of scrutiny by considering "how severely, if at all, the law burdens [the Second Amendment] right"); Heller II, 670 F.3d at 1261 (determining "the appropriate standard of review by assessing how severely the prohibitions burden the Second Amendment right").

1

The Act here does not severely burden the core Second Amendment right to self-defense in the home for five reasons. First, the Act, which prohibits possession of magazines with capacities over ten rounds, does not categorically ban a class of firearms. The ban applies only to magazines capable of holding more than ten rounds and thus restricts "possession of only a subset of magazines that are over a certain capacity." Fyock, 779 F.3d at 999 (describing LCM ban as a restriction); S.F. Veteran Police Officers Ass'n v. City & Cty. of San Francisco, 18 F.Supp.3d 997, 1002-03 (N.D. Cal. 2014) (emphasizing that the law was not "a total ban on all magazines" but "a total ban only on magazines holding more than ten rounds"); see also App. 159 (testimony explicitly addressing that the law "does not ban any particular class of gun" because "it just deals with the size of the magazine").

Second, unlike the ban in Heller, the Act is not "a prohibition of an entire class of `arms' that is overwhelmingly chosen by American society for [self-defense in the home]." 554 U.S. at 628, 128 S.Ct. 2783. The firearm at issue in Heller, a handgun, is one that the Court described as the "quintessential self-defense weapon." Id. at 629, 128 S.Ct. 2783. The record here demonstrates that LCMs are not well-suited for self-defense. App. 225, 861, 863, 915, 1024.

Third, also unlike the handgun ban in Heller, a prohibition on "large-capacity magazines does not effectively disarm individuals or substantially affect their ability to defend themselves." Heller II, 670 F.3d at 1262 (citing Marzzarella, 614 F.3d at 97). Put simply, the Act here does not take firearms out of the hands of law-abiding citizens, which was the result of the law at issue in Heller. The Act allows law-abiding citizens to retain magazines, and it has no impact on the many other firearm options that individuals have to defend themselves in their home.20 Marzzarella, 614 F.3d at 97; App. 230-32, 917-18.

Fourth, the Act does not render the arm at issue here incapable of operating as intended. New Jersey citizens may still possess and utilize magazines, simply with five fewer rounds per magazine. Ass'n of N.J. Rifle & Pistol Clubs, 2018 WL 4688345, at *12; see also N.Y. State Rifle & Pistol Ass'n, 804 F.3d at 260 ("[W]hile citizens may not acquire high-capacity magazines, they can purchase any number of magazines with a capacity of ten or fewer rounds. In sum, numerous alternatives remain for law-abiding citizens to acquire a firearm for self-defense." (internal quotation marks and citation omitted)).

Fifth, "it cannot be the case that possession of a firearm in the home for self-defense is a protected form of possession under all circumstances. By this rationale, any type of firearm possessed in the home would be protected merely because it could be used for self-defense." Marzzarella, 614 F.3d at 94.

For these reasons, while the Act affects a type of magazine one may possess, it does not severely burden, and in fact respects, the core of the Second Amendment right. See N.Y. State Rifle & Pistol Ass'n, 804 F.3d at 258; Marzzarella, 614 F.3d at 94 (observing that machine guns are not protected by the Second Amendment even though they may be used in the home for self-defense). As a result, intermediate scrutiny applies.21

2

"[U]nder intermediate scrutiny[,] the government must assert a significant, substantial, or important interest; there must also be a reasonable fit between that asserted interest and the challenged law, such that the law does not burden more conduct than is reasonably necessary." Drake, 724 F.3d at 436; Marzzarella, 614 F.3d at 98 (requiring serial numbers on guns reasonably fits government interest). The law need not be the least restrictive means of achieving that interest. Drake, 724 F.3d at 439.22

"The State of New Jersey has, undoubtedly, a significant, substantial and important interest in protecting its citizens' safety." Id. at 437. Given the context out of which the Act was enacted, this clearly includes reducing the lethality of active shooter and mass shooting incidents. Thus, the State has asserted a qualifying interest.

New Jersey's LCM ban reasonably fits the State's interest in promoting public safety. LCMs are used in mass shootings. App. 1057 (stating that "LCM firearms are more heavily represented among guns used in murders of police and mass murders"); see App. 269 (noting 23 mass shootings using LCMs), 1118-26 (describing weapons used in sixty-one mass shootings, eleven of which used fifteen-round magazines, two of which used thirteen, and two of which used fourteen round magazines). LCMs allow for more shots to be fired from a single weapon and thus more casualties to occur when they are used. App. 562 (noting, however, that this does not imply that LCMs "caused shooters to inflict more casualties"), 865, 895-98. By prohibiting LCMs, the Act reduces the number of shots that can be fired from one gun, making numerous injuries less likely.

Not only will the LCM ban reduce the number of shots fired and the resulting harm, it will present opportunities for victims to flee and bystanders to intervene. App. 919-20. Reducing the capacity of the magazine to which a shooter has access means that the shooter will have fewer bullets immediately available and will need to either change weapons or reload to continue shooting.23 Weapon changes and reloading result in a pause in shooting and provide an opportunity for bystanders or police to intervene and victims to flee. As the Commissioner of the Baltimore Police Department explained, if a shooter uses a ten-round magazine, rather than a 30, 50, or 100-round magazine, the chances to act increase:

App. 865; see also Kolbe, 849 F.3d at 128 (same).

