NEWSOM v. TRUMP (2026)

OPINION

In early June 2025, federal enforcement of immigration laws gave rise to public demonstrations in Los Angeles. In response, the President invoked 10 U.S.C. § 12406 to order 4,000 members of California's National Guard into federal service for 60 days to protect federal property and personnel. The State of California and its Governor, Gavin Newsom (collectively, “Plaintiffs”), sued the President, the Secretary of Defense, and the Department of Defense (collectively, “Defendants”) in federal court. Plaintiffs applied for a temporary restraining order, and, after a hearing, the district court issued an order enjoining Defendants “from deploying members of the California National Guard in Los Angeles” and directing Defendants “to return control of the California National Guard to Governor Newsom.” Newsom v. Trump, 786 F. Supp. 3d 1235, 1263 (N.D. Cal. 2025) (hereinafter “TRO”). Defendants immediately appealed and filed an emergency motion to stay the TRO pending appeal. On June 19, we issued an order granting Defendants' emergency stay motion. 1 Newsom v. Trump, 141 F.4th 1032, 1040–41 (9th Cir. 2025) (per curium) (hereinafter “Stay Order”).

Now, with the benefit of intervening Supreme Court precedent and full briefing, we consider the merits of Defendants' appeal. We conclude, as we did in the Stay Order, that the President's statutory authority to federalize a state's National Guard is conditioned on the existence of specified exigencies, and the President's determination of whether a required exigency exists is subject to judicial review. Although this review is highly deferential, we must evaluate, at a minimum, whether the President's determination “reflects a colorable assessment of the facts and law within a ‘range of honest judgment.’ ” Stay Order, 141 F.4th at 1051 (quoting Sterling v. Constantin, 287 U.S. 378, 399 (1932)).

When granting a stay of the district court's injunction, we concluded that the President “likely acted within his authority in federalizing the National Guard under 10 U.S.C. § 12406(3),” which authorizes federalization when “the President is unable with the regular forces to execute the laws of the United States.” Id. at 1052. However, after oral argument and the conclusion of briefing in this appeal, the Supreme Court issued an order denying the government's motion to stay a district court's temporary restraining order barring the federalization of the Illinois National Guard in a case involving similar facts. Trump v. Illinois, 146 S. Ct. 432 (2025) (per curiam). In Illinois, the Supreme Court concluded that the President lacks statutory authority to federalize the National Guard under § 12406(3) unless the President makes two threshold showings: first, that he has “statutory or constitutional authority to execute the laws with the regular military”; and second, that he is “ ‘unable’ with those forces to perform that function.” Id. at 434. Defendants have not made either of the required threshold showings here. Thus, they have not carried their burden to show that § 12406(3) authorized the Guard's federalization and deployment.

The Supreme Court's order in Illinois does not end our inquiry, however, because Defendants argue that the California federalization order “was independently warranted under [§ 12406(2),] the provision authorizing him to call the National Guard into federal service whenever ‘there is a rebellion or danger of a rebellion against the authority of the Government of the United States.’ ” We conclude that Plaintiffs are likely to succeed on their claim that the exigency determination at issue here—that the conditions in Los Angeles on June 7, 2025, constituted “a rebellion or danger of rebellion” under § 12406(2)—does not reflect a colorable assessment within a “range of honest judgment.” Accordingly, § 12406(2) did not authorize the President's order federalizing the California National Guard.

Because neither § 12406(3) nor § 12406(2) authorized the President's initial federalization order, the order exceeded the President's statutory authority and encroached on Congress's calling forth power, in violation of the constitutional separation of powers. Because the order was ultra vires and unconstitutional, the balance of equities favors granting injunctive relief. Accordingly, the district court did not abuse its discretion in issuing the TRO, and we affirm. 2

I. APPELLATE JURISDICTION

We have jurisdiction to hear an interlocutory appeal of a preliminary injunction. 28 U.S.C. § 1292(a)(1). Here, the district court issued a temporary restraining order, not a preliminary injunction. However, as we explained in our Stay Order, because the district court's TRO “ ‘possesses the qualities of a preliminary injunction,’ it is reviewable under 28 U.S.C. § 1292(a)(1).” 141 F.4th at 1043 (quoting Serv. Emps. Int'l Union v. Nat'l Union of Healthcare Workers, 598 F.3d 1061, 1067 (9th Cir. 2010)).

II. STANDARD OF REVIEW

We review the district court's issuance of a preliminary injunction for abuse of discretion. Am. Trucking Ass'ns v. City of Los Angeles, 559 F.3d 1046, 1052 (9th Cir. 2009). We review its conclusions of law de novo and its findings of fact for clear error. All. for the Wild Rockies v. Cottrell, 632 F.3d 1127, 1131 (9th Cir. 2011). Preliminary injunctive relief is appropriate where a plaintiff shows “that he is likely to succeed on the merits, that he is likely to suffer irreparable harm in the absence of preliminary relief, that the balance of equities tips in his favor, and that an injunction is in the public interest.” Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 20 (2008).

III. DISCUSSION

A. The Power to Federalize a State's National Guard

This appeal concerns the nature and scope of the President's authority to federalize a state's National Guard, an authority historically referred to as the power to “call forth” the militia. The President's power to “call forth” members of a state's National Guard is statutory, not constitutional. Although the Constitution provides that “[t]he President shall be Commander in Chief ․ of the Militia of the several States, when called into the actual Service of the United States,” U.S. CONST. art. II, § 2, cl. 1, it expressly grants Congress the power to “call[ ] forth” the state militias into federal service to address three specific exigencies: “to execute the Laws of the Union, suppress Insurrections and repel Invasions.” Id. art. I, § 8, cl. 15. 3

Although the Constitution grants the “calling forth” power to Congress, Congress, in turn, has statutorily delegated a portion of that power to the President. Congress first partially delegated its “calling forth” power to the President in the Militia Act of 1792, ch. 28, § 1, 1 Stat. 264, 264 (repealed 1795). Congress renewed that delegation in the Militia Act of 1795, ch. 36, § 1, 1 Stat. 424, 424. The 1795 Act was a precursor to the Militia Act of 1903, Pub. L. No. 57-33, §§ 1, 4, 32 Stat. 775, 775–76; and Section 4 of the 1903 Act was a precursor to 10 U.S.C. § 12406, see Wiener, supra note 3, at 186–88.

Through these enactments, Congress has authorized the President to exercise its “calling forth” power, but—consistent with the limits set in Article I—only in response to three specific exigencies. The current version of the “calling forth” statute, 10 U.S.C. § 12406, provides:

Whenever—

(1) the United States, or any of the Commonwealths or possessions, is invaded or is in danger of invasion by a foreign nation;

(2) there is a rebellion or danger of a rebellion against the authority of the Government of the United States; or

(3) the President is unable with the regular forces to execute the laws of the United States;

the President may call into Federal service members and units of the National Guard of any State in such numbers as he considers necessary to repel the invasion, suppress the rebellion, or execute those laws. Orders for these purposes shall be issued through the governors of the States or, in the case of the District of Columbia, through the commanding general of the National Guard of the District of Columbia.

10 U.S.C. § 12406.

B. Judicial Review of § 12406 Federalization Orders

In our Stay Order, we rejected Defendants' argument that, under Dalton v. Specter, 511 U.S. 462 (1994), the President's decision to federalize the National Guard under § 12406 is insulated from judicial review. See Stay Order, 141 F.4th at 1047. In Dalton, the Supreme Court held that judicial review of the President's decision approving the closure of a naval base was unavailable because he acted pursuant to a statute that “authorized unfettered discretion by the President to either approve or disapprove the package of base closures” proposed by an independent commission. Sierra Club v. Trump, 929 F.3d 670, 696 (9th Cir. 2019). Defendants do not further develop this argument in their merits briefs; they only cite Dalton and assert that it is analogous to this case. We reject Defendants' argument that § 12406 grants the President unfettered discretion for the reasons articulated in our Stay Order. We summarize them here.

Unlike the base-closure statute at issue in Dalton, § 12406 “enumerates three predicate conditions for the President's decision to call forth the National Guard.” Stay Order, 141 F.4th at 1047. The President may call forth the militia under § 12406 only if there is an “invasion,” a “rebellion,” a “danger of” invasion or rebellion, or an inability “to execute the laws with the regular forces.” Accordingly, as the Supreme Court recognized in Martin v. Mott, exercising the “calling forth” power involves two distinct determinations: first, that the statutorily required exigency exists; and second, that a state militia should be called forth to address that exigency. 25 U.S. (12 Wheat.) 19, 31–32 (1827). If the requisite exigency does not exist—that is, if the President's first determination is incorrect—and he nonetheless decides to call forth the militia, he acts without constitutional authority and violates separation-of-powers principles. See Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579, 585-89 (1952) (invalidating wartime presidential order because President acted with neither statutory nor constitutional authority); see also id. at 637–38 (Jackson, J., concurring) (observing that the President's “power is at its lowest ebb” when he “takes measures incompatible with the expressed or implied will of Congress”).

The question then follows: to what extent has Congress, in § 12406, committed the determination of whether the required exigency exists to the President's discretion? As we explained in our Stay Order, “[t]his question is purely a matter of statutory interpretation, and it is justiciable.” 141 F.4th at 1046 (citing Trump v. J.G.G., 604 U.S. 670, 672 (2025) (per curiam)). The courts have “[t]he responsibility of determining the limits of statutory grants of authority,” Stark v. Wickard, 321 U.S. 288, 310 (1944), and the “power to construe the statutes involved to determine whether [a government official] did exceed his powers,” Harmon v. Brucker, 355 U.S. 579, 582 (1958) (per curiam).

We interpret the meaning of a statute de novo. Loper Bright Enters. v. Raimondo, 603 U.S. 369, 392 (2024). As usual, “[w]e start with the text.” Thompson v. United States, 604 U.S. 408, 413 (2025). “Interpretation of a word or phrase depends upon reading the whole statutory text, considering the purpose and context of the statute, and consulting any precedents or authorities that inform the analysis.” Dolan v. U.S. Postal Serv., 546 U.S. 481, 486 (2006).

Section 12406 does not include express language making the President the sole judge of whether a required exigency exists. Consequently, if we were considering the text of § 12406 alone, we might conclude that the President's exigency determination is subject to review like certain other factual findings that are preconditions for executive action under a statute. See Doe #1 v. Trump, 957 F.3d 1050, 1066–67 (9th Cir. 2020).

In our Stay Order, however, we concluded that the “history of Congress's statutory delegations of its calling forth power,” and a line of cases interpreting those delegations, beginning with Martin, 25 U.S. (12 Wheat.) 19, “strongly suggest that our review of the President's determinations in this context is especially deferential.” 141 F.4th at 1047. We also concluded that we are largely bound by the Martin Court's interpretation of the calling forth provision of the 1795 Militia Act, a statutory predecessor of § 12406. See id. at 1047–50. In Martin, the Supreme Court concluded that, because the 1795 Act granted the President a “discretionary power” to federalize a state's militia when the required exigency exists, the Act impliedly made the President “the judge of the existence of the exigency in the first instance.” 25 U.S. at 31. We also recognized, however, that the Supreme Court clarified in Sterling that “courts may at least review the President's [exigency] determination to ensure that it reflects a colorable assessment of the facts and law within a ‘range of honest judgment.’ ” Stay Order, 141 F.4th at 1051 (quoting Sterling, 287 U.S. at 399).

In this appeal, the parties contest our readings of those cases. Plaintiffs contend that Martin is limited to its facts—a militiaman's defense to charges for failing to report to duty. See 25 U.S. at 28. Outside of that narrow factual context, Plaintiffs argue, the President's authority under § 12406 is not discretionary, and courts may review his exigency determination with little, if any, deference. We continue to disagree, for the reasons stated in our Stay Order. See Stay Order, 141 F.4th at 1048–51. 4

Defendants agree with our conclusion that the Court's reasoning in Martin controls here, even though that case involved materially different factual circumstances. But Defendants argue that we erred in reaching the same conclusion about the Court's reasoning in Sterling. That is, Defendants argue that Martin is not limited to its facts, but Sterling is. Thus, Defendants contend, the Martin Court's description of the President's exigency determination as “conclusive upon all other persons” means that, even if the President federalized the Guard based on no evidence whatsoever, the courts would be unable to review his determination that an exigency existed.

We continue to conclude that Sterling, as much as Martin, provides controlling reasoning, and that Sterling forecloses the Defendants' extreme position that the executive's exigency determination is unreviewable in all cases. In Sterling, the Supreme Court reaffirmed Martin's construction of the President's statutory power to call forth the militia to respond to an exigency. Sterling, 287 U.S. at 399 (discussing Martin, 25 U.S. (12 Wheat.) at 29–30). But the Court clarified that, notwithstanding Martin's statement that the executive is the “sole and exclusive judge” of the existence of an exigency, a court can review that determination when a plaintiff colorably claims that the exercise of military authority is causing it constitutional injury and the constitutionality of the executive's military order turns on whether the exigency exists. See id. at 399–401.

Sterling arose when, amid widespread oil overproduction in Texas, the Governor issued a proclamation “declaring ‘martial law’ ” in certain counties and “called out” the Texas National Guard to “shut down” the oil wells to give a state commission time to issue orders limiting oil production. Id. at 387, 389–90. Much like § 12406, the Texas Constitution empowered the Governor, as the commander-in-chief of the State's militia, to “call forth the militia to execute the laws of the state, to suppress insurrection, repel invasion, and protect the frontier.” Id. at 394 (quoting Tex. Const. art. IV, § 7). And much like the Executive Order challenged here, the Governor's proclamation asserted that an organized group of oil and gas producers were “in a state of insurrection against the [oil] conservation laws,” and “that the civil officers did not have a sufficient force to compel them to obey.” Id. at 389.

After the state commission issued an order limiting oil production, the wells reopened but “the proclamation of martial law was not rescinded nor the troops entirely withdrawn.” Id. at 390. Oil and gas producers obtained a temporary restraining order against the commission's order in federal court, but in response, the Governor issued more military orders imposing and enforcing comparable production limits. Id.at 387–88. The oil and gas producers then challenged the Governor's “military and executive orders,” alleging they were unjustified because there was no actual insurrection. See id. at 388. The Governor answered the complaint by “setting forth the executive proclamation and orders, and the declaration of martial law,” and “asserting [their] validity.” Id. Additionally, the Governor argued that the federal court was “without jurisdiction over [those] actions” “during the continuance of the proclaimed state of war ․ by virtue of the claim ․ that war conditions were prevailing in the field, and that military necessity required the action.” Id. at 390 (internal quotation marks omitted). Thus, both the justiciability and merits of plaintiffs' constitutional claims turned on whether “the Governor, under the Constitution and statutes of the state, could lawfully exercise the authority he had assumed.” Id. at 388. And that question turned on whether the asserted exigency—an insurrection—actually existed. See id. at 390–92.

The district court considered the evidence submitted by both parties regarding the existence of an exigency and determined that the “evidence shows no insurrection nor riot.” Constantin v. Smith, 57 F.2d 227, 232 (E.D. Tex. 1932). Accordingly, the district court enjoined the Governor “from enforcing [his] military or executive orders regulating or restricting the production of oil from complainants' wells.” Sterling, 287 U.S. at 386.

On appeal to the Supreme Court, the Governor, much like the Defendants here, argued that “courts may not review the sufficiency of facts upon which martial law is declared.” Id. at 393. Even though the district court had found that the Governor's action was “unjustified by any existing exigency” and that injunctive relief was appropriate, the Governor argued that the court “was powerless thus to intervene” because “the Governor's order had the quality of a supreme and unchallengeable edict, overriding all conflicting rights of property and unreviewable through the judicial power of the federal government.” Id. at 397.

The Supreme Court soundly rejected that “extreme position” and articulated several reasons why the executive's exigency determination was subject to judicial review. See id. at 393–404. Many of those considerations are equally applicable here.

