The Military and Elections, Part II: Deploy First, Litigate Later

Editor's note: Listen to a discussion with the authors on the Lawfare Daily podcast here. Find Part I here.

***

As we discussed in our companion piece, a longstanding legal regime protects against use of the military in connection with elections. As an overarching matter, the military’s authority to operate domestically is circumscribed; constitutional and statutory law both establish distinct roles for civilian law enforcement and members of the armed forces. Additional legislation, codified in law and preserved uninterrupted since the mid-19th century, specifically insulates elections from military interference.

This seems like good news for those who fear military involvement in the 2026 election and beyond. But zooming out, the picture blurs. While it’s true that numerous laws restrict use of the military domestically and for elections in particular, other laws affirmatively permit domestic deployments under certain, sometimes ill-defined, circumstances. Still other legal authorities at least arguably permit the use of the military, even if they don’t do so explicitly.

Democrats in Congress seem to fear the existing rules aren’t enough: Members of the Senate Armed Services Committee reportedly attempted last month to amend the National Defense Authorization Act to bar the use of any funds to deploy federal troops to seize ballots, voting rolls, voting machines, or other election materials—and to require Congress to be notified of any federal troop deployment to polling places. Republicans on the committee blocked the measure. In mid-June, a group of Democratic senators also introduced a bill to modify § 592 to require a Joint Resolution before the President could invoke the “armed enemies” exception. It has not come up for a vote.

The power to involve the military in U.S. elections could conceivably come from two places: statute and the Constitution. Statutory sources include the Insurrection Act (an exception to the Posse Comitatus Act) and National Guard authorities under Title 32 and Title 10. Constitutional theories include the “protective power” and the Article II power to respond to attacks on the United States. An expansive reading of any of these might support an argument that the laws restricting use of the military, including those specific to elections, can be overcome by relying on these permissive authorities.

Such expansive readings are dubious. Sounder analysis would likely conclude that the military generally cannot be used in elections. But our purpose here is to outline what authorities might be cited in support of such deployments, not to evaluate whether those claims would prevail in court. After all, that’s how it would play out on the ground: A president could deploy troops for elections on the basis of a weak legal theory, the deployments would be challenged as unlawful, parties would litigate, and—likely months later—a court would adjudicate. But in the meantime, the deployment already happened.

Many commentators worry that a president might invoke the Insurrection Act to sidestep other statutory limits. That’s a reasonable concern. But the Insurrection Act is not the only mechanism for a president to deploy the military at or near a polling site in a manner at least arguably consistent with the law. And it may not be the most likely option for a president eager to use the troops. That’s particularly true because the interaction of these legal regimes—one permitting the president’s use of the military, the other restricting it—remains largely untested. A president with a propensity for pushing the limits of executive authority, including with respect to domestic deployments specifically, might be tempted to capitalize on the uncertainty by sending in the troops first and letting legal challenges follow.

The Insurrection Act and the PCA

The Insurrection Act is best understood for present purposes as an exception to the Posse Comitatus Act, which sharply restricts domestic use of the military. In its modern form, the Posse Comitatus Act provides that

Whoever, except in cases and under circumstances expressly authorized by the Constitution or Act of Congress, willfully uses any part of the Army, the Navy, the Marine Corps, the Air Force, or the Space Force as a posse comitatus or otherwise to execute the laws shall be fined under this title or imprisoned not more than two years, or both.

The statute’s express exception—“in cases and under circumstances expressly authorized by the Constitution or Act of Congress”—is where the Insurrection Act comes in.

The Insurrection Act grants the president authority to deploy the military domestically if one of three conditions is met. The first, 10 U.S.C. § 251, is an “insurrection in any State against its government” where the state legislature or governor requests federal assistance to “suppress the insurrection.” The second, 10 U.S.C. § 252, is the existence of “unlawful obstructions, combinations, or assemblages or rebellion against authority of the United States” that make it “impracticable to enforce the law” of the United States…by judicial proceedings.” The third, 10 U.S.C. § 253, is “any insurrection, domestic violence, unlawful combination, or conspiracy” that either (1) “hinders the execution” of state or federal law such that “any part or class of its people is deprived of a right, privilege, immunity, or protection named in the Constitution and secured by law” and the local state authorities are “unable, fail, or refuse” to act, or (2) “opposes or obstructs” federal law  or “impedes the course of justice under those laws.”