This view is corroborated by other items in the record demonstrating that a delay occurs when a shooter needs to reload, see App. 114 (eight to ten seconds for inexperienced shooter or two to four seconds for trained shooter), and that such delay can be consequential. Videos from the Las Vegas shooting in 2017 show that "concert attendees would use the pauses in firing when the shooter's high capacity magazines were spent to flee." App. 914. During the Navy Yard shooting, one victim had a chance to escape when the shooter was forced to reload. App. 1045 (describing Navy Yard shooting where shooter attempted to kill a woman, was out of ammunition, and left to reload, at which time she found a new hiding spot and ultimately survived); see also App. 658-59, 1027 (describing escape during reloading in 2012 Newtown shooting). There are multiple instances when individuals have intervened in mass shootings and active shooter incidents to stop the shooter. App. 830 (Waffle House shooting), 969 (Florida's Gold Leaf Nursery shooting where "shooter was restrained by a citizen while attempting to reload his gun"), 1113 (Arizona's Giffords shooting), 1142 (Seattle Pacific University shooting where shooter was confronted/pepper-sprayed by student while reloading). While each incident may not have involved delay due to a need to reload, see App. 282 (distinguishing Waffle House incident on the basis that the intervener "said he didn't know one way or another, and when he was interviewed the first possibility he offered was the guy's — the shooter's gun jammed"), it was the pause in shooting that allowed individuals and bystanders to act. See App. 865, 979, 1142. In light of this evidence, the District Court did not clearly err when it concluded that the evidence "established that there is some delay associated with reloading, which may provide an opportunity for potential victims to escape or for a bystander to intercede and somehow stop a shooter." Ass'n of N.J. Rifle & Pistol Clubs, 2018 WL 4688345, at *12. Therefore, the ban reasonably fits New Jersey's interest.24 See Drake, 724 F.3d at 437.

Plaintiffs attempt to discount the need for the LCM ban by describing mass shootings as rare incidents, and asserting that the LCM ban burdens the rights of law-abiding gun owners to address an infrequent occurrence.25 The evidence adduced before the District Court shows that this statement downplays the significant increase in the frequency and lethality of these incidents. See, e.g., App. 906, 1133-34; see also App. 1042-43 (noting that pre-2015, there was never a year with more than five gun massacres, and 2015 had seven "massacres" as defined by Mother Jones, but acknowledging discrepancies with Mother Jones' definition of massacre or mass shooting). Despite Plaintiffs' assertion to the contrary, New Jersey has not been spared from a mass shooting. Just days after the Act was passed, a mass shooter injured twenty-two individuals and killed one at an arts festival in Trenton. Ass'n of N.J. Rifle & Pistol Clubs, 2018 WL 4688345, at *3; App. 1288-95. Even if this event had not occurred, "New Jersey need not wait for its own high-fatality gun massacre before curtailing access to LCMs." Giffords Law Ctr. Amicus Br. at 3; App. 247.

Lastly, the Act does not burden more conduct than reasonably necessary. As we have already discussed, the prohibition on LCMs does not disarm an individual. While the Act does limit access to one tool — magazines that hold over ten rounds — it imposes no limit on the number of firearms or magazines or amount of ammunition a person may lawfully possess.26 In any event, the record does not show that LCMs are well-suited or safe for self-defense.27 App. 844-51, 861, 863, 923, 1024. Thus, the Act is designed to "remove these especially lethal items from circulation so that they will be unavailable, or at least less available, to mass murderers," S.F. Veteran Police Officers Ass'n, 18 F.Supp.3d at 1004; see also Friedman v. City of Highland Park, 784 F.3d 406, 412 (7th Cir.2015); App. 195, 221-22, 846, 800-01, 853, and it does not burden a gun owner's right to self-defense, Drake, 724 F.3d at 439 (upholding a gun law that "takes into account the individual's right to protect himself from violence as well as the community at large's interest in self-protection" and general public safety).

For these reasons, the Act survives intermediate scrutiny, and like our sister circuits, we hold that laws restricting magazine capacity to ten rounds of ammunition do not violate the Second Amendment.28 See Kolbe, 849 F.3d 114 (upholding Maryland ten round limit); N.Y. State Rifle & Pistol Ass'n, 804 F.3d at 263-64 (upholding New York and Connecticut's ten-round limit); Friedman, 784 F.3d at 411-12 (upholding city's ten-round limit); Fyock, 779 F.3d at 1000 (same)29; Heller II, 670 F.3d at 1262-64 (upholding D.C.'s ten-round limit).30

IV

Plaintiffs' Fifth Amendment Takings claim also fails. The Takings Clause provides that "private property [shall not] be taken for public use, without just compensation."31 U.S. Const. amend. V. "The paradigmatic taking requiring just compensation is a direct government appropriation or physical invasion of private property." Lingle v. Chevron U.S.A., Inc., 544 U.S. 528, 537, 125 S.Ct. 2074, 161 L.Ed.2d 876 (2005). In addition, a government regulation "may, in some instances, be so onerous that its effect is tantamount to a direct appropriation or ouster," and "such `regulatory takings' may be compensable under the Fifth Amendment." Id.