To begin, the Court explained that insulating the Governor's exigency determination from judicial review would conflict with the limits on governmental power established in the federal Constitution. Accepting the Governor's position would mean that

the fiat of a state Governor, and not the Constitution of the United States, would be the supreme law of the land; [and] the restrictions of the Federal Constitution upon the exercise of state power would be but impotent phrases, the futility of which the state may at any time disclose by the simple process of transferring powers of legislation to the Governor to be exercised by him, beyond control, upon his assertion of necessity. Under our system of government, such a conclusion is obviously untenable.

Id. at 397-98.

The Court thus held that the Governor's exigency determination was reviewable. Because the plaintiffs had made “a substantial showing that the exertion of state power has overridden private rights secured by that Constitution,” a federal court could review the exigency determination “in an appropriate proceeding directed against the individuals charged with the transgression.” Id. at 398. Further, the Court instructed, “when questions of law and fact are so intermingled as to make it necessary, in order to pass upon the [constitutional] question, the court may, and should, analyze the facts.” Id.

Importantly, the Court concluded that the Governor's exigency determination was reviewable even though it also concluded that Martin's construction of the President's statutory calling forth power applied equally to the Governor's power to call forth the militia under state law:

By virtue of his duty to ‘cause the laws to be faithfully executed,’ the executive is appropriately vested with the discretion to determine whether an exigency requiring military aid for that purpose has arisen. His decision to that effect is conclusive. That construction, this Court has said, in speaking of the power constitutionally conferred by the Congress upon the President to call the militia into actual service, ‘necessarily results from the nature of the power itself, and from the manifest object contemplated.’ The power ‘is to be exercised upon sudden emergencies, upon great occasions of state, and under circumstances which may be vital to the existence of the Union.’ Martin v. Mott, [25 U.S. (12 Wheat.) at 30]. Similar effect, for corresponding reasons, is ascribed to the exercise by the Governor of a state of his discretion in calling out its military forces to suppress insurrection and disorder. Luther v. Borden, [48 U.S. (7 How.) 1, 45 (1849)]; Moyer v. Peabody, 212 U.S. 78, 83 [(1909)].

Id. at 399 (emphasis added).

The Court acknowledged that the nature of the executive power at issue “also necessarily implies that there is a permitted range of honest judgment as to the measures to be taken in meeting force with force, in suppressing violence and restoring order, for, without such liberty to make immediate decisions, the power itself would be useless.” Id. at 399–400. “Such measures, conceived in good faith, in the face of the emergency, and directly related to the quelling of the disorder or the prevention of its continuance, fall within the discretion of the executive in the exercise of his authority to maintain peace.” Id. at 400.

Yet, the Court cautioned, “[i]t does not follow from the fact that the executive has this range of discretion, deemed to be a necessary incident of his power to suppress disorder, that every sort of action the Governor may take, no matter how unjustified by the emergency or subversive of private right and the jurisdiction of the courts, otherwise available, is conclusively supported by mere executive fiat. The contrary is well established.” Id. at 400-01. Further, the Court underscored, “the allowable limits of military discretion, and whether or not they have been overstepped in a particular case, are judicial questions.” Id. at 401. “Thus, in the theater of actual war,” there are occasions in which military officers may take private property, “and the officer may show the necessity” for the taking “in defending an action for trespass.” Id. But, as the Court held in Mitchell v. Harmony, “[i]t is the emergency that gives the right, and the emergency must be shown to exist before the taking can be justified.” Id. (quoting Mitchell, 54 U.S. (13 How.) 115, 134 (1851)). That principle, the Court concluded, applied equally to the Governor's defense in Sterling because “[t]here is no ground for the conclusion that military orders in the case of insurrection have any higher sanction or confer any greater immunity.” Id.

The Court then rejected the Governor's assertion that military “action can be taken as conclusive proof of its own necessity” as having “no support in the decisions of this Court.” Id. at 402. The Court noted that it had rejected a similar proposition in Ex parte Milligan, 71 U.S. (4 Wall.) 2, 124 (1866), because martial law imposed based on a military commander's determination that “the exigencies of the country demand it, and of which he is to judge,” and which subjected “citizens as well as soldiers to the rule of his will,” “destroys every guaranty of the Constitution, and effectually renders the ‘military independent of and superior to the civil power.’ ” Sterling, 287 U.S. at 402–03 (quoting Ex parte Milligan, 71 U.S. at 124).

Having concluded that the Governor's exigency determination was subject to judicial review, the Court reviewed the district court's factual findings and determined that they were “fully supported by the evidence.” Id. at 403– 04. Accordingly, the Court affirmed the district court's conclusion that “[t]here was no exigency which justified the Governor in attempting to enforce by executive or military order the restriction [on oil production] which the district court had restrained pending proper judicial inquiry.” Id. at 404.

Reading Martin and Sterling together, we conclude that the existence of an exigency required to invoke § 12406 is reviewable under the circumstances presented here. Both Martin and Sterling involved challenges to military orders in which the executive's authority to issue the order depended on the existence of an exigency. Further, the Court held that the Governor's exigency determination in Sterling was “conclusive” for the same reasons that the President's exigency determination was “conclusive” in Martin. But the Court concluded that the exigency determination in Sterling was reviewable, even though the exigency determination in Martin was not. The key difference between Martin and Sterling lies in the nature of the challenge to the executive's order. In Martin, the existence of the exigency underlying the President's calling-forth order was challenged by a militia member, Mott, as a defense to charges for disobeying the order to appear for service. Martin, 25 U.S. (12 Wheat.) at 28. The militia member argued only that he did not have to report for duty because the order was invalid, not that the federalization order violated his constitutional rights. In Sterling, by contrast, the plaintiffs challenged the existence of the exigency asserted in the Governor's martial-law proclamation in the context of their claim that the Governor's ensuing orders violated their constitutional rights. If the exigency existed, then it justified the Governor's martial-law orders, and the Governor did not violate the Due Process Clause. If, however, the exigency did not exist, the orders were unauthorized and violated the plaintiffs' federal constitutional rights. Sterling, 287 U.S. at 401. As the Court explained, the federal court could hear the plaintiffs' “cogniza[ble]” claim “of an alleged invasion of a right,” and “the judicial power necessarily extends to the granting of the relief found to be appropriate according to the circumstances of the case.” Id. at 403.

Reconciling Martin with Sterling, we conclude that § 12406 grants the President the authority to federalize the National Guard to respond to an exigency, and because that power “is to be exercised upon sudden emergencies, upon great occasions of state, and under circumstances which may be vital to the existence of the Union,” the President has “discretion to determine whether an exigency requiring military aid ․ has arisen.” Id. at 399 (quoting Martin, 25 U.S. (12 Wheat.) at 30). If a subordinate officer challenges the validity of the President's order to determine whether he must obey it, the order and the underlying exigency determination are unreviewable. Martin, 25 U.S. (12 Wheat.) at 30. But if a plaintiff colorably alleges that the President's federalization order has unjustifiably infringed upon her constitutional rights and the merits of that claim depend on whether the President had the authority to issue the order, then a court may review the President's predicate exigency determination. Because it is the exigency that gives the President the authority to act, the exigency “must be shown to exist” before his action can be justified. Sterling, 287 U.S. at 401 (quoting Mitchell, 54 U.S. (13 How.) at 134).

Applying those principles here, we conclude the present case is akin to Sterling, rather than Martin. Plaintiffs allege constitutional violations that infringe the State's sovereign interests. See Complaint ¶¶ 69, 95, 98, Newsom v. Trump, 786 F. Supp. 3d 1235 (N.D. Cal. 2025) (No. 25-cv-4870), Dkt. No. 1, 2025 WL 1640287 (hereinafter “Complaint”).

Although Sterling involved alleged violations of individuals' private due process rights, whereas this case involves alleged violations of a State's constitutionally protected interests under the Tenth Amendment, the constitutional violations alleged here are at least as significant as the due process violations at issue in Sterling. They strike at the heart of our federal system and the limits on governmental power that are the foundation of our democracy.

Defendants argue that Sterling “says nothing” about Martin's holding that the President's exigency determination is “conclusive” because Sterling “involved judicial review of the conduct of the militia once deployed, not of the Governor's determination that deployment of the militia was warranted in the first instance.” That argument is unpersuasive for several reasons. To begin, we agree that Sterling did not disturb Martin's holding that the President's exigency determination is “conclusive.” Indeed, the Court held that the Governor's exigency determination was “conclusive” for the same reasons it had set forth in Martin. Sterling, 287 U.S. at 399. But, as explained above, the Court held both that the Governor's exigency determination was conclusive and that it was subject to judicial review. Id. at 400.

Defendants' argument to the contrary misreads Sterling. The plaintiffs in that case did not challenge how the militia members conducted themselves when carrying out the Governor's orders; the plaintiffs challenged the validity of the orders themselves. See id. at 388 (“[C]omplainants filed an amended bill ․ alleging that the above-mentioned military and executive orders limiting production were without justification in law or in fact ․”). The Court viewed the Governor's exigency determination in Sterling as analogous to the President's exigency determination at issue in Martin, yet concluded that the Governor's determination was reviewable. Id. at 399.

In arguing that Sterling has no bearing on the reviewability of the President's exigency determination, Defendants note that the Court stated that the question was “not of the power of the Governor to proclaim that a state of insurrection” exists, but was limited to “the Governor's attempt to regulate by executive order the lawful use of complainants' properties in the production of oil.” Id. at 401–02. Read in context, the Court was merely reframing the issue presented—instead of asking whether the Governor had the power to declare that an exigency existed and call forth the military to respond, the Court asked whether the Governor's use of that power unjustifiably interfered with the plaintiffs' use of their properties. Because the answer to the latter question turned on whether the Governor properly used that power to respond to an exigency, the factual existence of the exigency was subject to judicial review. See id. at 402. 5

The same reframing is appropriate here. The issue is not whether the President has the power to proclaim that an exigency exists and call forth the National Guard; § 12406 plainly grants him that authority. Rather, the question is whether the President's use of that authority violates Plaintiffs' constitutional rights. Because the answer to that question depends on the existence of the required exigency, we may review the President's determination that it exists. 6

In a “statement respecting the grant of rehearing en banc” in Oregon v. Trump, Judge Tung asserted that statutory history shows that “Congress now vests the President with exclusive discretion in determining whether [the required] exigency exists.” 165 F.4th 1158, 1189 (9th Cir. 2025). Specifically, he pointed out that § 2 of the Militia Act of 1792 required judicial certification as a predicate to the President's authority to call forth the militia, but Congress removed that requirement when it enacted the Militia Act of 1795. Id. 7 This statutory history does not change our conclusion for three reasons.

First, § 2 of the Militia Act of 1792 is a predecessor of a provision of the Insurrection Act currently codified at 10 U.S.C. § 252, not of § 12406. 8

Second, this amendment to § 252's predecessor does not suggest that Congress intended to restrict judicial review of the President's exigency determination. Under the 1792 Militia Act, the judicial-certification requirement was, like the existence of a specified exigency, a predicate for the President's statutory authority to call forth the militia, rather than a restriction on judicial review. The Constitution, not the statute, provides the right to judicial review “because of [the] alleged invasion of a [constitutional] right.” Sterling, 287 U.S. at 398, 403.

Third, we agree with Judge Tung that § 12406 “vests the President with exclusive discretion in determining whether [the required] exigency exists.” Oregon, 165 F.4th at 1189 (Tung, J., statement respecting the grant of rehearing en banc). The Supreme Court established this in Martin and reiterated it in Sterling. See supra at 22-23, 28-29; Sterling, 287 U.S. at 399 (“[T]he executive is appropriately vested with the discretion to determine whether an exigency requiring military aid ․ has arisen.”). Yet the Court in Sterling still concluded that the executive's exigency determination was reviewable. 287 U.S. at 400 (“It does not follow from the fact that the executive has this range of discretion” that his action “is conclusively supported by mere executive fiat.”); see supra at 28-30; see also Panama Refin. Co. v. Ryan, 293 U.S. 388, 446 (1935) (Cardozo, J., dissenting) (“A court will not revise the discretion of the Executive, sitting in judgment on his order as if it were the verdict of a jury. Martin v. Mott, supra. On the other hand, we have said that his order may not stand if it is an act of mere oppression, an arbitrary fiat that overleaps the bounds of judgment. Sterling v. Constantin, 287 U.S. 378, 399–401.”).

The Supreme Court reaffirmed these principles of judicial review in Trump v. Cook, 146 S. Ct. 2234, 2247 (2026). Even when a statute authorizes “only the President” to make a decision, “that does not mean that he may make that decision for any reason, or no reason.” Id. And “[e]ven when a statute ‘delegates discretionary authority’ to the Executive Branch, ․ our role ‘is, as always, to independently interpret the statute and effectuate the will of Congress subject to constitutional limits.’ ” Id. (quoting Loper Bright, 603 U.S. at 395).

Here, as in Sterling, the question whether the executive's military action violates the Plaintiffs' constitutional rights turns on whether the action was authorized. That second question turns on the existence of an exigency, a statutory precondition for the executive's exercise of authority. We may review the executive's determination that the required exigency exists.

C. Standard of Review for Exigency Determinations

Having concluded that the President's exigency determination is reviewable, we now consider the standard of review. Under Martin and Sterling, our review must be highly deferential, but the President's discretion has “limits” that may be “overstepped in a particular case,” and it is the judiciary's responsibility to define what those limits are. Sterling, 287 U.S. at 401. Although the Court has not precisely defined the level of deference we owe in this context, Sterling makes clear that we may “at least review the [President's exigency] determination to ensure that it reflects a colorable assessment of the facts and law within a ‘range of honest judgment.’ ” Stay Order, 141 F.4th at 1051 (quoting Sterling, 287 U.S. at 399).

We glean three lessons from the Court's application of that standard in Sterling. First, to determine whether the required exigency actually exists, the district court should perform its normal fact-finding function, developing a detailed account of “who did what, when or where, how or why,” U.S. Bank Nat'l Ass'n ex rel. CWCapital Asset Mgmt. LLC v. Vill. at Lakeridge, LLC, 583 U.S. 387, 394 (2018) (citation omitted). See Sterling, 287 U.S. at 389-92 (“[T]he District Court received the evidence submitted by both parties, and ․ stated that it established the following facts ․”).

Second, the district court must apply the legal standard for the asserted exigency (e.g., the meaning of “rebellion”) to its own factual findings, not to the President's factual assertions. See id. at 403-04 (“In the present case, the findings of fact made by the District Court ․ leave no room for doubt that there was no military necessity which, from any point of view, could be taken to justify the action of the Governor ․”). The interpretation of the governing legal standard is a question of law for the court. See Loper Bright Enters., 603 U.S. at 392.

Third, the court must assess the President's exigency determination under an objective standard. In Sterling, the Court explained that executive measures “conceived in good faith, in the face of the emergency, and directly related to” that emergency “fall within the discretion of the executive in the exercise of his authority to maintain peace.” 287 U.S. at 400. In assessing the Governor's exigency determination, the Court considered only whether there was a reasonable factual basis for the governor's determination that the exigency existed. See id. at 403–04. It undertook no inquiry into the governor's state of mind, subjective knowledge, or intent. See id. Thus, the Court left open the possibility that evidence of subjective bad faith may be relevant but made clear that no showing of subjective bad faith is required. Under this objective standard of review, a plaintiff can prevail when the record shows that the President's determination does not fall within a broad range of reasonable judgment.

Defendants urge us to apply an even more deferential standard of review because “Plaintiffs' only path to judicial review of the President's federalization order ․ is a non-statutory ultra vires claim.” In support of this argument, they cite only the Supreme Court's admonition in Nuclear Regulatory Commission v. Texas that such a claim is a “Hail Mary pass” that “rarely succeeds.” 605 U.S. 665, 681-82 (2025) (quoting Nyunt v. Chairman, Broad. Bd. of Governors, 589 F.3d 445, 449 (D.C. Cir. 2009)). But Nuclear Regulatory Commission concerned when an ultra vires cause of action is available, not the level of deference we should apply when reviewing an executive action challenged as ultra vires.