Once one of these criteria is met, the Insurrection Act requires the president, before deploying troops, to issue a public proclamation ordering the insurgents to “disperse and retire peaceably to their abodes within a limited time.”

The precise extent of these powers remains uncertain, but they can plainly be read expansively by a president inclined to do so. Section 252—which covers situations making it “impracticable” to enforce federal law—explicitly leaves it to the president to determine whether the facts satisfy the requirements (“whenever the President considers…”). The other two provisions lack similar statutory text, but a Justice Department 1964 Office of Legal Counsel (OLC) Memorandum concluded that the president is empowered to act whenever a state asks for assistance, a court order is involved, or there is a “serious and general breakdown of the authority of state and local government in the area affected.”

Even the OLC, however, emphasized that the Insurrection Act was not as broad as it may seem. The memo noted that the “Constitution and tradition” limit the president’s authority, especially under the provisions now at § 252 and § 253 (at the time of the memo, these provisions were codified at § 333 and § 334 of the Insurrection Act). Those provisions, the OLC opined, were

designed to deal with situations where state and local law enforcement have completely broken down, either because local officials are themselves opposing and obstructing federal law or because they are unable or unwilling to control private groups that are in command of the situation. . . . [T]he use of military force to execute the laws has traditionally been regarded with disfavor—as a course of action that can be lawfully and properly pursued only as a last resort.

That interpretation reflects the Insurrection Act’s long history. It and its predecessors have mostly been invoked to do exactly what they appear designed for: to quell violent disorder, from the 1794 Whiskey Rebellion to the 1992 Los Angeles Riots following the police beating of Rodney King. President Dwight Eisenhower invoked the Act to enforce a desegregation order in Little Rock, Arkansas, after a mob tried to block nine Black students from entering the city’s all-white Central High School. Of course, not every invocation of the Insurrection Act has been uncontroversial. When President Grover Cleveland used the statute to dispatch federal troops to end the Pullman railroad car labor strike, for example, the governor of Illinois called the move unconstitutional. The general in charge said the proclamation meant “whoever disobeys is a public enemy, and as such is to be destroyed.” Approximately 30 people were killed, and many more wounded.

The Insurrection Act and Limiting Statutes

However restricted the Insurrection Act may be by the Constitution, tradition, or statutory text, it undeniably allows the president to bypass the Posse Comitatus Act and use the military domestically, at least under some circumstances. But, as we discussed in our companion piece, three other statutes specifically insulate elections from military involvement in elections: 18 U.S.C. §§ 592-593 and 52 U.S.C. § 10102. Does the Insurrection Act create an exception to those prohibitions as well?

The executive branch wrestled with this very question in 1968—a year marred by the assassinations of Martin Luther King Jr. and Robert F. Kennedy and by anti-Vietnam War protests that turned into riots at the Democratic National Convention. OLC wrote a hasty page-and-a-half opinion concluding that § 592 did not restrict the president’s ability to use the military under the Insurrection Act when its conditions were met—for instance, when a governor requests federal military support “to put down an Election Day civil disorder that is beyond the ability of local authorities to control.” The memo promised a more thorough analysis to come—though if such a document exists, it has never been made public.

The existing opinion is itself largely free of legal analysis. What little reasoning it offers rests on the claim that § 592’s legislative history “shows” the statute was “primarily intended to prohibit subordinate military or civil officers from acting at their own discretion to ‘keep the peace’ at the polls.”