Here, the compliance measures in the Act do not result in either an actual or regulatory taking.32 There is no actual taking because owners have the option to transfer or sell their LCMs to an individual or entity who can lawfully possess LCMs, modify their LCMs to accept fewer than ten rounds, or register those LCMs that cannot be modified. See N.J. Stat. Ann. 2C:39-19, 2C:39-20. With these alternatives, "[t]he ban does not require that owners turn over their magazines to law enforcement." Wiese v. Becerra, 806 F.Supp.3d 1190, 1198 (E.D. Cal. 2018); see Rupp v. Becerra, No. 8:17-cv-00746, 2018 WL 2138452, at *8 (C.D. Cal. May 9, 2018) (dismissing takings claim where "[t]he law offers a number of options to lawful gun owners that do not result in the weapon begin surrendered to the government").

The Act also does not result in a regulatory taking because it does not deprive the gun owners of all economically beneficial or productive uses of their magazines. See Murr v. Wisconsin, ___ U.S. ___, 137 S.Ct. 1933, 1942, 198 L.Ed.2d 497 (2017) (stating that "a regulation which denies all economically beneficial or productive use of land will require compensation under the Takings Clause" (internal quotation marks and citation omitted)); see also Lucas v. S.C. Coastal Council, 505 U.S. 1003, 1030, 112 S.Ct. 2886, 120 L.Ed.2d 798 (1992) (describing a "total taking" where a regulation "declares `off-limits' all economically productive or beneficial uses of land"). Simply modifying the magazine to hold fewer rounds of ammunition than before does not "destroy[] the functionality of the magazine." Wiese, 306 F.Supp.3d at 1198 (internal quotation marks omitted). Indeed, there is no assertion that a gun owner cannot use a modified magazine for its intended purpose. A gun owner may also retain a firearm with a fixed magazine that is "incapable of being modified to accommodate 10 or less rounds" or one that only "accepts a detachable magazine with a capacity of up to 15 rounds which is incapable of being modified to accommodate 10 or less rounds" so long as the firearm is registered. N.J. Stat. Ann. 2C:39-20(a). Thus, owners may keep their unmodifiable LCMs and modified versions. These magazines may be used in the same way expected: to hold multiple rounds of ammunition in a single magazine. In short, the Act does not result in a taking.

V

Finally, Plaintiffs' Equal Protection claim fails. The Equal Protection Clause provides that no state shall "deny to any person within its jurisdiction the equal protection of the laws." U.S. Const. amend. XIV, § 1. "This is essentially a direction that all persons similarly situated should be treated alike." Shuman ex rel. Shertzer v. Penn Manor Sch. Dist., 422 F.3d 141, 151 (3d Cir.2005) (citing City of Cleburne v. Cleburne Living Ctr., 473 U.S. 432, 439, 105 S.Ct. 3249, 87 L.Ed.2d 313 (1985)). Thus, to establish an equal protection claim, Plaintiffs "must demonstrate that they received different treatment from that received by other individuals similarly situated." Id. (citations omitted).

Plaintiffs assert that the Act violates the Fourteenth Amendment's Equal Protection Clause because it allows retired law enforcement officers to possess LCMs while prohibiting retired military members and ordinary citizens from doing so. N.J. Stat. Ann. 2C:39-3(g), 2C:39-17. Plaintiffs have not shown that retired law enforcement officers are similarly situated to other New Jersey residents. Retired law enforcement officers have training and experience not possessed by the general public. Kolbe, 849 F.3d at 147 (holding that retired law enforcement officers "are not similarly situated to the general public with respect to the assault weapons and large-capacity magazines banned"). Police officers in New Jersey must participate in firearms and defensive tactics training, including mandatory range and classroom training, under a variety of simulated conditions. App. 144; see, e.g., App. 1361, 1369, 1368, 1383. Law enforcement officers are also tested on a periodic basis after initial qualification and must re-qualify twice a year and meet certain shooting proficiency requirements. App. 144-45; see App. 1322-410 (describing standards, requirements, and full courses for law enforcement firearms qualification). Retired law enforcement officers must also satisfy firearms qualification requirements. N.J. Stat. Ann. 2C:39-6(l). Moreover, because the standard-issue weapon for many New Jersey law enforcement officers is a Glock 19 with a loaded fifteen round magazine, App. 116-17, these officers have experience carrying and using LCMs. Thus, law enforcement officers, both active and retired, have training and experience that distinguishes them from the general public.

Law enforcement officers are also different from members of the military. Unlike military personnel trained for the battlefield, law enforcement officers are trained for and have experience in addressing volatile situations in both public streets and closed spaces, and they operate in noncombat zones where the Constitution and other rules apply. App. 148-49. Even if some military members receive firearms training comparable to the training law enforcement officers receive, App. 140-41, the scope and nature of their training and experience are different, App. 141, 147-49.

For these reasons, retired law enforcement officers are not similarly situated to retired military personnel and ordinary citizens, and therefore their exemption from the LCM ban does not violate the Equal Protection Clause.

VI

For the foregoing reasons, we will affirm the order denying Plaintiffs' motion for a preliminary injunction.

BIBAS, Circuit Judge, dissenting.

The Second Amendment is an equal part of the Bill of Rights. We must treat the right to keep and bear arms like other enumerated rights, as the Supreme Court insisted in Heller. We may not water it down and balance it away based on our own sense of wise policy. 554 U.S. at 634-35, 128 S.Ct. 2783.