D. Cause of Action

Defendants argue only that Plaintiffs' ultra vires claim fails on the merits—they do not argue, at least in their opening brief, that Plaintiffs lack a valid ultra vires cause of action. Indeed, Defendants have repeatedly forfeited the cause-of-action issue: They did not mention it before the district court, in their briefing on the emergency stay motion, or in their opening merits brief. 9 See Martinez v. High, 91 F.4th 1022, 1028 n.4 (9th Cir. 2024) (“As a general rule, we will not consider arguments that are raised for the first time on appeal.” (citation omitted)); Padgett v. Wright, 587 F.3d 983, 985 n.2 (9th Cir. 2009) (per curiam) (“This court will not ordinarily consider matters on appeal that are not specifically and distinctly raised and argued in appellant's opening brief.” (internal quotation marks and citation omitted)).

The dissent contends Defendants' merits arguments raise the cause-of-action issue. Dissent at 87-88. But in our view, whether Plaintiffs have a valid cause of action and whether they should prevail on the merits are distinct issues—not, as the dissent contends, merely different arguments relating to the same issue. While we are “not limited to the particular legal theories advanced by the parties,” Kamen v. Kemper Fin. Servs., Inc., 500 U.S. 90, 99 (1991), we are limited to addressing issues “bearing a fair resemblance to the case shaped by the parties,” United States v. Sineneng-Smith, 590 U.S. 371, 380 (2020); see Kamen, 500 U.S. at 100 n.5. Whether a cause of action exists and whether a plaintiff states a claim are separate questions, see, e.g., Ashcroft v. Iqbal, 556 U.S. 662, 675-76 (2009), and Defendants adequately raised only the latter issue.

Even if Defendants' failure to raise the cause-of-action issue before the district court may be excused, the dissent offers no justification for their failure to raise it in their opening brief. While parties may raise new arguments on appeal, see Lebron v. Nat'l R.R. Passenger Corp., 513 U.S. 374, 379 (1995), Defendants never challenged the existence of a cause of action in their opening brief. Instead, they addressed the issue only in reply. Even if a party may make a new argument in its opening brief on appeal, we have long recognized that “arguments not raised by a party in its opening brief” are forfeited. Smith v. Marsh, 194 F.3d 1045, 1052 (9th Cir. 1999). 10

The dissent also contends that, even if Defendants forfeited the cause-of-action issue, we must address it. Dissent at 88. Again, we disagree. Whether a plaintiff has a cause of action is not a jurisdictional question that we must address before reaching the merits of a claim. See Lexmark Int'l, Inc. v. Static Control Components, Inc., 572 U.S. 118, 128 n.4 (2014); Bell v. Hood, 327 U.S. 678, 682 (1946) (“[T]he failure to state a proper cause of action calls for a judgment on the merits and not for a dismissal for want of jurisdiction.”). The cause-of-action issue is therefore subject to ordinary principles of waiver and forfeiture.

In any event, Plaintiffs have a valid ultra vires cause of action. In Trump v. Illinois, the plaintiffs advanced an ultra vires claim similar to the one Plaintiffs make here, and in seeking a stay from the Supreme Court, Defendants similarly cited Nuclear Regulatory Commission to argue for greater deference to the President's decision to federalize the National Guard. See Application to Stay the Order Issued by the United States District Court for the Northern District of Illinois at 26–27, 146 S. Ct. 432 (2025) (No. 25A443), 2025 WL 2959866. If Defendants' and the dissent's theory were correct, there would have been no basis for an injunction in that case, and the Court should have granted a stay. Yet the Court denied Defendants' motion for a stay without finding it necessary to address either Defendants' theory of deference or whether the plaintiffs had a cause of action. See Illinois, 146 S. Ct. at 433–34. We have been rather dramatically admonished that the Supreme Court's stay orders, even when not fully explained, should “inform” our resolution of “like cases.” Trump v. Boyle, 145 S. Ct. 2653, 2654 (2025) (per curium); Nat'l Insts. of Health v. Am. Pub. Health Ass'n, 145 S. Ct. 2658, 2663 (2025) (Gorsuch, J., concurring in part and dissenting in part) (“Lower court judges may sometimes disagree with this Court's decisions, but they are never free to defy them.”). We disagree with the dissent's apparent view that Supreme Court simply overlooked a dispositive issue that should have led it to the opposite result in Illinois.

Further, our understanding of Illinois is consistent with the relevant precedent. As the dissent explains, where, as here, there is no applicable waiver of sovereign immunity, a plaintiff generally has no right to sue, but there are “two narrow exceptions—review of statutory ultra vires claims and review for claimed violations of the Constitution.” Dissent at 81-82 (citing Murphy Co. v. Biden, 65 F.4th 1122, 1128-29 (9th Cir. 2023)). The dissent argues that Plaintiffs' claim does not fall within the first exception because their ultra vires cause of action does not meet the standard established in Leedom v. Kyne, 358 U.S. 184 (1958), and clarified in Nuclear Regulatory Commission, 605 U.S. 665. Dissent at 82-87.

But it is unnecessary to decide whether the Kyne standard is met here because Plaintiffs' ultra vires claim—in addition to their Tenth Amendment claim—is a constitutional claim falling within the second exception. The plaintiffs in Nuclear Regulatory Commission argued the Commission exceeded the authority granted to it by the Atomic Energy Act, but they did not argue that the agency exercised that authority in an unconstitutional manner. 605 U.S. at 673–74. The Supreme Court addressed the standard for determining when nonstatutory judicial review is available for claims alleging that executive action exceeded only statutory authority, but it had no opportunity to pass on the judicial-review standard that applies to claims alleging that executive actions exceeded both statutory and constitutional authority. 11

As we explained in Murphy Co., when the plaintiff alleges that the President had neither statutory authority nor background constitutional authority to take the challenged action, their ultra vires claim is a constitutional claim:

Contemporary Ninth Circuit jurisprudence weighs in favor of justiciability by taking an expansive view of the constitutional category of claims highlighted in Dalton. In Sierra Club v. Trump, for example, we held that a challenge to presidential action will be considered constitutional, and therefore justiciable under Franklin [v. Massachusetts, 505 U.S. 788 (1992)], so long as a plaintiff claims that the President has “violat[ed] ․ constitutional separation of powers principles” because the President's action lacked both “statutory authority” and “background constitutional authority.” 929 F.3d 670, 696-97 (9th Cir. 2019); see also Sierra Club v. Trump, 963 F.3d 874, 889-90 (9th Cir. 2020) (reiterating that claims alleging the President violated the Constitution by exceeding statutory authority are justiciable as constitutional claims), vacated and remanded on other grounds sub nom. Biden v. Sierra Club, ․ 142 S. Ct. 46, 210 L.Ed.2d 985 (2021).

65 F.4th at 1130.

We recognize that “an action taken by the President in excess of his statutory authority [does not] necessarily violate[ ] the Constitution.” Dalton, 511 U.S. at 473. In Dalton, “the Court stated that ‘every action by the President, or by another executive official, in excess of his statutory authority is [not] ipso facto in violation of the Constitution.’ ” Sierra Club, 929 F.3d at 696 (quoting Dalton, 511 U.S. at 472). But “[t]he Court did not say ․ that action in excess of statutory authority can never violate the Constitution or give rise to a constitutional claim. Statutory and constitutional [ultra vires] claims are not mutually exclusive.” Id. In other words, an ultra vires claim can be, but is not necessarily, a constitutional ultra vires claim. See Dalton, 511 U.S. at 472-74; see also id. at 476-77 (“The [plaintiff's] claim that the President exceeded his authority under the [act at issue] [wa]s not a constitutional claim, but a statutory one.”).

To state a constitutional ultra vires claim, a plaintiff must make “specific allegations regarding separation of powers.” Murphy Co., 65 F.4th at 1130. Here, Plaintiffs contend that the President's federalization order, in addition to exceeding his statutory authority under 10 U.S.C. § 12406, encroached on Congress's calling forth power. Plaintiffs are not “simply alleging that the President has exceeded his statutory authority,” Dalton, 511 U.S. at 473, and do not advance a “typical statutory-authority argument” that must fit within the Leedom v. Kyne framework. Nuclear Regul. Comm'n, 605 U.S. at 682. Because the Constitution gave Congress—not the President—the limited power to federalize a state's National Guard, the President has no background constitutional authority to federalize the Guard. Supra at 18-20. The President has the power to federalize the Guard only to the extent authorized by Congress. If, as Plaintiffs claim, the President's federalization order exceeded that statutory authority, then the President necessarily encroached on Congress's authority under a specific provision of the Constitution and thus violated the separation of powers. See Youngstown Sheet & Tube Co., 343 U.S. at 588-89; see also id. at 637 (Jackson, J., concurring) (“When the President takes measures incompatible with the expressed or implied will of Congress, his power is at its lowest ebb, for then he can rely only upon his own constitutional powers minus any constitutional powers of Congress over the matter.”). Plaintiffs' ultra vires claim is best characterized as “constitutional and therefore reviewable.” Murphy Co., 65 F.4th at 1130; see Sierra Club, 929 F.3d at 696–97 (recognizing a cause of action because plaintiffs' claim—“that to the extent [d]efendants did not have statutory authority to reprogram funds they acted in violation of constitutional separation of powers principles because [d]efendants lack any background constitutional authority to appropriate funds”—was “fundamentally a constitutional one”).

The dissent argues that Murphy Co. is relevant only when the challenged action is “permitted by the authorizing statute but prohibited by a second statutory provision.” Dissent at 85-86. The dissent misreads Murphy Co. and muddles the distinct standards for (1) nonstatutory judicial review of a constitutional claim (including a constitutional ultra vires claim), (2) nonstatutory judicial review of a statutory ultra vires claim, and (3) nonstatutory judicial review of an executive action taken pursuant to a statute that grants the President discretion. Recall that Dalton distinguished between statutory and constitutional ultra vires claims. 511 U.S. at 472-74. After the Court held that Dalton's claim was a statutory ultra vires claim, not a constitutional one, the Court considered whether the statutory claim was reviewable. 511 U.S. at 474. The Court assumed that “some claims that the President has violated a statutory mandate are judicially reviewable outside the framework of the [Administrative Procedure Act],” meaning, the Court assumed nonstatutory judicial review was available. Id. But because the statute that granted the President authority to act “d[id] not by its terms circumscribe the President's discretion,” the Court concluded that the President's decision was “not reviewable” for abuse of discretion. See id. at 470, 474-476. See also Stay Order, 141 F.4th at 1047 (discussion Dalton).

Dalton's recognition of two distinct forms of ultra vires actions (constitutional and statutory) traces back to Larson v. Domestic & Foreign Commerce Corp., which acknowledged that an action may be ultra vires if it is “not within the officer's statutory powers or, if within those powers, only if the powers, or their exercise in the particular case, are constitutionally void.” 337 U.S. 682, 702 (1949). See also Dalton, 511 U.S. at 472 (explaining Larson “specif[ied] unconstitutional and ultra vires conduct as separate categories”). Starting in Leedom, 358 U.S. 184, the Supreme Court has placed strict limits on the availability of nonstatutory judicial review of statutory ultra vires claims. See Nuclear Regul. Comm'n, 605 U.S. at 681–82. But the Supreme Court has not extended those limits to claims of unconstitutional action. See Youngstown, 343 U.S. 579. Instead, the Supreme Court has repeatedly recognized claims brought by plaintiffs invoking an “implied private right of action directly under the Constitution to challenge governmental action under ․ separation-of-powers principles.” Free Enter. Fund v. Pub. Co. Accounting Oversight Bd., 561 U.S. 477, 491 n.2 (2010); see, e.g., Collins v. Yellen, 594 U.S. 220, 245 (2021).

In Murphy Co., we held that the plaintiff satisfied the requirements for both nonstatutory judicial review of constitutional ultra vires claims and nonstatutory judicial review of statutory ultra vires claims because the plaintiff adequately alleged that the challenged action was unconstitutional and, relatedly, that it violated an independent statute. First, as noted above, we explained that “specific allegations regarding separation of powers may suffice” to qualify an ultra vires claim as constitutional and therefore reviewable. 65 F.4th at 1129-30 (relying on Franklin, which held that the President's actions may be “reviewed for constitutionality,” 505 U.S. at 801, and describing Dalton, 511 U.S. at 467-72, as “reaffirming the Franklin principle that ‘Presidential decisions are reviewable for constitutionality’ ”). Second, we explained that a plaintiff may overcome the stringent limits applicable to nonstatutory judicial review of a statutory ultra vires claim if the plaintiff alleges that the executive action independently violates another statute. Id. at 1130–31. We did not hold that alleging such a secondary statutory violation is required to state a constitutional ultra vires claim. Rather, we held that the plaintiff's allegations in Murphy Co. satisfied both standards. See 65 F.4th at 1129-30 (“[W]hether characterized as ultra vires or constitutional, the result is the same: we resolve that Murphy's claims against the President ․ are justiciable.”). Murphy Co.'s alternative holding regarding nonstatutory judicial review of a statutory ultra vires claim aligns with Nuclear Regulatory Commission's recognition that such review may be available when the executive acts “ ‘contrary to a specific prohibition’ in a statute.” 605 U.S. at 681 (citation omitted).

Additionally, because the plaintiff in Murphy Co. alleged that the President's discretionary action violated a second statute, our conclusion that the action was reviewable was consistent with Dalton's second holding that, when a statute grants the President discretion to act, nonstatutory judicial review is available only when the discretion is subject to a reviewable limit. See Murphy Co., 65 F.4th at 1129 (noting the plaintiff argued that the second statute provided a “reviewable limit” on the President's discretion). We did not hold that the reviewable limit must be supplied by a second statute. See id. Nor did we hold that the reviewable limit must be a specific prohibition. 12

Thus, the dissent errs in suggesting that Murphy Co. held that a constitutional ultra vires claim is reviewable only when a President's action contravenes an independent statute or specific prohibition.

Here, as in Sierra Club, Plaintiffs pursue the constitutional route, alleging not only that the President exceeded the scope of his statutory authority but also that the President's action lacked any “background constitutional authority.” Murphy Co., 65 F.4th at 1130 (quoting Sierra Club, 929 F.3d at 696–97). The President thus acted in the absence of any authority and violated “constitutional separation of powers principles.” Id. (quoting Sierra Club, 929 F.3d at 696–97). That is a constitutional claim that we may review. Id.; Free Enter. Fund, 561 U.S. at 491 n.2; Cook, 146 S. Ct. at 2251 n.2.

E. The June 2025 Federalization Order

We next consider whether the President exceeded the limits of his statutory grant of authority under § 12406 when he ordered the federalization of the California National Guard on June 7, 2025. Defendants assert that two of the required exigencies existed here: there was “a rebellion or danger of a rebellion against the authority of the Government of the United States” under § 12406(2), and the President was “unable with the regular forces to execute the laws” under § 12406(3).

1. Section 12406(3): “Unable with the Regular Forces to Execute the Laws”

a. Meaning of § 12406(3)

Section 12406(3) provides that the President may call forth a state's militia “[w]henever ․ the President is unable with the regular forces to execute the laws of the United States.” In Illinois, the Court concluded:

[T]he term “regular forces” in § 12406(3) likely refers to the regular forces of the United States military. This interpretation means that to call the Guard into active federal service under § 12406(3), the President must be “unable” with the regular military “to execute the laws of the United States.” Because the statute requires an assessment of the military's ability to execute the laws, it likely applies only where the military could legally execute the laws. Such circumstances are exceptional: Under the Posse Comitatus Act, the military is prohibited from “execut[ing] the laws” “except in cases and under circumstances expressly authorized by the Constitution or Act of Congress.” 18 U.S.C. § 1385. So before the President can federalize the Guard under § 12406(3), he likely must have statutory or constitutional authority to execute the laws with the regular military and must be “unable” with those forces to perform that function.