Yet the majority treats the Second Amendment differently in two ways. First, it weighs the merits of the case to pick a tier of scrutiny. That puts the cart before the horse. For all other rights, we pick a tier of scrutiny based only on whether the law impairs the core right. The Second Amendment's core is the right to keep weapons for defending oneself and one's family in one's home. The majority agrees that this is the core. So whenever a law impairs that core right, we should apply strict scrutiny, period. That is the case here.

Second, though the majority purports to use intermediate scrutiny, it actually recreates the rational-basis test forbidden by Heller. It suggests that this record favors the government, but make no mistake — that is not what the District Court found. The majority repeatedly relies on evidence that the District Court did not rely on and expert testimony that the District Court said was "of little help." 2018 WL 4688345, at *8. It effectively flips the burden of proof onto the challengers, treating both contested evidence and the lack of evidence as conclusively favoring the government.

Whether strict or intermediate scrutiny applies, we should require real evidence that the law furthers the government's aim and is tailored to that aim. But at key points, the majority substitutes anecdotes and armchair reasoning for the concrete proof that we demand for heightened scrutiny anywhere else. New Jersey has introduced no expert study of how similar magazine restrictions have worked elsewhere. Nor did the District Court identify any other evidence, as opposed to armchair reasoning, that illuminated how this law will reduce the harm from mass shootings. Id. at *12-13. So New Jersey cannot win unless the burden of proof lies with the challengers. It does not.

The majority also guts heightened scrutiny's requirement of tailoring. Alternatives to this ban may be less burdensome and as effective. New Jersey has already gone further than most states. It has a preexisting fifteen-round magazine limit and a restrictive permitting system. These laws may already do much to allay its public-safety concerns. New Jersey needs to show that these and other measures will not suffice.

The majority stands in good company: five other circuits have upheld limits on magazine sizes. These courts, like the New Jersey legislature, rightly worry about how best to reduce gun violence. But they err in subjecting the Second Amendment to different, watered-down rules and demanding little if any proof. So I would enjoin this Act until New Jersey provides real evidence to satisfy its burden of proving the Act constitutional.

I. STRICT SCRUTINY APPLIES TO LAWS THAT IMPAIR SELF-DEFENSE IN THE HOME

Unlike the majority, I would apply strict scrutiny to any law that impairs the core Second Amendment right to defend one's home. This law does so. And it fails strict scrutiny.

A. Other core constitutional rights get strict scrutiny

The Supreme Court has not set up tiers of scrutiny for gun regulations. Heller, 554 U.S. at 634, 128 S.Ct. 2783. That may be intentional: many rights do not have tiers of scrutiny. E.g., Duncan v. Louisiana, 391 U.S. 145, 88 S.Ct. 1444, 20 L.Ed.2d 491 (1968) (jury trial); Crawford v. Washington, 541 U.S. 36, 124 S.Ct. 1354, 158 L.Ed.2d 177 (2004) (Confrontation Clause). But our precedent mandates them for the Second Amendment, at least for laws that do not categorically ban commonly used weapons. See Marzzarella, 614 F.3d at 96-97.

As the majority recognizes, if we apply tiers of scrutiny, we apply strict scrutiny to the right's core. Maj. Op. at 117-18. For other rights, that is the end of the question. The "bedrock principle" of the Free Speech Clause forbids limiting speech just because it is "offensive or disagreeable." Texas v. Johnson, 491 U.S. 397, 414, 109 S.Ct. 2533, 195 L.Ed.2d 342 (1989). So content-based speech restrictions get strict scrutiny. Id. at 412, 109 S.Ct. 2533. The Free Exercise Clause was designed as a bulwark against "religious persecution and intolerance." Church of the Lukumi Babalu Aye, Inc. v. City of Hialeah, 508 U.S. 520, 532, 113 S.Ct. 2217, 124 L.Ed.2d 472 (1993) (internal quotation marks omitted). So laws that target religion or religious conduct get strict scrutiny. Id. at 533, 113 S.Ct. 2217. And the Equal Protection Clause targets classifications that historically were used to discriminate. See Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 236, 115 S.Ct. 2097, 132 L.Ed.2d 158 (1995). So laws that classify based on race get strict scrutiny. Id. at 235, 115 S.Ct. 2097.

B. The Second Amendment's core is self-defense in the home

The Second Amendment merits the same level of scrutiny. As Heller and McDonald confirm, and the majority acknowledges, its core turns on the weapon's function and its location: self-defense and the home. Maj. Op. 115-16, 117-18. Laws that tread on both warrant strict scrutiny.

Self-defense is the quintessential protected function of weapons. As Heller stressed, "it [i]s the central component of the right itself." 554 U.S. at 599, 128 S.Ct. 2783 (emphasis in original); accord id. at 628, 128 S.Ct. 2783. Heller thus focused on laws that deprive people of weapons commonly used for self-defense. Id. at 624, 629, 128 S.Ct. 2783. And McDonald focused on the history of colonists' and freedmen's defending themselves, whether from King George's troops or the Ku Klux Klan. 561 U.S. at 768, 772, 130 S.Ct. 3020 (majority opinion); id. at 857, 130 S.Ct. 3020 (Thomas, J., concurring in part and concurring in the judgment).

Not every gun law impairs self-defense. Our precedent applies intermediate scrutiny to laws that do not affect weapons' function, like serial-number requirements. Marzzarella, 614 F.3d at 97. But for laws that do impair self-defense, strict scrutiny is apt.

And the home is the quintessential place protected by the Second Amendment. In the home, "the need for defense of self, family, and property is most acute." McDonald v. City of Chicago, Ill., 561 U.S. 742, 767, 130 S.Ct. 3020, 177 L.Ed.2d 894 (2010) (quoting Heller, 554 U.S. at 628, 128 S.Ct. 2783). So the core is about using weapons in common use for self-defense in the home.

C. This Act burdens the core right

A ban on large magazines burdens that core right. Large magazines, unlike machineguns, are in common use. The ban extends to the home. Indeed, that is the main if not only locale of the law, as New Jersey can already deny most people permits to carry large magazines publicly. See N.J. Stat. Ann. § 2C:58-4(c). And the ban impairs using guns for self-defense. The government's entire case is that smaller magazines mean more reloading. That may make guns less effective for ill — but so too for good. The government's own police detective testified that he carries large magazines because they give him a tactical "advantage[]," since users must reload smaller magazines more often. App. 116-18. And he admitted that "law-abiding citizens in a gunfight" would also find them "advantageous." App. 119. So the ban impairs both criminal uses and self-defense.

The law does not ban all magazines, so it is not per se unconstitutional. But it does impair the core Second Amendment right. We usually would stop there. How much the law impairs the core or how many people use the core right that way does not affect the tier of scrutiny. So like any other law that burdens a constitutional right's core, this law warrants strict scrutiny.

D. The majority's responses are unconvincing

The majority tries to justify using intermediate scrutiny. But it errs twice over.

1. Forbidden interest-balancing. First and most fundamentally, the majority weighs the merits of the right to possess large magazines. It extends a passing phrase from Marzzarella into a requirement that a burden "severely burden the core Second Amendment right to self-defense in the home" before it will receive strict scrutiny. Maj. Op. at 117 (emphasis added) (citing Marzzarella, 614 F.3d at 97); accord id. at 118. It demands evidence that people commonly fire large magazines in self-defense. The challengers offer some data, and the government offers different data. The majority observes that the record is unclear on how many people fire more than ten rounds in self-defense. Maj. Op. at 112-13 & n.8. And it argues that people can use smaller magazines and "many other firearm options" anyway. Id. at 118; accord id. at 118, 121-22.

But the Second Amendment provides a right to "keep and bear Arms." U.S. Const. amend. II (emphasis added). It protects possessing arms, not just firing them. So the majority misses a key part of the Second Amendment. The analysis cannot turn on how many bullets are fired.

And we never demand evidence of how severely a law burdens or how many people it hinders before picking a tier of scrutiny. That demand is backwards and explicitly forbidden by Heller. We should read our precedent in keeping with the Supreme Court's instructions. Polling defensive gun uses and alternatives to set a level of scrutiny, as the majority does, boils down to forbidden interest-balancing. Any gun regulation limits gun use for both crime and self-defense. And any gun restriction other than a flat ban on guns will leave alternative weapons. So the majority's test amounts to weighing benefits against burdens.

That balancing approach is a variant of the position of Justice Breyer's dissent in Heller; the Heller majority rejected it. Compare 554 U.S. at 634-35, 128 S.Ct. 2783 (majority), with id. at 689-90, 128 S.Ct. 2783 (Breyer, J., dissenting). It makes no difference whether we break out the balancing into two steps or one. Maj. Op. at 119 n.22. And looking to smaller magazines and other options is the same argument, adapted to magazines, that the Court dismissed in Heller: "It is no answer to say ... that it is permissible to ban the possession of [large magazines] so long as the possession of other [] arms [like small magazines] is allowed." Id. at 629, 128 S.Ct. 2783. In picking a tier of scrutiny, our job is to ask only whether the ban extends to the home and impairs the gun's self-defense function.

Otherwise, we put the cart before the horse. Deciding the severity of the burden before picking a tier of scrutiny is deciding the merits first. It is backwards. That upends Heller's careful approach. The Supreme Court insisted that the Second Amendment has already made the basic policy choice for us. Id. at 634-36, 128 S.Ct. 2783. By enacting it, the Framers decided that the right to keep and bear arms is "really worth insisting upon." Id. at 634, 128 S.Ct. 2783 (emphasis in original). So the Court needed no data on how many people wield handguns defensively. It did not evaluate alternatives. It was enough that banning handguns impaired self-defense in the home. Id. at 628, 128 S.Ct. 2783.

That is how we approach other constitutional rights. The level of scrutiny for speech restrictions does not change if speech is unpopular or hateful. See Snyder v. Phelps, 562 U.S. 443, 458, 131 S.Ct. 1207, 179 L.Ed.2d 172 (2011). Nor does it change if a content-based burden is modest. See Reed v. Town of Gilbert, ___ U.S. ___, 135 S.Ct. 2218, 2224-27, 192 L.Ed.2d 236 (2015). Our scrutiny of classifications does not depend on how many people the law burdens. See United States v. Virginia, 518 U.S. 515, 531-34, 542, 116 S.Ct. 2264, 135 L.Ed.2d 735 (1996) (VMI) (noting that "most women would not choose VMI"). So it should not change our scrutiny of gun laws, no matter how unclear the record is on how many times "more than ten shots were used in self-defense." Maj. Op. at 112 & n.8.