146 S. Ct. at 434. Unless the Government makes both threshold showings, it “has not carried its burden to show that § 12406(3) permits the President to federalize the Guard.” Id.

Although Illinois “reflects only likely legal conclusions and comes at a ‘preliminary stage’ ” of that case, it is “a form of precedent throughout the Federal Judiciary that requires a temporary injunction (either a preliminary injunction or an injunction pending appeal or certiorari) in like circumstances, unless and until this Court reaches a contrary conclusion.” Id. at 436 n.3 (Kavanaugh, J., concurring in the judgment). Because this case involves like circumstances, we treat Illinois as binding precedent.

The Supreme Court did not explain why it concluded that “regular forces” likely means “the regular forces of the United States military.” However, applying the usual tools of statutory interpretation, we reach the same conclusion. We start with the text and “interpret the words consistent with their ‘ordinary meaning ․ at the time Congress enacted the statute.’ ” Wis. Cent. Ltd. v. United States, 585 U.S. 274, 277 (2018) (quoting Perrin v. United States, 444 U.S. 37, 42 (1979)). Congress first added the term “regular forces” to the statutory predecessor of § 12406 in 1908. Compare 1903 Militia Act § 4, 32 Stat. at 776 (President may call forth state militia members when he is “unable, with the other forces at his command, to execute the laws” (emphasis added)) with Militia Reform Act of 1908 § 4, 35 Stat. at 400 (President may call forth state militia members when he is “unable with the regular forces at his command to execute the laws” (emphasis added)).

“[W]e may refer to dictionary definitions to help determine” the “plain meaning” of statutory text. United States v. Saini, 23 F.4th 1155, 1161 (9th Cir. 2022). In 1908, legal dictionaries defined “regular forces” or “forces” to mean military forces. For example, the second edition of the Judicial Dictionary, published in 1903, defined “regular forces” only by referring the reader to the entry for “military forces.” See Regular Forces, 3 The Judicial Dictionary 1696 (F. Stroud ed., 2d ed. 1903) (“V. military forces”). In turn, the Judicial Dictionary defined “military forces” as “officers and soldiers who ․ are liable to render continuously for a term military service.” Military Forces, 2 Judicial Dictionary 1198 (F. Stroud ed., 2d ed. 1903) (citation modified); see Forces, Bouvier's Law Dictionary 428 (William Edward Baldwin ed., Baldwin's Century ed. 1926) (defining “regular forces” as “those persons who are impressed with military character by the national authority, to such, in short, as constitute the regular army and navy, (now include the aviation forces)”); see also Forces, Black's Law Dictionary 503 (1st ed. 1891) (“The military and naval power of the country.”); Forces, 1 Bouvier's Law Dictionary and Concise Encyclopedia 1255 (8th ed. 1914) (“The military and naval resources of a country.”).

Likewise, multiple dictionaries from that time period defined “regular” as describing the standing army or a member of the standing army. See, e.g., Regular, 8 A New English Dictionary on Historical Principles 379 (James A.H. Murray & W.A. Craigie eds.,1914) (“Of forces or troops: Properly and permanently organized; constituting the standing army․”); Regular, 8 The Century Dictionary and Cyclopedia 5050 (William Dwight Whitney & Benjamin E. Smith eds., rev. & enl. ed. 1911) (“permanent; standing; opposed to volunteer”); Regular, A Military Dictionary 583 (William Duane ed., Philadelphia, W. Duane 1810) (“Regular, when applied to the army, signifies those troops that are inlisted for a regular period, do duty as soldiers and nothing else; contradistinguished from those who are citizens occasionally exercising the duties of soldiers.”). To our knowledge, no dictionary defined “regular forces” to include civilian law enforcement.

“[S]tatutory structure and history” also inform our interpretation. Niz-Chavez v. Garland, 593 U.S. 155, 165 (2021). These tools confirm that “regular forces” means the regular forces of the U.S. military.

Congress first added an inability “to execute the laws” clause to a predecessor of § 12406 in 1903. See 1903 Militia Act § 4, 32 Stat. at 776. Between 1903 and 1908, the clause stated that the President may call forth state militia members when he is “unable, with the other forces at his command, to execute the laws.” Id. (emphasis added). “The other forces” likely referred to the United States military forces because the 1903 Militia Act primarily addressed the relationship between those forces and the state militias. See generally 32 Stat. 775.

Other sections of the 1903 Militia Act used the term “regular” to refer to the United States military forces. See, e.g., § 9, 32 Stat. at 776 (“[T]he militia, when called into the actual service of the United States, shall be subject to the same Rules and Articles of War as the regular troops of the United States.” (emphasis added)). See also §§ 3, 10, 14, 15, 16, 18, 32 Stat. at 775–78. Congress “generally uses a particular word with a consistent meaning in a given context.” Graham Cnty. Soil & Water Conservation Dist. v. United States ex rel. Wilson, 559 U.S. 280, 290 (2010) (citation modified). Thus, the statutory context further confirms that in 1908, when Congress amended § 4 to replace “the other forces” with “the regular forces,” it intended to clarify that it was referring to the United States military forces.

The codification history of the relevant provisions also supports this understanding of section 12406(3). Section 4 of the 1908 Militia Act was codified as 32 U.S.C. § 81a in 1940. Congress recodified the provision in 1956 when Title 10 was enacted into positive law. See Acts of Aug. 10, 1956, ch. 1041, § 49, 70A Stat. 640; see also 70A Stat. 199 (1956) (codified at 10 U.S.C. § 3500 (1958)). Title 10 governs the military's structure and operations. Throughout Title 10, statutes use the term “regular” to refer to members or components of the armed forces, often to distinguish “regular” from “reserve” components. See, e.g., 10 U.S.C. § 12502(a) (“The President may detail a regular or reserve officer of the Army as chief of staff.”); id. § 12501 (“The Secretary concerned shall detail such members of the regular and reserve components under his jurisdiction as are necessary to effectively develop, train, instruct, and administer those reserve components.”). 13

When Congress recodified the calling forth provision in 1956, it amended the provision to remove the words “at his command” from the phrase, “regular forces at his command.” 70A Stat. 199 (1956). The Historical and Revision Notes indicate that those words were “omitted as surplusage.” 10 U.S.C. § 3500 (1958) (Historical and Revision Notes). Congress clarified in the recodification statute that “it [wa]s the legislative purpose to restate, without substantive change,” the recodified provisions. § 49(a), 70A Stat. 640 (1956). Section 3500 was recodified as § 12406 in 1994, see National Defense Authorization Act for Fiscal Year 1995, Pub. L. 103-337, 108 Stat. 2663, 2994 (1994), but the text has not changed since 1956.

Thus, the plain language, context, and statutory history of § 12406(3) convince us that the term “regular forces” continues to mean the regular U.S. military forces. 14

b. June 7, 2025 “Unable with the Regular Forces” Determination

“[B]efore the President can federalize the Guard under § 12406(3), he likely must have statutory or constitutional authority to execute the laws with the regular military and must be ‘unable’ with those forces to perform that function.” Illinois, 146 S. Ct. at 434. In other words, the President cannot federalize the Guard under § 12406(3) unless he is legally able to execute the laws with the regular military but factually unable to do so under the circumstances. Further, because the Posse Comitatus Act generally prohibits use of the military “to execute the laws” “except in cases and under circumstances expressly authorized by the Constitution or Act of Congress,” 18 U.S.C. § 1385, Defendants must identify a source of authority that would have allowed the military to “execute the laws” in California when the June 2025 federalization order was issued. Illinois, 146 S. Ct. at 434.

Defendants argue that § 12406 itself authorizes the President to use the military to execute the laws. We conclude that it does not.

To begin, the Supreme Court impliedly rejected that argument in Illinois. There as here, the President invoked § 12406 to federalize the Guard. Illinois, 146 S. Ct. at 434. 15 The Supreme Court concluded that the “President has not invoked a statute that provides an exception to the Posse Comitatus Act.” Id. 16 Examining the text of § 12406, we reach the same conclusion. Section 12406 expressly authorizes the President to “call into Federal service members and units of the National Guard of any State in such numbers as he considers necessary to ․ execute th[e] laws” of the United States. It does not expressly authorize the President to “use” the military to execute those laws.

Defendants argue that the authorization to call the Guard into federal service “in such numbers” as the President “considers necessary to ․ execute” the laws must include the authorization to use the Guard to execute the laws. We disagree. “ ‘[S]tatutes addressing the same subject matter’ generally should be interpreted consistently with each other,” especially when “the language in an earlier act is the same as, or similar to, the language in the later act.” Advanced Integrative Med. Sci. Inst. v. Garland, 24 F.4th 1249, 1256 (9th Cir. 2022) (quoting Wachovia Bank v. Schmidt, 546 U.S. 303, 305 (2006)). The Insurrection Act addresses the same subject matter as § 12406—the federalization of state militias—and both include provisions that can be traced back to the Militia Act of 1792.

In the Insurrection Act, Congress has authorized the President to “call into Federal service” the militia of any State when insurrection makes it “impracticable to enforce the laws”—and has separately authorized the President to “use such of the armed forces ․ to enforce those laws.” 10 U.S.C. § 252. If § 252's express authorization to federalize the state militias when “necessary to enforce the laws” included the authorization to use those militias to enforce the laws, then the clause that expressly authorizes such use would be superfluous. It is “one of the most basic interpretive canons that a statute should be construed so that effect is given to all its provisions, so that no part will be inoperative or superfluous, void or insignificant.” Corley v. United States, 556 U.S. 303, 314 (2009) (citation modified). Accordingly, in the context of § 252 of the Insurrection Act, we conclude that the authorization of federalization does not include authorization of use.

Because both the Insurrection Act and § 12406 address the federalization of state militias, we presume that the authorization of federalization under § 12406, like the authorization of federalization under § 252, does not include authorization of use. And because § 252 expressly authorizes both federalization and use of the militia to execute the laws, but § 12406 expressly authorizes only federalization, we presume that Congress intended not to authorize such use under § 12406. See Russello v. United States, 464 U.S. 16, 23 (1983) (“Where Congress includes particular language in one section of a statute but omits it in another section of the same Act, it is generally presumed that Congress acts intentionally and purposely in the disparate inclusion or exclusion.” (alteration accepted) (quoting United States v. Wong Kim Bo, 472 F.2d 720, 722 (5th Cir. 1972))).

Defendants also argue that the “President has an ‘inherent’ protective and emergency power derived from the Take Care Clause” of the Constitution that authorizes him to use the military to protect federal personnel and property. Even if that is correct, it does not mean that the Take Care Clause provides an exception to the Posse Comitatus Act. By Defendants' own admission, this protective power is implied, not express—and the Act prohibits use of the military to execute the laws unless such use is “expressly authorized by the Constitution or an Act of Congress.” 18 U.S.C. § 1385 (emphasis added).

The Supreme Court in Illinois rejected Defendants' argument that, when the military carries out a protective mission pursuant to the President's constitutional authority to protect federal personnel and property, it is not “execut[ing] the laws” within the meaning of the Posse Comitatus Act. See Illinois, 146 S. Ct. at 434. The Court concluded that the President could not rely on “inherent constitutional authority that, according to the Government, allows him to use the military to protect federal personnel and property” because, if “performing such protective functions does not constitute ‘execut[ing] the laws’ within the meaning of the Posse Comitatus Act,” then “it is hard to see how performing those functions could constitute ‘execut[ing] the laws’ under § 12406(3).” Id. The same reasoning applies here.

Because Defendants have failed to identify a source of authority that would allow the military to legally execute the laws in California, they have not carried their burden to show that the President had the authority to federalize California's National Guard under § 12406(3).

2. Section 12406(2): “A Rebellion or Danger of a Rebellion”

Defendants argue that the President's action was independently warranted under 10 U.S.C. § 12406(2).

a. Meaning of § 12406(2)

Section 12406(2) authorizes the President to call forth a state's militia “[w]henever ․ there is a rebellion or danger of a rebellion against the authority of the Government of the United States.” Because the statutory text uses the present tense, we conclude that § 12406 permits federalization when the asserted exigency exists at the time the order is issued. See United States v. Wilson, 503 U.S. 329, 333 (1992) (“Congress' use of a verb tense is significant in construing statutes.”). This means that our review must focus on the conditions that existed when the federalization order was issued, not conditions that existed in the past. 17

In our Stay Order, in the context of defining “unable to execute the laws,” we commented that “rebellions” and “invasions” are “unusual and extreme exigencies ․ that threaten the normal operations of civil government.” Stay Order, 141 F.4th at 1051. However, we did not reach the merits of the parties' dispute regarding the meaning of “rebellion.” Id. We do so here.

The district court interpreted “rebellion” as “a violent, armed, organized, open and avowed uprising against the government as a whole.” TRO, 786 F. Supp. 3d at 1253. Defendants argue that the district court's interpretation is too narrow. For the following reasons, we continue to conclude that a “rebellion” is an unusual and extreme exigency, and we agree with the district court's interpretation.

We begin once more with the plain text of the statute, and with dictionary definitions of the disputed term. Congress added “rebellion” as a predicate condition to the “calling forth” provision in 1903. See Militia Act of 1903, § 4. Dictionaries from that time period defined “rebellion,” either exclusively or primarily, as an open, organized, and armed renunciation of and resistance to an established government. See, e.g., Rebellion, Webster's International Dictionary of the English Language (1903) (“The act of rebelling; open and avowed renunciation of the authority of the government to which one owes obedience, and resistance to its officers and laws, either by levying war, or by aiding others to so; an organized uprising of subjects for the purpose of coercing or overthrowing their lawful ruler or government by force; revolt; insurrection.”); Rebellion, The Cyclopedic Dictionary of Law (1901) (“The taking up of arms traitorously against the government; the forcible opposition and resistance to the laws and process lawfully issued.”); Rebellion, An American Dictionary of the English Language (1897) (“An open and avowed renunciation of the authority of the government to which one owes allegiance; or the taking of arms traitorously to resist the authority of lawful government; revolt.”); Rebellion, Black's Law Dictionary (1st ed. 1891) (“Deliberate, organized resistance, by force and arms, to the laws or operations of the government, committed by a subject.”).

The primary definition of rebellion in all these dictionaries thus supports the district court's interpretation of rebellion as “a violent, armed, organized, open and avowed uprising against the government as a whole,” see TRO, 786 F. Supp. 3d at 1253, rather than opposition to the government's handling of a single law or issue. Indeed, one historical dictionary distinguished “rebellion” from “insurrection” on that very ground: It explained that “[r]ebellion differs from insurrection” because “[i]nsurrection may be a rising in opposition to a particular act or law, without a design to renounce wholly all subjection to the government.” Rebellion, An American Dictionary of the English Language (1897). See also id. (“Insurrection may be, but is not necessarily, rebellion.”). 18

Defendants cite the latest edition of Black's Law Dictionary, but we are guided by the common understanding of “rebellion” when the statute was enacted, rather than by its contemporary meaning. Defendants also ask us to ignore the primary definition of “rebellion” and to instead rely on the broader secondary and tertiary definitions. Two historical dictionaries provided a secondary definition of “rebellion” as open resistance to lawful authority. See Rebellion, An American Dictionary of the English Language (1897) (“Open resistance to lawful authority.”); Rebellion, Webster's International Dictionary of the English Language (1903) (“Open resistance to, or defiance of, lawful authority.”). Thus, Defendants' interpretation would permit federalization whenever there is “open resistance to lawful authority,” including resistance that is not violent, armed, organized, or against the government as a whole.

The plain language of Section 12406 forecloses Defendants' broad interpretation. Section 12406 authorizes the President to federalize members only when such action is “necessary” to “suppress the rebellion.” Id. The term “necessary” indicates that Congress intended to authorize federalization only when the conditions require such military intervention. Resistance that is not violent, armed, organized, and against the government as a whole generally does not require a military response.

Our interpretation of “rebellion” follows the “rule of statutory interpretation that a law's terms are best understood by “the company [they] kee[p],” and the Supreme Court's interpretation in Martin of § 12406's statutory predecessor, § 1 of the 1795 Militia Act. United States v. Taylor, 596 U.S. 845, 856 (2022) (quoting Gustafson v. Alloyd Co., 513 U.S. 561, 575 (1995)). The Court in Martin interpreted the 1795 Militia Act as granting the President power to call forth the militia “upon sudden emergencies, upon great occasions of state, and under circumstances which may be vital to the existence of the Union.” Martin, 25 U.S. at 30. Although “invasion” and “imminent danger of invasion” were the only predicate exigencies at that time, we conclude that “rebellion” also refers to a “great occasion[ ] of state” potentially “vital to the existence of the Union,” comparable to an “invasion.” 19 Martin, 25 U.S. at 29–30.

Statutory history reinforces our conclusion. Although Congress added “rebellion” to § 12406's predecessor statute in 1903, it had added “rebellion” to the Insurrection Act much earlier, in July 1861—shortly after the outbreak of the Civil War in April 1861. See Militia Act of 1903, ch. 196, § 4, 32 Stat. 775, 776; Militia Act of 1861, ch. 25, § 1, 12 Stat. 281, 281 (authorizing the President to call forth a state's militia “whenever, by reason of unlawful obstructions, combinations, or assemblages of persons, or rebellion against the authority of the Government of the United States, it shall become impracticable ․ to enforce, by the ordinary course of judicial proceedings, the laws of the United States” (emphasis added)). The timing suggests that Congress added the term “rebellion” to address the extraordinary exigency created by the Civil War. See Biden v. Texas, 597 U.S. 785, 804 (2022) (“The historical context in which [a] provision was adopted [can] confirm[ ] the plain import of its text.”). As noted, the Insurrection Act is related to § 12406, and courts should “not lightly assume that Congress silently attaches different meanings to the same term in the same or related statutes.” Azar v. Allina Health Servs., 587 U.S. 566, 574 (2019).

Defendants argue that the primary, narrow definition of “rebellion” would not “encompass numerous instances ․ in which the President has called the militia into federal service to address open defiance of federal authority in situations that fell short of organized efforts to overthrow the government.” But in Defendants' historical examples, the President invoked the Insurrection Act (including its statutory predecessors), rather than § 12406, and the Insurrection Act has always authorized the President to exercise the calling-forth power to address exigencies falling short of “rebellion.” 20 For example, President Washington's 1794 proclamation calling forth the militia to quell violent opposition to the enforcement of a federal excise tax on alcohol (later dubbed the “Whiskey Rebellion”) invoked Section 2 of the Militia Act of 1792. 21 Section 2 of the 1792 Act authorized the President to call forth a state's militia “[w]henever the laws of the United States shall be opposed or the execution thereof obstructed, in any state, by combinations too powerful to be suppressed by the ordinary course of judicial proceedings.” Militia Act of 1792, ch. 28, § 2, 1 Stat. 264, 264. 22 No court held that the President acted without statutory authority in these historical examples, and we have no reason to believe that Congress added “rebellion” to the Insurrection Act and § 12406 to encompass them.

Further, even if President Washington's invocation of Section 2 of the 1792 Militia Act were to inform our interpretation of a term added to the Insurrection Act and § 12406 decades later, that example supports our conclusion that “rebellion” means an open, organized, and armed renunciation of and resistance to the federal government that requires military intervention. Violent incidents had been occurring and escalating for over two years before President Washington invoked his statutory calling forth power in 1794. See Michael Hoover, The Whiskey Rebellion, U.S. Department of the Treasury, Alcohol and Tobacco Tax and Trade Bureau (last updated Jan. 9, 2020); Leland D. Baldwin, Whiskey Rebellion, in 8 Dictionary of American History 469 (Stanley I. Kutler ed., 3d ed. 2003). President Washington did not conclude a military response was required and call forth the militia until several thousand armed rebels had gathered outside Pittsburgh in a coordinated effort “to burn and plunder the[ ] town,” H.M. Brackenridge, History of the Western Insurrection in Western Pennsylvania 99 (1794), and state authorities had declined to respond, Gregory J. W. Urwin, The Army of the Constitution: The Historical Context, in Papers From the Conference on Homeland Protection 50 (Max G. Manwaring ed., 2000). 23 The conditions facing President Washington were not only unusual and extreme, but also constituted open, organized, and armed opposition to the federal government requiring military intervention.

Finally, we note that the Supreme Court in Illinois implicitly rejected the government's reading of “rebellion.” In their application for a stay in that case, Defendants made an argument essentially identical to the one they present here, asserting that “[t]he violence targeted at federal personnel and property in Illinois in recent weeks also constitutes a ‘rebellion or danger of rebellion’ under the ordinary meaning of those terms, which encompass violent opposition to federal authorities' enforcement of particular federal laws.” Application for Stay at 4, No. 25A443, Illinois, supra (Oct. 17, 2025). If that argument were correct, it would have been an alternative basis on which Defendants could have prevailed in their appeal from the preliminary injunction entered by the district court in that case, and, thus, an alternative basis for the Supreme Court to grant a stay. Yet the Court evidently deemed the argument so lacking in merit as not to be worth mentioning in its opinion denying a stay.

b. June 7, 2025 “Rebellion” Determination

We now consider the district court's factual findings and assess whether the President's determination that there was “a rebellion or danger of a rebellion” in Los Angeles as of June 7, 2025, “reflects a colorable assessment of the facts and law within a ‘range of honest judgment.’ ” Stay Order, 141 F.4th at 1051 (quoting Sterling, 287 U.S. at 399). 24

The district court found:

Some members of the public gathered in protest. A group of people assembled at the site of an ICE operation in Los Angeles's Garment District and tried to prevent ICE from leaving. Another one of the primary protests on June 6 took place at the Metropolitan Detention Center. The protests were explicitly about the immigration raids. The Governor's Office of Emergency Services, which was monitoring the protests, was in regular communication with representatives of the Los Angeles Police Department and the Los Angeles County Sheriff's Department. Both stated that they did not need additional resources, and LAPD reassigned additional officers to the area.

That evening, protesters reportedly marched in downtown Los Angeles. There were about 800 protesters at two sites. Some protesters threw “concrete chunks, bottles of liquid, and other objects at Federal Protective Service officers guarding a parking lot gate; some protesters attempted “to use large rolling commercial dumpsters as a battering ram.” Officers protected the gate entrance with pepper balls and other nonlethal force, until LAPD arrived and pushed the crowd away. Some of the protesters used “chairs, dumpsters, and other items as weapons.” LAPD declared an “unlawful assembly,” and while some individuals resisted, the protesters departed by 11:00 p.m. Two federal buildings were vandalized and sustained minor damage.

On June 7, both the ICE operations and responsive protests continued. In addition to the protest at the Metropolitan Detention Center, protests also emerged in Paramount and Compton. LASD provided 200 deputies to respond to the Paramount and Compton protests, including a team with specialized training in handling civil unrest. Customs and Border Protection officers arrived from San Diego to assist with immigration enforcement operations.

During the evening hours of June 7, the protest at the Metropolitan Detention Center was “fairly in control,” with some protesters targeting the building. Federal officers pushed protesters away while LAPD officers moved in front and declared an unlawful assembly; by 2:15 a.m., “most of the protestors had left the area.” In Paramount and Compton, there were about 300 to 400 protesters present. Some protesters threw rocks and other objects (including a Molotov cocktail), burned a vehicle, looted a gas station, and vandalized property. There was an extended clash between some protesters and officers, with a crowd boxing in the officers and throwing fireworks, rocks, and mangos at them, and “using shopping carts to barricade the street.” One officer was briefly trapped inside her law enforcement vehicle when a crowd surrounded it, shook it, and threw stones at it. A DHS fence was damaged, and three government vehicles were damaged. Local law enforcement brought the situation under control by 4:00 a.m. and LASD was able to demobilize its teams. About 11 people were arrested for engaging in unlawful behavior at the protests that night.

TRO, 786 F. Supp. 3d at 1242–43 (record citations omitted). Although the district court found it “beyond debate that some individuals used the protests as an excuse for violence and destruction,” it also found there could “be no debate that most protesters demonstrated peacefully.” Id. at 1253. 25

As of the June 12 hearing before the district court, there was no evidence “of any protesters carrying firearms,” and “no evidence of any organized (as opposed to sporadic) violence.” Id. at 1245. “Nor [wa]s there evidence that any of the violent protesters were attempting to overthrow the government as a whole; the evidence is overwhelming that protesters gathered to protest a single issue—the immigration raids.” Id. at 1253. 26

On this record, a determination that the conditions in Los Angeles on June 7, 2025, constituted “a rebellion or danger of rebellion” does not reflect a colorable assessment within a range of honest judgment. We are mindful that the President's range of discretion in this context is necessarily broad. See Sterling, 287 U.S. at 400. Even so, the “limits” of that discretion may be “overstepped in a particular case.” Id. at 400–01. Here, while many individuals peacefully protested against the federal government's handling of immigration issues, there were bad actors who committed serious breaches of the peace and acts of violence. But the district court found there was “no evidence of organized, as apart from sporadic or impromptu, violence.” TRO, 786 F. Supp. 3d at 1253. Moreover, hundreds of civilian law enforcement officers, “including a team with specialized training in handling civil unrest,” responded to the protests, “brought the situation under control,” and arrested people who engaged in unlawful behavior. Id. at 1243. Compare Sterling, 287 U.S. at 391, 403–04 (affirming district court's finding that there was no danger of “actual riot, tumult, or insurrection” because, even “ ‘if all of the [predicted] conditions had come to pass, they would have resulted merely in breaches of the peace, to be suppressed by the militia as a civil force, and not at all in a condition constituting, or even remotely resembling, a state of war’ ” (quoting district court order, Constantin v. Smith, 57 F.2d at 231)).

* * *

In sum, on this record, we conclude that neither § 12406(2) nor § 12406(3) authorized the President to federalize members of California's National Guard on June 7, 2025. Therefore, we agree with the district court that Plaintiffs are likely to succeed on their claims that the Defendants' actions were ultra vires. 27

3. The Tenth Amendment

The parties agree that Plaintiffs' Tenth Amendment claim rises and falls with their ultra vires claim. 28 Indeed, Defendants argued below that Plaintiffs' Tenth Amendment claim “is wholly derivative” of their § 12406 claim, and the district court concluded that, because Plaintiffs are likely to succeed on the merits of their ultra vires claim, they are also likely to succeed on the merits of their Tenth Amendment claim. TRO, 786 F. Supp. 3d at 1259-60. Defendants do not challenge this reasoning on appeal. 29

The dissent, however, disputes the parties' shared understanding that Plaintiffs' Tenth Amendment claim rises and falls with their ultra vires claim. We briefly explain why the parties are correct.

The federal government—whether the President or Congress—violates the Tenth Amendment by commandeering state personnel to enforce or administer federal law. Printz v. United States, 521 U.S. 898, 925 (1997) (Under the Tenth Amendment, “the Federal Government may not compel the States to implement, by legislation or executive action, federal regulatory programs.”). If, for example, the President commandeered state police to help federal officers enforce federal immigration laws, that would violate the Tenth Amendment. See United States v. California, 921 F.3d 865, 889-90 (9th Cir. 2019).

Plaintiffs contend that effectively happened here. California National Guard members are state personnel, unless properly federalized. See 10 U.S.C. § 12401. The President purported to convert the Guard into federal personnel by issuing federalization orders. If those federalization orders had been valid, then the President would have deployed federal officers in Los Angeles, and no commandeering of state personnel would have occurred. But because the President lacked authority to federalize the California National Guard, the federalization orders were invalid, and the deployed Guard members remained state personnel. See Complaint ¶ 99, 2025 WL 1640287 (“[T]he federal government is not merely intruding on the province of the State, but doing so by taking command of the State's own resources, the California National Guard, which remains under State control unless properly federalized. Proper federalization has not happened here.”). 30

In other words, because the deployed members of the California National Guard were not properly federalized, they remained the legal equivalent (for Tenth Amendment purposes) of the state police. And because the challenged executive orders compelled those state officers to help federal officers enforce federal law, the orders violated the Tenth Amendment.

The dissent argues that even if the President's federalization order exceeded his statutory authority, the order would not violate the Tenth Amendment because it would infringe Congress's authority, not California's, and the Constitution expressly empowers Congress to federalize the National Guard. Dissent at 91-96. In other words, the dissent argues that, because Congress could have authorized the President to issue the challenged federalization orders, there cannot be a Tenth Amendment violation. 31

The Tenth Amendment protects States against overreach by the “Federal Government,” whether “by legislation or executive action.” Printz, 521 U.S. at 925, 935. The President cannot command state officials to enforce federal law even if a federal statute would authorize him, under some other circumstances, to call those officials into federal service. If the statute did not authorize the President to federalize the California Guard members here (as we conclude and the dissent assumes), then the federalization orders were invalid, the Guard members remained state personnel, and Defendants violated the Tenth Amendment by commandeering those state personnel to administer federal immigration laws. 32

Defendants are therefore correct that Plaintiffs' Tenth Amendment claim rises and falls with their ultra vires claim. And because Plaintiffs are likely to succeed on the merits of their ultra vires claim, they are also likely to succeed on the merits of their Tenth Amendment claim.

F. Remaining Equitable Factors

In addition to the merits, we consider three other factors when assessing whether the district court's award of preliminary injunctive relief was appropriate: the possibility of irreparable harm in the absence of relief, the balance of equities, and the public interest in an injunction. Winter, 555 U.S. at 20.

1. Irreparable Injury

As we recognized in our Stay Order, Plaintiffs have “significant interests” in preventing the “upset [of the] constitutional balance of power between federal and state government.” 141 F.4th at 1055. Plaintiffs allege that the challenged federalization orders infringe California's sovereign rights and Tenth Amendment interests.

When plaintiffs allege a constitutional violation and injury and show they are likely to prevail on the merits, that showing may demonstrate irreparable harm even if the violation is brief. See Melendres v. Arpaio, 695 F.3d 990, 1002 (9th Cir. 2012). Although Defendants argue that we concluded in our Stay Order that Plaintiffs' purported harms are “speculative,” 141 F.4th at 1055, they quote this language out of context. We used that term in response to Plaintiffs' argument that “the continued presence of National Guard members in Los Angeles risks worsening, not improving, tensions on the ground,” and said only that Plaintiffs' “concerns about escalation and interference with local law enforcement, at present, are too speculative.” Id. (internal quotation marks omitted). Nothing in our Stay Order suggests that all of Plaintiffs' asserted harms were speculative.

2. Balance of Equities

Although Defendants have articulated a legitimate interest in protecting federal personnel and property, that interest may not be exercised “beyond the bounds of their authority.” TRO, 786 F. Supp. 3d at 1263. Congress has not authorized the President to protect federal personnel and property by federalizing the California National Guard under the circumstances present here. Because the government “cannot suffer harm from an injunction that merely ends an unlawful practice,” the balance of equities favors Plaintiffs. Rodriguez v. Robbins, 715 F.3d 1127, 1145 (9th Cir. 2013).

3. The Public Interest

It is “always in the public interest to prevent the violation of a party's constitutional rights.” Riley's Am. Heritage Farms v. Elsasser, 32 F.4th 707, 731 (9th Cir. 2022) (quoting Melendres, 695 F.3d at 1002). Here, the public interest is served by upholding the district court's award of preliminary injunctive relief.

IV. CONCLUSION

For the reasons above, we AFFIRM the district court's order preliminarily enjoining Defendants from deploying members of California's National Guard in Los Angeles and directing Defendants to return control of the Guard to Governor Newsom.