Nor does the availability of alternatives lower our tier of scrutiny. Bans on flag-burning get strict scrutiny even though there are other ways to express one's views. See Johnson, 491 U.S. at 412, 109 S.Ct. 2533. Racial preferences for college applicants face the toughest scrutiny even though applicants can always go to other colleges. See Gratz v. Bollinger, 539 U.S. 244, 270, 123 S.Ct. 2411, 156 L.Ed.2d 257 (2003). The availability of alternatives bears on whether the government satisfies strict scrutiny, not on whether strict scrutiny applies in the first place. We focus on whether the government can achieve its compelling goal by using other restrictions, not on whether the rights-holder still has other avenues to exercise the right.

So the only question is whether a law impairs the core of a constitutional right, whatever the right may be. Any other approach puts the cart before the horse by weighing the merits of the case to pick a tier of scrutiny.

2. Limiting Heller's core to handgun bans. Second, though it denies it, the majority effectively cabins Heller's core to bans on handguns. Compare Maj. Op. at 115 n.14 (denying that Heller is so limited), with id. at 118 (stressing that this law, unlike the law in Heller, "does not take firearms out of the hands of law-abiding citizens" and leaves them with "many other firearm options"). But that is like cabining VMI to military institutes. Heller never limited its reasoning to handguns or complete bans, and for good reason. No other right works that way. Strict scrutiny applies to laws that burden speech or religion even if they do not nearly eliminate the right to speak or believe. E.g., Trinity Lutheran Church of Columbia, Inc. v. Comer, ___ U.S. ___, 137 S.Ct. 2012, 2021, 198 L.Ed.2d 551 (2017); Reed, 135 S.Ct. at 2225-27.

People commonly possess large magazines to defend themselves and their families in their homes. That is exactly why banning them burdens the core Second Amendment right. For any other right, that would be the end of our analysis; for the Second Amendment, the majority demands something much more severe.

So I would apply strict scrutiny to this Act, at least insofar as it limits keeping magazines to defend one's home. But as discussed below, the government has not shown that this Act can survive even intermediate scrutiny.

II. EVEN UNDER INTERMEDIATE SCRUTINY, ON THIS RECORD, THE LAW FAILS

Our precedent holds that intermediate scrutiny governs limits on weapons outside the home. Drake v. Filko, 724 F.3d 426, 436 (3d Cir.2013). The majority purports to apply that test. But its version is watered down — searching in theory but feeble in fact. It takes a record on which the District Court did not rely and construes everything in favor of the government, effectively flipping the burden onto the challengers. Even then, its analysis boils down to anecdotes and armchair reasoning. And the majority overlooks tailoring. None of that would be enough for other rights. I would apply true intermediate scrutiny, demanding evidence for the government's assertions and some showing of tailoring. Under either strict or true intermediate scrutiny, the law fails.

A. Intermediate scrutiny must be searching, not feeble

Though the Supreme Court has yet to specify a tier of scrutiny for gun laws, it forbade rational-basis review. Heller, 554 U.S. at 628 n.27, 128 S.Ct. 2783. So our scrutiny must not be so deferential that it boils down to a rational-basis test.

Intermediate scrutiny requires much more. As the majority concedes, the government bears the burden of proof. Maj. Op. at 120-21 n.24; Binderup, 836 F.3d at 353 (Ambro, J., controlling opinion). This is true even for preliminary injunctions. Gonzales v. O Centro Espirita Beneficente Uniao do Vegetal, 546 U.S. 418, 429, 126 S.Ct. 1211, 163 L.Ed.2d 1017 (2006). It must prove that the Act advances a substantial governmental interest. Marzzarella, 614 F.3d at 98. And though we may give some deference to the legislature's predictive judgments, those judgments must rest on real, hard evidence. Compare Drake, 724 F.3d at 436-37 ("accord[ing] substantial deference to the [legislature's] predictive judgments") (internal quotation marks omitted), with Frontiero v. Richardson, 411 U.S. 677, 689-90, 93 S.Ct. 1764, 36 L.Ed.2d 583 (1973) (castigating government's armchair, supposedly empirical reasoning unsupported by "concrete evidence").

It is not enough to base sex classifications on armchair reasoning. Frontiero, 411 U.S. at 689-90, 93 S.Ct. 1764 (applying intermediate scrutiny); see VMI, 518 U.S. at 541-43, 116 S.Ct. 2264 (same). So that should not be enough for gun laws either. Almost any gun law would survive an armchair approach; there are always plausible reasons to think that limiting guns will hinder criminals. That starts to look like rational-basis review.

The government must also prove that its law does not "burden more [conduct] than is reasonably necessary." Marzzarella, 614 F.3d at 98. To be sure, intermediate scrutiny does not demand the least restrictive means possible. But the government may not impair a constitutional right simply because doing so is convenient. McCullen v. Coakley, ___ U.S. ___, 134 S.Ct. 2518, 2534, 189 L.Ed.2d 502 (2014). It must make some showing that alternatives will not work. Id. at 2540. True intermediate scrutiny thus requires proof of tailoring.