I join the court's opinion in full. As the opinion explains, the argument that no ultra vires cause of action is available in this case has been forfeited twice over: Defendants did not raise it in the district court, and they did not raise it here until their reply brief. The opinion also correctly explains that our decision in Murphy Co. v. Biden dictates that such a cause of action is available. 65 F.4th 1122 (9th Cir. 2023). There is room to debate whether Murphy Co. is consistent with Supreme Court precedent. See, e.g., NRC v. Texas, 605 U.S. 665, 682 (2025) (describing an ultra vires claim as “a Hail Mary pass” that “rarely succeeds” (quoting Nyunt v. Chairman, Broad. Bd. of Governors, 589 F.3d 445, 449 (D.C. Cir. 2009))); Trump v. Illinois, 146 S. Ct. 432, 433 (2025) (per curiam) (declining to stay an injunction predicated on an ultra vires challenge to the federalization of the National Guard). But because we are bound by Murphy Co.—and because, to repeat, the argument has been forfeited—we need not engage in that debate here.

For the reasons stated in my dissent from the denial of Defendants' motion to dismiss, I would not decide this appeal. And I write separately here too because, respectfully, the majority makes at least three errors in affirming the preliminary injunction.

First, this appeal's unique procedural posture counsels against reaching the difficult and important questions that the majority opinion resolves. Second, the majority disregards the legal standard that governs statutory ultra vires claims. And third, the majority concludes that Plaintiffs are likely to succeed on the merits of their Tenth Amendment claim—a claim the parties have barely briefed and that I believe is unlikely to succeed.

I

After this appeal was fully briefed, the Supreme Court decided Trump v. Illinois, 146 S. Ct. 432 (2025). That case concerns a central issue in this appeal. And the Court's decision rested on grounds the parties here never raised.

Defendants reacted by moving to dismiss this appeal. As a result, we have received only scant briefing on the effect of this directly relevant intervening Supreme Court authority. But the majority insists on pressing forward, relying primarily on outdated briefs prepared amid a shifting legal landscape.

The circumstances under which the President can federalize members of a State's national guard is a question of tremendous consequence and national importance. But it is not a question that must—or should—be resolved by this panel today. Because this appeal comes to us in an interlocutory posture, we could have granted Defendants' motion to dismiss the appeal and sent Plaintiffs' case onward toward trial. The parties could have then litigated the effect of Trump v. Illinois before the district court. And with the benefit of that process, we could have addressed these difficult issues in turn after entry of a final judgment.

The majority instead takes a different route, issuing at this premature stage a published opinion that will bind future three-judge panels, absent intervening precedent from the Supreme Court or an en banc panel. In choosing that path, I believe the majority disregards the cardinal principle of judicial restraint. See Dobbs v. Jackson Women's Health Org., 597 U.S. 215, 348 (2022) (Roberts, C.J., concurring in the judgment) (“[A] simple yet fundamental principle of judicial restraint: If it is not necessary to decide more to dispose of a case, then it is necessary not to decide more.”).

II

The majority also errs by refusing to apply the legal standard that governs statutory ultra vires claims. “[I]n all but the most unusual circumstances, prescribing a cause of action is a job for Congress, not the courts ․” Egbert v. Boule, 596 U.S. 482, 486 (2022). And “[s]overeign immunity generally bars suits against the United States and its officials sued in their official capacity unless Congress has expressly waived immunity by statute.” Murphy Co. v. Biden, 65 F.4th 1122, 1128 (9th Cir. 2023) (citing Lane v. Pena, 518 U.S. 187, 192 (1996)); see also Minnesota v. United States, 305 U.S. 382, 387 (1939) (“The exemption of the United States from being sued without its consent extends to a suit by a state.”). Thus, a suit against federal officials is generally not justiciable unless Congress expressly provides otherwise by statute.

Congress has not done so here. No statute gives any State or person a right of action against the President (or any other Executive Branch official) for violating 10 U.S.C. § 12406. Nor, as relevant here, does any statute expressly waive the sovereign immunity of the United States and its officials. 1 In these circumstances, a plaintiff would ordinarily have no right to sue the President or another official to enforce § 12406. The law recognizes two narrow exceptions—review of statutory ultra vires claims and review for claimed violations of the Constitution, see Murphy Co., 65 F.4th at 1128–29—but neither applies here.

As to the first, courts have long found “a right to equitable relief where an agency's action was ultra vires—that is, ‘unauthorized by any law and ․ in violation of the rights of the individual.’ ” Nuclear Reg. Comm'n v. Texas, 605 U.S. 665, 680 (2025) (quoting Am. Sch. of Magnetic Healing v. McAnnulty, 187 U.S. 94, 110 (1902)). The “leading case” on ultra vires review is Leedom v. Kyne, 358 U.S. 184 (1958). Nuclear Reg. Comm'n, 605 U.S. at 681. In Kyne, the Supreme Court held that ultra vires review is available when the challenged executive action is: (1) “an attempted exercise of [a] power that had been specifically withheld”; and (2) “contrary to a specific prohibition” in a statute. 358 U.S at 188–89; see also Nuclear Reg. Comm'n, 605 U.S. at 681.

“Because ultra vires review could become an easy end-run around the limitations of the Hobbs Act and other judicial-review statutes, [the Supreme] Court's subsequent cases have strictly limited nonstatutory ultra vires review[² ] to the ‘painstakingly delineated procedural boundaries of Kyne.’ ” Nuclear Reg. Comm'n, 605 U.S. at 681 (quoting Boire v. Greyhound Corp., 376 U.S. 473, 481 (1964)). So as the Supreme Court stressed in the October 2024 Term, ultra vires review “applies only when an agency has taken action entirely ‘in excess of its delegated powers and contrary to a specific prohibition’ in a statute.” Id. (quoting Ry. Clerks v. Ass'n for Benefit of Non-contract Emps., 380 U.S. 650, 660 (1965)).

By contrast, ultra vires review “does not apply simply because an agency has arguably reached ‘a conclusion which does not comport with the law.’ ” Id. (quoting Boire, 376 U.S. at 481). “Ultra vires review is also unavailable if, as is usually the case, a statutory review scheme provides aggrieved persons ‘with a meaningful and adequate opportunity for judicial review,’ or if a statutory review scheme forecloses all other forms of judicial review.” Id. (quoting Bd. of Govs. v. MCorp Fin., Inc., 502 U.S. 32, 43 (1991)). “Given all that,” the Supreme Court has cautioned, a statutory ultra vires claim “is essentially a Hail Mary pass—and in court as in football, the attempt rarely succeeds.” Id. at 681–82 (quoting Nyunt v. Chairman, Broad. Bd. of Govs., 589 F.3d 445, 449 (D.C. Cir. 2009)).

Plaintiffs nevertheless seek to enjoin the federalization and deployment of California National Guardsmen through a statutory ultra vires claim. In today's majority opinion, that Hail Mary pass scores. The majority holds that Plaintiffs' claim is reviewable because it is a “constitutional ultra vires claim,” as the claim alleges that the President's initial federalization order “exceeded the President's statutory authority” under 10 U.S.C. § 12406 and “encroached on Congress's calling forth power.” Majority Op. at 16, 43 n.11. But in doing so, the majority invents out of whole cloth a so called “constitutional ultra vires claim,” thereby improperly avoiding the onerous ultra vires review standard. 3

The majority argues that “it is unnecessary to decide whether the Kyne standard is met here because Plaintiffs' ultra vires claim ․ is a constitutional claim.” Majority Op. at 43. But the gravamen of Plaintiffs' claim is that Defendants violated 10 U.S.C. § 12406—that is a statutory ultra vires claim subject to the Kyne standard. Indeed, if the majority is correct that “[t]he President's power to ‘call forth’ members of a state's National Guard is statutory, not constitutional,” see Majority Op. at 18, then by definition a claim that such power was violated “concern[s] only issues of statutory interpretation,” and “no constitutional question what[so]ever is raised,” Dalton v. Specter, 511 U.S. 462, 474 n.6 (1994) (citation modified). By improperly characterizing Plaintiffs' claim as a “constitutional ultra vires claim,” the majority permits Plaintiffs to bypass the rigorous scrutiny of ultra vires review by simply “dress[ing] up a typical statutory-authority argument as a[ ] [constitutional] ultra vires claim.” See Nuclear Reg. Comm'n, 605 U.S. at 681. The majority's approach erodes ultra vires review by creating “an easy end-run around the limitations of the Hobbs Act and other judicial-review statutes.” Id. This contravenes the Supreme Court's explicit instruction.

In concluding that the statutory ultra vires claim here can be reviewed as a constitutional claim, the majority misconstrues our decision in Murphy Co. v. Biden, 65 F.4th 1122. There, the plaintiff alleged that “the President violated separation of powers by directing the Secretary [of the Interior] to act in contravention of a duly enacted law,” and we determined that such a claim “could be considered constitutional and therefore reviewable.” Id. at 1130. But the plaintiff's claim was not that the President had exceeded the authority granted by the statute under which he acted. The “duly enacted law” the President allegedly contravened was instead a separate statute prohibiting conduct otherwise permitted by the statute under which he acted. See id. This distinction makes a difference: the cases on which Murphy Co. based its finding of judicial reviewability all involved action permitted by the authorizing statute but prohibited by a second statutory provision. 4 So we determined that the “particularized allegations that [an independent statute] restricts the President's designation powers under the [authorizing statute] satisfies the jurisdictional standard.” Id. at 1131. That conclusion is consistent with the Supreme Court's instruction in Kyne and Nuclear Regulatory Commission that ultra vires review is available only to challenge an executive action that “was an attempted exercise of power that had been specifically withheld” and “violated a ‘specific prohibition’ in” a statute. Nuclear Reg. Comm'n, 605 U.S. at 681 (quoting Kyne, 358 U.S. at 188– 89).

Unlike in Murphy Co. and the cases on which it relies, there is no claim before us that the federalization order independently violated a specific statutory prohibition or a statutory provision specifically withholding authority; the only claim is that federalization was not authorized under § 12406 itself. By expanding Murphy Co. to cover claims which allege that presidential action is simply unauthorized by statute and that such violation alone contravenes separation-of-powers principles, the majority permits plaintiffs without a cause of action to construe any (judicially unreviewable) statutory ultra vires claim as a (reviewable) constitutional claim. But Supreme Court “cases do not support the proposition that every action by the President, or by another executive official, in excess of his statutory authority is ipso facto in violation of the Constitution.” Dalton, 511 U.S. at 472. The majority “thus open[s] the floodgates to frivolous judicial challenges that hinder the President's power to respond to pressing issues” and “empower[s] future objectors to frame any unpopular presidential action as ‘ultra vires.’ ” Murphy Co., 65 F.4th at 1131.

Taking a different tack, the majority contends that Defendants forfeited their reliance on Nuclear Regulatory Commission. Majority Op. at 40–42. Here too I disagree. “When an issue or claim is properly before the court, the court is not limited to the particular legal theories advanced by the parties, but rather retains the independent power to identify and apply the proper construction of governing law.” Kamen v. Kemper Fin. Servs., Inc., 500 U.S. 90, 99 (1991).

In the district court, in the briefing on the emergency stay motion, and in the briefs on appeal, Plaintiffs have alleged that Defendants' actions were ultra vires under 10 U.S.C. § 12406, and Defendants have consistently contested those allegations. With the ultra vires issue properly before us, we retain the power to “identify and apply the proper construction of governing law.” Id. And Supreme Court case law culminating in Nuclear Regulatory Commission governs the availability of ultra vires review. Put another way, the proper standard under which to review Plaintiffs' ultra vires claim is “an issue ‘antecedent to ․ and ultimately dispositive of’ the dispute before [us],” even if “the parties fail[ed] to identify and brief” it. See U.S. Nat'l Bank of Or. v. Indep. Ins. Agents of Am., Inc., 508 U.S. 439, 447 (1993) (quoting Arcadia v. Ohio Power Co., 498 U.S. 73, 77 (1990)). Thus, forfeiture does not prevent us from faithfully applying the relevant legal framework.

And even if Defendants failed to preserve the argument that Plaintiffs lack a valid ultra vires cause of action, that alone would not give us, members of the unelected and non-political branch, a basis to circumvent Congress's authority and grant injunctive relief to a party who lacks a right of action. See Alexander v. Sandoval, 532 U.S. 275, 286–87 (2001). “[F]ederal courts,” the Supreme Court has explained, “do not exercise general oversight of the Executive Branch; they resolve cases and controversies consistent with the authority Congress has given them.” Trump v. CASA, Inc., 606 U.S. 831, 861 (2025). So “[w]hen a court concludes that the Executive Branch has acted unlawfully, the answer is not for the court to exceed its power, too.” Id. Because Plaintiffs lack a right of action, the injunction against Defendants exceeds the limits on federal judicial authority. See id. at 858. Plaintiffs are thus unlikely to succeed on the merits of their ultra vires claim. 5

III

The majority also determines that Plaintiffs are likely to succeed on their claim that Defendants violated the Tenth Amendment. 6 See Majority Op. at 76–77. The parties have scarcely briefed this issue, so the prudent course would have been not to reach Plaintiffs' constitutional claim. Yet the majority again exercises its discretion—wrongly, in my view—to move forward and decide questions that it need not resolve today.

The above provides a more than sufficient reason not to decide the Tenth Amendment claim. But in addition, even based on scant briefing, I am at best skeptical that Plaintiffs are likely to succeed on the merits of their constitutional claim. “[T]he President's actions c[an] ‘be reviewed for constitutionality,’ even though they [are] ‘not reviewable for abuse of discretion’ under the Administrative Procedure Act.” Murphy Co., 65 F.4th at 1129 (quoting Franklin v. Massachusetts, 505 U.S. 788, 801 (1992)). The Supreme Court has long recognized that federal courts have jurisdiction “to issue injunctions to protect rights safeguarded by the Constitution.” Bell v. Hood, 327 U.S. 678, 684 (1946); see also Ex parte Young, 209 U.S. 123, 149, 165 (1908). But, as noted, not all claims are “ipso facto in violation of the Constitution” just because they allege that the President, or another executive official, exceeded their statutory authority. Dalton, 511 U.S. at 472. Instead, courts “distinguish[ ] between claims of constitutional violations and claims that an official has acted in excess of his statutory authority.” Id.

When a plaintiff alleges that a federal officer “violates an express prohibition of the Constitution,” Sierra Club v. Trump, 963 F.3d 874, 889–90 (9th Cir. 2020), vacated and remanded on other grounds sub nom. Biden v. Sierra Club, 142 S. Ct. 46 (2021), or violates “constitutional separation of powers principles,” Sierra Club v. Trump, 929 F.3d 670, 696 (9th Cir. 2019), we have found that such claims are reviewable, see Murphy Co., 65 F.4th at 1130. Attempting to apply those principles here, Plaintiffs contend that Defendants violated an express prohibition in the Constitution—the Tenth Amendment. Their complaint alleges that the federal government “intrud[ed] on the province of the State ․ by taking command of the State's own resources, the California National Guard, which remains under State control unless properly federalized.”

The Tenth Amendment provides that “[t]he powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the States respectively, or to the people.” U.S. Const. amend. X. This provision generally prohibits Congress from commandeering state officials to enforce federal law. See Printz v. United States, 521 U.S. 898, 909, 935 (1997) (holding that Congress may not “command the States' executive power in the absence of a particularized constitutional authorization”). But Article I of the Constitution expressly delegates to Congress the power “[t]o provide for calling forth the Militia to execute the Laws of the Union, suppress Insurrections and repel Invasions” as well as the power “[t]o provide for organizing, arming, and disciplining, the Militia, and for governing such Part of them as may be employed in the Service of the United States.” U.S. Const. art. I, § 8, cls. 15–16. And Congress, in turn, “has delegated some of its power to call forth the militia to the President by statute, including 10 U.S.C. § 12406.” Newsom v. Trump, 141 F.4th 1032, 1045 (9th Cir. 2025) (per curiam).

Because Article I expressly empowers Congress to “call[ ] forth the Militia,” U.S. Const. art. I, § 8, cl. 15, federalization under the Militia Clauses is a power “delegated to the United States,” and is not “reserved to the States” under the Tenth Amendment, U.S. Const. amend. X. Unlike in Printz, therefore, here we have no “absence of a particularized constitutional authorization.” 521 U.S. at 909. So even if the majority is correct that Defendants exceeded their authority under § 12406, that statutory showing alone is not enough to establish a constitutional violation.

The majority turns the Tenth Amendment analysis on its head. It starts with the proposition from Printz that, generally, the federal government “violates the Tenth Amendment by commandeering state personnel to enforce or administer federal law.” Majority Op. at 74 (citing Printz, 521 U.S. at 925). It then reasons that state National Guardsmen “are state personnel, unless properly federalized” and federal personnel if federalized in accordance with the requirements of § 12406. Id. at 74–75. The implication drawn by the majority is that, if the Guardsmen were federal personnel (or “properly federalized”), the federal government could freely press them into service to enforce federal law, because then “no commandeering of state personnel would have occurred.” See id. But, the majority reasons, “because the President lacked authority [under § 12406] to federalize the California National Guard, the federalization orders were invalid, and the deployed Guard members remained state personnel.” Id. at 75. And so, the majority concludes, using them to enforce federal law constitutes unconstitutional commandeering under Printz. See id. at 75–77.

But this analysis is inconsistent with Printz and with the structure of the Tenth Amendment. Under the majority's rule, Congress may, by statute, define what does and does not count as commandeering. As applied here, so long as the President (vested with the executive power of the federal government) follows federalization procedures set out by Congress (vested with the legislative power of the federal government), the Guardsmen are federal personnel, so there is no commandeering and no consequent infringement of the protections of the Tenth Amendment. 7 But Congress has not been given “the power to determine what constitutes a constitutional violation.” City of Boerne v. Flores, 521 U.S. 507, 519 (1997). It is not up to Congress to determine what counts as an intrusion upon state sovereignty or as a violation of the Tenth Amendment.

In fact, in Printz, the Supreme Court considered the validity of a federal statute, not the validity of executive action—the question was whether an act of Congress improperly commandeered state personnel. See 521 U.S. at 905 (“Petitioners ․ contend that congressional action compelling state officers to execute federal laws is unconstitutional.” (emphasis added)). As a result, the Court was primarily concerned with the vertical separation of powers: that between the federal government—as a whole—and each State. See id. at 922 (“We have thus far discussed the effect that federal control of state officers would have upon ․ the division of power between State and Federal Governments.”). Thus, Printz made no suggestion that the statute, which the Court overturned, would be constitutional so long as the President implemented it as written. See id. at 933, 935. And Printz did not derive—and could not have derived—the line between permissible federal regulation and impermissible federal commandeering by reference to the requirements laid out in the challenged statute.

Instead, the Court looked, as we must, to the Constitution. See id. at 905. Looking to the Constitution, Printz did not conclude, in absolute and universal terms, that the federal government may never commandeer state personnel to enforce federal law. The Court only rejected the contention that “the Federal Government may command the States' executive power in the absence of a particularized constitutional authorization.” Id. at 909 (emphasis added). In fact, Printz identified and distinguished several historical examples of the early Congress commandeering state personnel, where the federalization at issue “was in direct implementation ․ of the Extradition Clause of the Constitution itself,” id., or where authority to impose an obligation on state personnel “was perhaps implicit in one of the provisions of the Constitution, and was explicit in another,” id. at 907. Such cases, the Printz Court found, were not counterexamples to the general rule against commandeering because they involved express constitutional authorization. See id. at 907–09. Indeed, express authorization of that kind would have been dispositive in Printz; the Court sought an answer to the general question of whether state personnel can ordinarily be “pressed into federal service” in the “historical understanding and practice” only “[b]ecause there is no constitutional text speaking to this precise question.” See id. at 905.

This case presents a different question. We do not confront whether state personnel may ordinarily be pressed into federal service in the absence of a constitutional provision on point. Instead, we must decide whether the federal government may press state personnel into federal service when the Constitution expressly empowers it to do so. See U.S. Const. art. I, § 8, cl. 15 (“The Congress shall have Power ․ [t]o provide for calling forth the Militia to execute the Laws of the Union ․”). Perhaps that is why, despite Printz's exhaustive review of historical statutes that potentially violate the anti-commandeering principle, neither the majority opinion nor any of the five separate writings ever discuss § 12406 or its predecessors. See generally 521 U.S. 898.

In sum, Printz's general rule does not control here because this case involves federalization in accordance with a specific constitutional power to press state personnel into federal service, which the Constitution reserves to Congress. So we must ask whether the President's federalization order directly violates the Tenth Amendment. The answer is obviously not. The federal government has the power to press state National Guardsmen into federal service, and this power is expressly guaranteed to Congress by the Constitution. See U.S. Const. art. I, § 8, cl. 15. Thus, control of the National Guard is not a power “reserved to the States.” U.S. Const. amend. X. Put differently, no violation of the vertical separation of powers occurred. The federal government therefore could not have violated the Tenth Amendment by federalizing the California National Guard.

The result might be different if Plaintiffs alleged that Congress exceeded the bounds of its Article I authority in enacting § 12406, rather than that the President exceeded the bounds of his delegated statutory authority by issuing the federalization order. But Plaintiffs do not make, and therefore have forfeited, any such claim. The only surviving question is whether Defendants violated the horizontal separation of powers—that is, between the coordinate branches of the federal government. But that question, discussed in the previous section, is one of statutory interpretation and Article II authority, not a question of state sovereignty under the Tenth Amendment.

Because the alleged violation of § 12406 does not implicate any power “reserved to the States,” it falls outside the scope of the Tenth Amendment. Thus, I cannot say that Plaintiffs are likely to succeed on the merits of their constitutional claim. Indeed, I think they likely will not.

IV

For these reasons, I respectfully dissent.

FOOTNOTES

1.   Plaintiffs also claimed that the Guard's deployment violated the Posse Comitatus Act's prohibition of the “use [of] any part of the [military] ․ to execute the laws.” 18 U.S.C. § 1385. The district court held a bench trial on this claim. On September 2, 2025, the court issued an order permanently enjoining Defendants from using the deployed Guard members to execute the laws in violation of the Posse Comitatus Act. Newsom v. Trump, 797 F. Supp. 3d 1092 (N.D. Cal. 2025). Additionally, in the months following the Stay Order, Defendants issued multiple orders extending the Guard's federalization and deployment. Plaintiffs separately challenged the extension orders, and the district court issued an order preliminarily enjoining those orders on December 10, 2025. Newsom v. Trump, 811 F. Supp. 3d 1086, 1116 (N.D. Cal. 2025). Each of these orders—the Posse Comitatus Act order and the preliminary injunction of the extension orders—is the subject of a separate appeal.

2.   Defendants filed a motion to dismiss this appeal on January 23, 2026. Dkt. #152. The dissent contends we should grant Defendants' motion. Dissent at 79-81. We explain why we deny that motion in an order filed concurrently with this opinion.

3.   The Constitution's delegation of federal authority over the state militias to Congress rather than the President reflects the Founders' concerns about the dangers posed by a standing army and unchecked executive power. See Frederick Bernays Wiener, The Militia Clause of the Constitution, 54 Harv. L. Rev. 181, 184–85 (1940) (“The [Constitutional] Debates reflect, in part at least, widespread and exaggerated fears of standing armies․ The Constitution as adopted represented a compromise. A standing army was authorized, but the militia was not abolished. It was to be available for federal service in three specified contingencies. It was to be organized, armed and disciplined by Congress, but, except when in federal service, was to be governed by the States.”); Stephen I. Vladeck, Note, Emergency Power and the Militia Acts, 114 YALE L.J. 149, 157 (2004) (“The central point of contention at [the Constitutional Convention in] Philadelphia over the militia was not the scope of the militia's authority, but who would be responsible for invoking it. Questions of federalism dominated ․ Most everyone at the Convention dreaded a powerful standing army, and nearly as many feared a central, dominant Chief Executive. The consensus thus clearly favored vesting the primary responsibility for responding to threats in the militias of the several states, though the federal government—through Congress, not the President—would exercise ultimate control.”).

4.   For the most part, we addressed Plaintiffs' arguments in our Stay Order. Their merits brief raises only two new arguments. First, Plaintiffs argue that 10 U.S.C. § 12406's statutory history supports their position. Specifically, Plaintiffs assert that § 3 of the Militia Acts of 1792 and 1795 (the “dispersal provision”) is a predecessor of 10 U.S.C. § 12406, and, because the dispersal provision contained express discretion-granting language but § 12406 does not, we can infer that Congress intended to eliminate the president's discretion. The problem with Plaintiffs' argument is that the dispersal provision is not a predecessor of § 12406. Rather, it is a predecessor of a provision of the Insurrection Act, which is now codified at 10 U.S.C. § 254. Compare Militia Act of 1795, ch. 36, § 3, 1 Stat. 424, 424 (“That whenever it may be necessary, in the judgment of the President, to use the military force hereby directed to be called forth, the President shall forthwith, by proclamation, command such insurgents to disperse, and retire peaceably to their respective abode, within a limited time.”), and Militia Act of 1861, ch. 25, § 2, 12 Stat. 281, 282 (“That whenever, in the judgment of the President, it may be necessary to use the military force hereby directed to be employed and called forth by him, the President shall forthwith, by proclamation, command such insurgents to disperse and retire peaceably to their respective abodes, within a limited time.”), with Insurrection Act, 10 U.S.C. § 254 (“Whenever the President considers it necessary to use the militia or the armed forces under this chapter, he shall, by proclamation, immediately order the insurgents to disperse and retire peaceably to their abodes within a limited time.”).Second, Plaintiffs argue that we should not grant any deference to the President's exigency determination because, under Loper Bright, “courts should accord—at most—‘due respect’ to the executive's interpretation of a statute, 603 U.S. at 385, to the extent that it is ‘based upon specialized experience ․ and informed judgment,’ id. at 388.” As explained below, the President's determination that the required exigency exists is not a question of statutory interpretation—it is a mixed question of law and fact. Loper Bright does not require that we review that determination without deference. Consistent with Loper Bright, we decide without deference the underlying statutory interpretation questions, such as the meaning of “unable with the regular forces to execute the laws” and “rebellion.”

5.   In arguing that Sterling authorizes judicial review of the conduct the militia once deployed, not the Governor's exigency determination, Defendants emphasize that the Court stated: “If it be assumed that the Governor was entitled to declare a state of insurrection and to bring military force to the aid of civil authority, the proper use of that power in this instance was to maintain the federal court in the exercise of its jurisdiction, and not to attempt to override it.” 287 U.S. at 404. Again, we must read that statement in context. The Court did not refrain from reviewing the exigency determination, as Defendants suggest. Rather, the Court “assume[d], without deciding,” that Texas law authorized the Governor to declare martial law and use the militia to aid civil forces. Id. at 396. Further, the Court refrained from deciding that issue only because it was a “matter of local law, as to which the courts of the state would in any event have the final word.” Id. at 396. That federalism problem is not present here because § 12406 is a federal statute on which federal courts have the final word.

6.   In a concurrence to the now-vacated stay order issued in Oregon v. Trump, Judge Nelson argued that we erred in relying on Sterling because it addressed the authority of a governor, not a president. 157 F.4th 1013, 1035 (9th Cir. 2025) (R. Nelson, J., concurring), vacated on grant of reh'g en banc, 165 F.4th 1158 (9th Cir. 2025), appeal dismissed, 171 F.4th 1144 (9th Cir. 2026). We disagree. As explained above, the Supreme Court viewed the executive authority and the required exigencies at issue in Martin and Sterling as analogous. See supra at 28-29 (discussing Sterling, 287 U.S. at 399).

7.   Compare Militia Act of 1792, ch. 28, § 2, 1 Stat. 264, 264 (repealed 1795) (“That whenever the laws of the United States shall be opposed, or the execution thereof obstructed, in any state, by combinations too powerful to be suppressed by the ordinary course of judicial proceedings, or by the powers vested in the marshals by this act, the same being notified to the President of the United States, by an associate justice or the district judge, it shall be lawful for the President of the United States to call forth the militia of such state to suppress such combinations, and to cause the laws to be duly executed.”) (emphasis added), with Militia Act of 1795, ch. 36, § 2, 1 Stat. 424, 424 (“That whenever the laws of the United States shall be opposed, or the execution thereof obstructed, in any state, by combinations too powerful to be suppressed by the ordinary course of judicial proceedings, or by the powers vested in the marshals by this act, it shall be lawful for the President of the United States, to call forth the militia of such state, or of any other state or states, as may be necessary to suppress such combinations, and to cause the laws to be duly executed; ․”).

8.   Compare Militia Act of 1792, ch. 28, § 2, 1 Stat. 264, 264 (repealed 1795), quoted in full supra note 7, with 10 U.S.C. § 252 (“Whenever the President considers that unlawful obstructions, combinations, or assemblages, or rebellion against the authority of the United States, make it impracticable to enforce the laws of the United States in any State by the ordinary course of judicial proceedings, he may call into Federal service such of the militia of any State, and use such of the armed forces, as he considers necessary to enforce those laws or to suppress the rebellion.”).

9.   The dissent argues that Defendants could not have “forfeited their reliance on Nuclear Regulatory Commission” because, when an issue is properly before us, “we retain the power to ‘identify and apply the proper construction of governing law,’ ” regardless of whether the parties identified the relevant legal authorities. Dissent at 87 (quoting Kamen v. Kemper Fin. Servs., Inc., 500 U.S. 90, 99 (1991)). But we hold that Defendants forfeited the argument that Plaintiffs lack a valid cause of action, not their reliance on a particular case.

10.   Although we used the term “waived” in Smith, 194 F.3d at 1052, “a party's failure to address a claim or issue in briefing is best described as a ‘forfeiture,’ ” Damiano v. Grants Pass Sch. Dist. No. 7, 140 F.4th 1117, 1152 (9th Cir. 2025).

11.   In Dalton, the Supreme Court described a claim alleging the president exceeded statutory authority but not constitutional authority (that is, a nonconstitutional ultra vires claim) as a “statutory” ultra vires claim. 511 U.S. at 474. In Nuclear Regulatory Commission, the Court described judicial review of an ultra vires claim that was not authorized by a judicial-review statute as “nonstatutory ultra vires review.” 605 U.S. at 680. But, as explained above, the ultra vires claim at issue in Nuclear Regulatory Commission was a statutory ultra vires claim (under Dalton), not a constitutional ultra vires claim. 605 U.S. at 673–74. In other words, Nuclear Regulatory Commission addressed the standard for nonstatutory judicial review of a statutory ultra vires claim. For clarity, we use the term “statutory ultra vires claim” to refer to a nonconstitutional ultra vires claim. We use the term “constitutional ultra vires claim” to refer to a claim that the challenged action exceeded both statutory and constitutional authority. And we use the term “nonstatutory judicial review” to refer to judicial review that is not specifically authorized by statute.

12.   In our Stay Order, we unanimously concluded that § 12406 places reviewable limits on the President's discretion by conditioning its exercise on the existence of a specified exigency. 141 F.4th at 1047 (discussing Dalton); see also supra at 20-21 (affirming that conclusion). See also Illinois, 146 S. Ct. at 433–34.

13.   See also, e.g., 10 U.S.C. § 10102 (“The purpose of each reserve component is to provide trained units and qualified persons available for active duty in the armed forces, in time of war or national emergency, and at such other times as the national security may require, to fill the needs of the armed forces whenever more units and persons are needed than are in the regular components.”); id. § 10103 (“Whenever Congress determines that more units and organizations are needed for the national security than are in the regular components of the ground and air forces, the Army National Guard of the United States and the Air National Guard of the United States, or such parts of them as are needed, together with units of other reserve components necessary for a balanced force, shall be ordered to active duty and retained as long as so needed.”); id. § 10209 (“Laws applying to both Regulars and Reserves shall be administered without discrimination—(1) among Regulars; (2) among Reserves; and (3) between Regulars and Reserves.”).