So we must require that the government introduce substantial proof. We may not reflexively defer to its justifications. And we must look for tailoring. None of these requirements is met here.

B. The government has not met its burden of proof

New Jersey has not met its burden to overcome intermediate scrutiny, let alone strict scrutiny. True, the government has a compelling interest in reducing the harm from mass shootings. No one disputes that. But New Jersey has failed to show how the ban advances its interest. Nor does it provide evidence of tailoring.

1. The record lacks evidence that magazine restrictions reduce mass-shooting deaths. This record lacks any evidence tying that interest to banning large magazines. The reader could be forgiven for any surprise at that statement: the majority acts as if the record abounds in this evidence. But that is not what the District Court found. That Court offered three rationales for upholding the ban. None of them withstands scrutiny.

First, the District Court, like the majority here, reasoned that people can still own many, smaller magazines. 2018 WL 4688345, at *13. But Heller rejected that very argument. See 554 U.S. at 629, 128 S.Ct. 2783.

Second, the District Court stressed its deference to the legislature's judgment about the local needs of densely populated urban states. 2018 WL 4688345, at *13. In doing so, it relied not on the majority opinion in Heller but on Justice Breyer's dissent. Id. (quoting 554 U.S. at 705, 128 S.Ct. 2783 (Breyer, J., dissenting)). That citation alone shows how the deferential decision below conflicts with our governing instructions from above.

Third, the District Court detailed the testimony and evidence of all four expert witnesses. But it then "f[ou]nd the expert testimony is of little help in its analysis." Id. at *8. It found that evidence "of little help" in figuring out how the law would impair self-defense and how it would reduce the harm from mass shootings. Id. So none of this satisfied the government's burden of proof.

The only expert finding on which the District Court could rely was a vague and general one: "[T]he expert testimony established that there is some delay associated with reloading, which may provide an opportunity for potential victims to escape or for a bystander to intercede and somehow stop a shooter." Id. at *12. In other words, it rested on the armchair proposition that smaller magazines force shooters to pause more often to reload. When shooters must reload, potential victims should have more chances to escape or tackle the shooter. This speculation is plausible. But the Court cited no concrete causal link between that plausible speculation and its effect on mass-shooting deaths.

So with no support from the District Court, the majority digs through the record to link large magazines with the harm from mass shootings. By construing a record that the District Court found unhelpful in favor of the government, the majority effectively flips the burden of proof onto the challengers. It cites many portions of the record never mentioned by the District Court. It details the rise of mass shootings. It cites reports of mass shootings to show that people can escape when the shooter stops shooting. And it quotes a police chief as evidence that smaller magazines require more reloading.

The District Court was admirably clear about the state of the record. It did not rely on any of this "anecdotal evidence." Compare 2018 WL 4688345, at *3 (noting "anecdotal evidence"), with id. at *12 (not relying on it). And rightly so. The majority cannot tell us how many mass shooters use large magazines. It cannot tell us how often mass shooters use magazines with ten to fifteen rounds. And it cannot tell us any specifics about the increase in reload time. In short, the majority has no record citation, let alone evidence relied on by the District Court, that specifically links large magazines to mass-shooting deaths.

It has no citation because there isn't one. The government's own experts never examined the causal link between these magazines and crime. Its best evidence came from a lone CNN article that mentioned a study linking large magazines to mass shootings. But the government never introduced the actual study, the expert, or the underlying data. Nor was the study ever peer-reviewed. Without examination or cross-examination of the study, we cannot rely on it.

So to link reports of mass shootings to generalities about reload times, the majority resorts to saying: "[T]here is some delay associated with reloading, which may provide an opportunity for potential victims to escape or for a bystander to intercede." Maj. Op. at 120 (quoting 2018 WL 4688345, at *12). With no support for this analysis, the majority's case thus boils down to the same armchair reasoning that the District Court relied on, plus some "anecdotal evidence." 2018 WL 4688345, at *3. Though the majority insists otherwise, finding for the government on this basis alone effectively flips the burden of proof. Maj. Op. at 120-21 n.24. And the majority offers no limiting principle: its logic would equally justify a one-round magazine limit.

This reasoning would be enough for rational-basis review. And it could be enough for intermediate scrutiny too. But the government has produced no substantial evidence of this link. It could compile that evidence by, for example, studying other jurisdictions that have restricted magazine size. Until it does so, we should grant the preliminary injunction.

2. There is no evidence of tailoring. The majority does not even demand evidence of tailoring. But tailoring is not limited to the First Amendment, as our precedent makes clear. Marzzarella, 614 F.3d at 98. Tailoring is fundamental to intermediate scrutiny, wherever applied. McCullen, 134 S.Ct. at 2534; Caban v. Mohammed, 441 U.S. 380, 392 & n.13, 99 S.Ct. 1760, 60 L.Ed.2d 297 (1979) (illegitimacy).

If anything, the evidence shows that other effective laws are already on the books. In a footnote, the majority suggests that these other laws prove tailoring. Maj. Op. at 123 n.28. But far from it. If other laws already restrict guns, New Jersey has to show that the laws already on the books will not suffice. See McCullen, 134 S.Ct. at 2538-39. It has not done so.