14.   Defendants correctly note that Plaintiffs did not argue that “regular forces” means “regular military forces” either before the district court or in their merits briefs on appeal. We note that neither party argued that “regular forces” means “regular military forces” in Illinois, see 146 S. Ct. at 437–38 (Alito, J., dissenting), yet the Court adopted that interpretation and denied Defendants' stay motion on that ground.

15.   See Memorandum on Department of War Security for the Protection of Federal Personnel and Property in Illinois, 2025 Daily Comp. Pres. Doc. 978 (Oct. 4, 2025).

16.   Defendants argued before the district court in Illinois that § 12406 provides an express exception to the Posse Comitatus Act. See Defendants' Opposition to Plaintiffs' Motion for Temporary Restraining Order and Preliminary Injunction, Illinois v. Trump, No. 1:25-cv-12174, 2025 WL 3454945 (N.D. Ill. Oct. 8, 2025) (“Section 12406(3) provides express authorization for the conduct complained of here, and that is sufficient to bring it within the PCA's express exception.”).

17.   Past events are not categorically irrelevant to whether the asserted exigency exists. Events that took place in the past may inform the assessment of conditions in the present. But the ultimate question is whether, at the time the order was issued, there was “a rebellion or danger of a rebellion.”

18.   Statutory context confirms that Congress intended “rebellion” to mean something different from “insurrection.” As explained in more detail below, Congress uses both of those terms in the closely related provisions of the Insurrection Act. See Insurrection Act, 10 U.S.C. §§ 251–55. Consequently, we must presume that “rebellion” and “insurrection” mean different things. Henson v. Santander Consumer USA Inc., 582 U.S. 79, 86 (2017) (“[W]hen we're engaged in the business of interpreting statutes we presume differences in language ․ convey differences in meaning.”).

19.   We also note that the Supreme Court construed the 1795 Militia Act as making “the President the sole and exclusive judge [of] whether the exigency has arisen” primarily because of the predicate exigency's extreme nature. Martin, 25 U.S. at 29–30 (“We think that this construction necessarily results from the nature of the power itself, and from the manifest object contemplated by the act of Congress. The power itself is to be exercised upon sudden emergencies, upon great occasions of state, and under circumstances which may be vital to the existence of the Union.”). Thus, construing “rebellion” as broadly as the Defendants suggest would undermine the justification for interpreting the statute as requiring deference to the President's judgment.

20.   See Militia Act of 1792, ch. 28, § 2, 1 Stat. 264, 264 (authorizing President to call forth the militia “whenever the laws of the United States shall be opposed, or the execution thereof obstructed, ․ by combinations too powerful to be suppressed by the ordinary course of judicial proceedings”); Militia Act of 1795, ch. 36, § 2, 1 Stat. 424, 424 (same); Militia Act of 1861, ch. 25, § 1, 12 Stat. 281, 281 (authorizing President to call forth the militia “whenever, by reason of unlawful obstructions, combinations, or assemblages of persons, or rebellion against the authority of the Government of the United States, it shall become impracticable, in the judgment of the President of the United States, to enforce, by the ordinary course of judicial proceedings, the laws of the United States”); Insurrection Act, 10 U.S.C. § 332 (1956), renumbered § 252 (2016) (authorizing President to call forth the militia “[w]henever the President considers that unlawful obstructions, combinations, or assemblages, or rebellion against the authority of the United States, make it impracticable to enforce the laws of the United States in any State by the ordinary course of judicial proceedings”). See also Militia Act of 1792, § 1 (authorizing President to call forth the militia “in case of an insurrection in any state, against the government thereof”); Militia Act of 1795, ch. 36, § 1 (same); Insurrection Act, 10 U.S.C. § 331 (1956), renumbered § 251 (2016) (authorizing President to call forth the militia “[w]henever there is an insurrection in any State against its government”); Insurrection Act, 10 U.S.C. § 333 (1956), renumbered § 253 (2016) (authorizing President to “take such measures as he considers necessary to suppress, in a State, any insurrection, domestic violence, unlawful combination, or conspiracy”).

21.   At the time, what is now widely known as the “Whiskey Rebellion” was called the “Western Insurrection” or the “Whiskey Insurrection.” Whiskey Rebellion, Library of Congress Study Guides, available at: https://guides.loc.gov/this-month-in-business-history/august/whiskey-rebellion [https://perma.cc/8XB5-A5AU]. See also generally H.M. Brackenridge, History of the Western Insurrection in Western Pennsylvania (1794).

22.   See George Washington, Proclamation (Aug. 7, 1794), available at https://founders.archives.gov/documents/Washington/05-16-02-0365#:~:text=Wherefore % 2C % 20and % 20in % 20pursuance % 20of,peaceably % 20to % 20their % 20respective % 20abodes [https://perma.cc/2445-4BYS]. Section 2 of the 1792 Militia Act, as amended, is currently codified as part of the Insurrection Act at 10 U.S.C. § 252.For another example, Defendants cite President Eisenhower's executive order calling forth Arkansas' National Guard to integrate Central High School in Little Rock. That order invoked multiple provisions of the Insurrection Act, 10 U.S.C. §§ 252, 253, 254. See Dwight D. Eisenhower, Executive Order 10730 (Sept. 23, 1957), available at https://www.archives.gov/milestone-documents/executive-order-10730 [https://perma.cc/7JM9-3YC8].

23.   Additionally, the regular army was unavailable because it was deployed in Ohio, fighting for control of the Northwest Territory. Id.

24.   The district court concluded that Defendants could not rely on “danger of rebellion” because the President's June 7 memo stated only that, “[t]o the extent that protests or acts of violence directly inhibit the execution of the laws, they constitute a form of rebellion,” not a danger of rebellion. TRO, 786 F. Supp. 3d at 1255 n.9. Plaintiffs, however, do not make this argument on appeal. Because Plaintiffs have forfeited that argument, we assume without deciding that Defendants may rely on “danger of rebellion.”

25.   On appeal, neither party disputes the district court's factual findings.

26.   The President issued his federalization memo on June 7 and the National Guard troops arrived in Los Angeles on June 8. Id. at 1243–45. The scale and intensity of the protests increased in the following days, but, as of June 12, when the district court issued the TRO, civil law enforcement was continuing to respond, including by making hundreds of arrests, and there was still no evidence of organized violence or an attempt to overthrow the government as a whole. See id. at 1245 (In the afternoon on June 8, “protesters increased to about 3,500 ․ LAPD, LASD, and the California Highway Patrol deployed to the protest area as protesters[ ] blocked Highway 101. Several blocks away, individuals reportedly vandalized Waymo driverless vehicles. Some people moved through downtown, setting off commercial-grade fireworks toward federal officers and throwing objects at passing law enforcement vehicles. Individuals lit fires in dumpsters and trash bins[,] looted at least one store, and vandalized buildings. At least 42 people were arrested in connection with their conduct during the June 8 protests.” (internal quotation marks and record cites omitted)); id. (On June 9, “a thousand demonstrators gathered, and one person drove by firing paintballs at the FPS inspectors. The crowd clashed with LAPD officers, injuring five. Over the course of the protests, state and local law enforcement arrests increased by the day, with over 400 individuals ultimately being arrested in total.” (internal quotation marks and record citations omitted)).

27.   Because we conclude that the President lacked statutory authority to federalize the California National Guard when he issued the order, we do not reach Plaintiffs' claim that the President violated § 12406's procedural requirement that “[o]rders ․ shall be issued through the governors of the States.” 10 U.S.C. § 12406.

28.   Defendants took the same position in Illinois. See Illinois Reply in Support of Motion to Stay, 2025 WL 2977873, at *12 (“As the court of appeals recognized, respondents' Tenth Amendment claim ‘rises and falls with the statutory claim.’ ”).

29.   The dissent contends Defendants concede only that Plaintiffs' Tenth Amendment claim falls with their ultra vires claim, not vice versa. See Dissent at 89 n.6. The district court, however, reasoned that Plaintiffs are likely to succeed on the merits of their Tenth Amendment claim because they are likely to succeed on the merits of their ultra vires claim, and Defendants do not challenge that reasoning.

30.   Plaintiffs also contend, and the district court concluded, that the challenged orders interfere with state sovereignty over local law enforcement in violation of the Tenth Amendment. See Complaint ¶¶ 95-98, 2025 WL 1640287; TRO, 786 F. Supp. 3d at 1260. We do not address that theory because we conclude that Plaintiffs are likely to succeed on their Printz theory.

31.   Because Plaintiffs do not argue that Congress lacked constitutional authority to authorize the President to federalize the California National Guard under the circumstances of this case, like the dissent, we assume without deciding that Congress could have authorized the President to issue the challenged federalization orders.

32.   We do not hold, as the dissent asserts, that “Congress may, by statute, define what does and does not count as commandeering.” Dissent at 92-93. Rather, we hold that Congress may define what counts as a valid federalization order. When the federal government validly federalizes National Guard members, it converts them from state personnel to federal personnel; when the federal government “commandeers,” it compels state personnel to administer federal law. We have no occasion to address whether there are circumstances under which the federal government's use of validly federalized Guard members could violate the anti-commandeering rule or any other Tenth Amendment protection of state sovereignty.

1.   Though Plaintiffs bring a claim under the Administrative Procedure Act, they did not “rely on” this claim “for purposes of their motion for a temporary restraining order.” Newsom v. Trump, 786 F. Supp. 3d 1235, 1246 (N.D. Cal. 2025).

2.   In Nuclear Regulatory Commission, the Court used the terms “nonstatutory review,” “nonstatutory ultra vires review,” and “ultra vires review” interchangeably to refer to the judicial review standard used to determine whether a statutory ultra vires cause of action is available. 605 U.S. at 681. I use the term “ultra vires review” for simplicity.

3.   Contrary to the majority's assertion, I do not “muddle[ ]” the applicable standards. See Majority Op. at 46–47. The ultra vires review standard under Kyne and its progeny should be applied to Plaintiffs' statutory ultra vires claim. Rather, it is the majority that muddles the standards by creating a new “constitutional ultra vires claim” category. See Murphy Co., 65 F.4th at 1128–29 (recognizing two claims as reviewable: “constitutional challenges” and statutory “ultra vires actions”).

4.   See Murphy Co., 65 F.4th at 1130–31 (explaining that the “restriction on reviewing presidential acts for abuse of discretion ‘is inapposite where the claim instead is that the presidential action independently violates another statute’ ” (citation modified) (quoting Mountain States Legal Found. v. Bush, 306 F.3d 1132, 1136 (D.C. Cir. 2002))); id. at 1130 (explaining that, in Chamber of Commerce v. Reich, the D.C. Circuit found a claim justiciable when “plaintiffs challenged President Clinton's executive order, issued pursuant to his Procurement Act authority,” but which “constituted ‘a palpable violation’ of the National Labor Relations Act” (quoting 74 F.3d 1322, 1324 (D.C. Cir. 1996))); id. at 1131 (explaining that, in Massachusetts Lobstermen's Association v. Ross, the D.C. Circuit found justiciable “claims ‘that interpreting the Antiquities Act to permit ocean-based monuments would render the Sanctuaries Act a practical nullity’ ” (quoting 945 F.3d 535, 541, 544 (D.C. Cir. 2019))).The majority also relies on Sierra Club v. Trump, 929 F.3d 670 (9th Cir. 2019). See Majority Op. at 49–50. There, plaintiffs alleged that defendants' reprogramming of funds under section 8005 of the Department of Defense Appropriations Act of 2019 was ultra vires because section 8005's requirements had not been met. Sierra Club, 929 F.3d at 676, 683. But section 8005 itself included both a provision setting out the scope of the executive's authority and a provision specifically prohibiting exercise of that authority under certain circumstances. Id. at 676. More specifically, section 8005 authorized the Secretary of Defense to reprogram funds “if the Secretary determines that the transfer is ‘for higher priority items, based on unforeseen military requirements.’ ” Id. (quoting Pub. L. No. 115-245, § 8005, 132 Stat. 2981, 2999 (2018)). But it also prohibited the Secretary from doing so if “the item for which funds are requested has been denied by the Congress.” Id. at 690 n.17 (quoting Pub. L. No. 115-245, § 8005, 132 Stat. 2981, 2999 (2018)). We determined “it [wa]s clear that Congress denied th[e] request” at issue, and we explained that “this conclusion alone undermine[d] Defendants' likelihood of success on the merits on appeal.” Id. 690–91. So in Sierra Club—like Murphy Co., Reich, and Ross—“nonstatutory review was available” only because the challenged order “was an attempted exercise of power that had been specifically withheld” and because it “violated a ‘specific prohibition’ in” a statute. See Nuclear Reg. Comm'n, 605 U.S. at 666 (emphases added) (quoting Kyne, 358 U.S. at 188–89). No statutory provision specifically withholds or specifically prohibits the President's actions in the case before us.

5.   The majority observes that in Trump v. Illinois, “the plaintiffs advanced an ultra vires claim similar to the one Plaintiffs make here, and in seeking a stay from the Supreme Court, Defendants similarly cited Nuclear Regulatory Commission to argue for greater deference to the President's decision to federalize the National Guard.” Majority Op. at 42. And it notes that the Supreme “Court denied Defendants' motion for a stay without finding it necessary to address either Defendants' theory of deference or whether the plaintiffs had a cause of action.” Id. That much is true but irrelevant. As the majority points out, “[w]hether a plaintiff has a cause of action is not a jurisdictional question that [a court] must address before reaching the merits of a claim.” Id. at 41. And the Court's opinion, which stemmed from its emergency docket, addressed only a single issue that was fatal to Defendants' application for a stay. So I do not read anything into the Supreme Court's silence on the question before us.

6.   The majority observes that “[t]he parties agree that Plaintiffs' Tenth Amendment claim rises and falls with their ultra vires claim,” Majority Op. at 73, and then accuses me of “disput[ing] the parties' shared understanding” on this point, id. at 74. All agree that the Tenth Amendment claim falls with the ultra vires claim—if the federalization order was authorized by § 12406, Plaintiffs' Tenth Amendment claim must fail too. But I see no indication the parties have also agreed that the Tenth Amendment claim rises with the ultra vires claim, at least if that means that a showing that Defendants exceeded their authority under § 12406 necessarily establishes a violation of the Tenth Amendment. Indeed, the majority recognizes that “Defendants argued below that Plaintiffs' Tenth Amendment claim ‘is wholly derivative’ of their § 12406 claim.” Id. at 73. Defendants' repeated statement that the constitutional claim is “derivative” of the statutory claim implies that the dependence is unilateral, not biconditional. Thus, I do not believe it is fair to construe Defendants' position as the majority does.

7.   The majority denies this characterization. Majority Op. at 76 n.32. But its attempt to recast its position fails. The majority reiterates that “[w]hen the federal government validly federalizes National Guard members, it converts them from state personnel to federal personnel,” whereas “when the federal government ‘commandeers,’ it compels state personnel to administer federal law.” Id. So, according to the majority, whether the National Guardsmen are commandeered depends not on the Constitution but on a decision of the political branches to federalize them. The majority adds a qualification that there may be “circumstances under which the federal government's use of validly federalized Guard members could violate the anti-commandeering rule.” Id. But it does not explain how lawfully federal personnel can be commandeered in violation of the Tenth Amendment. Moreover, the door the majority leaves open is inconsistent with at least two of its other conclusions: first, that commandeering applies only to state personnel; and second, that validly federalized Guardsmen are federal, not state, personnel. See id.

SUNG, Circuit Judge:

Opinion by Judge Sung; Concurrence by Judge Miller; Dissent by Judge Bennett