To start, since 1990 New Jersey has banned magazines that hold more than fifteen bullets. The ban affects everyone. The challengers do not contest that ban. And there is no evidence of its efficacy, one way or the other. Though the government cites mass shootings involving large magazines, these shooters often used magazines with thirty or more rounds. So we do not know if a ten-round limit is tailored.

New Jersey also has a may-issue permitting law, requiring people to show a "justifiable need" before they may carry handguns outside the home. Drake, 724 F.3d at 428. We have upheld that law. Id. at 440. So the only people who can carry large magazines outside the home are those who face "specific threats or previous attacks which demonstrate a special danger" to their lives. Id. at 428 (quoting N.J. Admin. Code § 13:54-2.4(d)(1)). This limited universe of people includes abused women, those being stalked, and those fleeing gangs. Banning large magazines thus harms those who need the Second Amendment most.

Given its may-issue law, the government offers nothing to explain why this added ban is necessary, let alone tailored to its interests. If only those with a justifiable need can carry these magazines, why is New Jersey's law not tailored enough already? The government's only answer is that the may-issue requirement does not currently extend to the home. And the majority's only response is that many previously law-abiding citizens commit crime. But these arguments run up against strict scrutiny in the home. At most, they would warrant extending a may-issue permit requirement to the home, rather than banning large magazines entirely. And once again, the majority lacks a limiting principle: since anyone could commit crime, the government could forbid anyone to have a gun.

3. The majority muddles defensive gun uses. Instead of a real tailoring analysis, the majority again demands evidence of how often people use large magazines for self-defense. But tailoring does not depend on how many times a right is impaired.

The majority cannot even decide what the evidence shows. In places, it concedes that large magazines "have also been used for self-defense." Maj. Op. at 112; accord id. at 116-17. If so, this undercuts the ban. Elsewhere, it notes that the record is unclear on how often people shoot more than ten rounds in self-defense. Maj. Op. at 112 & n.8; accord id. at 122 n.27. If so, then New Jersey has not borne its burden of proof. Relying on unclearness amounts to flipping the burden of proof onto the challengers. Lastly, the majority most often concludes — even in the same breath — that large magazines are not appropriate for self-defense. Maj. Op. at 112, 117-18. But that is not what the District Court found. That Court specifically observed that the evidence "failed to clearly convey ... the extent to which the law will impede gun owners from defending themselves." 2018 WL 4688345, at *8. These contradictory assertions cannot bolster the law, nor satisfy the government's burden of proof.

4. The majority's watered-down "intermediate scrutiny" is really rational-basis review. This law would never survive the intermediate scrutiny applied by the Supreme Court in speech or sex-discrimination cases. Those cases demand compelling evidence and tailoring. See McCullen, 134 S.Ct. at 2534; VMI, 518 U.S. at 524, 116 S.Ct. 2264.

In a footnote, the majority candidly admits that it is not applying intermediate scrutiny as we know it. It concedes that its approach does not come from the First Amendment or the Fourteenth Amendment (or any other constitutional provision, for that matter). Maj. Op. at 122-23 n.28. It offers only one reason: guns are dangerous. Id. (quoting and relying on the Tenth Circuit's decision in Bonidy, 790 F.3d at 1126). But as Heller explained, other rights affect public safety too. The Fourth, Fifth, and Sixth Amendments often set dangerous criminals free. The First Amendment protects hate speech and advocating violence. The Supreme Court does not treat any other right differently when it creates a risk of harm. And it has repeatedly rejected treating the Second Amendment differently from other enumerated rights. Heller, 554 U.S. at 634-35, 128 S.Ct. 2783; McDonald, 561 U.S. at 787-91, 130 S.Ct. 3020. The Framers made that choice for us. We must treat the Second Amendment the same as the rest of the Bill of Rights.

So the majority's version of intermediate scrutiny is too lax. It cannot fairly be called intermediate scrutiny at all. Intermediate scrutiny requires more concrete and specific proof before the government may restrict any constitutional right, period.

* * * * *

I realize that the majority's opinion aligns with those of five other circuits. But Heller overruled nine, underscoring our independent duty to evaluate the law ourselves. And unlike most other states, New Jersey has layered its law on top of not only a previous magazine restriction, but also a may-issue permit law that greatly limits public carrying. Those laws may have prevented or limited gun violence. That cuts against the law's necessity and its tailoring.

The majority's concerns are understandable. Guns kill people. States should be able to experiment with reasonable gun laws to promote public safety. And they need not wait for mass shootings before acting. The government's and the majority's position may thus be wise policy. But that is not for us to decide. The Second Amendment is an equal part of the Bill of Rights. And the Supreme Court has repeatedly told us not to treat it differently.

So we must apply strict scrutiny to protect people's core right to defend themselves and their families in their homes. That means holding the government to a demanding burden of proof. Here, the government has offered no concrete evidence that magazine restrictions have saved or will save potential victims. Nor has it made any showing of tailoring.

I would thus enjoin the law and remand to let the government provide evidence that the Act will advance its interests and is tailored to do so. On remand, the government would be free to introduce real studies of any causal evidence that large-magazine limits prevent harm from mass shootings or gun violence in general. It could also introduce proof of tailoring and discuss its existing laws and alternatives. The challengers could try to rebut those studies. And we could then find whether the government has met its burden to justify this law. But it has not yet done that. So the law may well irreparably harm the challengers by infringing their constitutional rights. I respectfully dissent.