That claim isn’t just unsupported—it’s also inaccurate. A faithful review of the legislative history, including conference reports and floor speeches, shows that the law was a direct response to military deployments to election sites in Kentucky, Maryland, Missouri, and Delaware—deployments ordered and led by high-ranking officers that provoked outrage among Democrats. OLC also never mentions that the senator who introduced the bill, L. W. Powell of Kentucky, believed that § 592 applied to the president himself: In an 1864 speech to the Senate, he argued that the president had no constitutional power to intervene in elections, saying “[h]e who attempts to interfere with this most inestimable right, whether he be President, major-general, or citizen, is an enemy to the public and deserves the harshest punishment.”

Nor does the memo address an obvious question: What about the congressional intent reflected every year since 1909, when Congress dropped the “keep the peace” exception from the statute? The modern language of § 592—unchanged since OLC’s 1968 analysis—allows troops only if “such force be necessary to repel armed enemies of the United States.”

The 1968 OLC memo also ignores an earlier episode. In 1879, President Hayes was presented with legislation that would have dropped the “keep the peace at the polls” language—the same language ultimately removed in 1909. Hayes vetoed the bill, explaining that existing statutes, including the Posse Comitatus Act, already protected against military interference in elections. If the “keep the peace” language was eliminated, Hayes warned,

there will be no power vested in any officer of the Government to protect from violence the officers of the United States engaged in the discharge of their duties. Their rights and duties under the law will remain, but the National Government will be powerless to enforce its own statutes.

A few weeks later, Hayes vetoed a second bill that would have allowed troops at polling sites only to repel armed enemies of the United States or to enforce Section 4 of the Fourteenth Amendment. Hayes said the restrictions were too expansive and the two exceptions too narrow. His chief objection was that the legislation would “abrogate” the Insurrection Act, which he called an “ancient and fundamental law” dating back to the founding, “on certain days and at certain places.”

In short, Hayes vetoed both bills because he feared the prohibition on military presence at the polls—the same prohibition in place today—would have become so absolute that not even the Insurrection Act could override it. His interpretation was precisely the opposite of the 1968 OLC memo that neglected to mention him. 

There’s additional historical evidence against reading the Insurrection Act as an override of the prohibitions now codified at §§ 592, 593, and 10102: Congress’s standoff with Hayes came right after Congress’s passage of the Posse Comitatus Act. It’s unlikely Congress would have kept fighting to amend 1865-era text if it believed the new law sufficiently protected elections. Yet, it tried—twice. And subsequent Congresses kept the election-specific protections in place alongside the Posse Comitatus Act, even revising them to become more restrictive in 1909. Congress also made the election-specific statutes carry more severe penalties: violating the Posse Comitatus Act draws up to two years in prison, while violations of modern-day §§ 592-593 draw up to five years and disqualification from office.

Finally, the Posse Comitatus Act’s exception applies “in cases and under circumstances expressly authorized by the Constitution or Act of Congress”—language that plainly covers the Insurrection Act. The statutes protecting elections from military interference, by contrast, contain no such carveout for acts of Congress. 

Whether the president could indeed invoke the Insurrection Act to override §§ 592-593 and § 10102 has never been tested in court. But the theoretical question may not matter much in practice: §§ 592-593 are criminal statutes that the Justice Department could simply decline to prosecute, or the president could pardon, and § 10102 is a civil statute with no enforcement mechanism. And troops would likely be on the ground long before any court could rule on an injunction. 

The Insurrection Act, though, is not the only avenue. A president determined to deploy the military for elections has other options.

National Guard Authorities

Although the National Guard is a reserve component of the United States military, it is structurally distinct and thus governed by different statutes. National Guard troops are simultaneously members of their state militia and reserve troops for the federal military. By default, the National Guard is under the command and control of their state’s governor. The president can federalize the National Guard under provisions within Title 10, making Army National Guard and Air Force National Guard forces part of the Army and Air Force, respectively. He can also utilize the National Guard under Title 32 “hybrid” authorities, in which National Guard troops remain under gubernatorial command but carry out a federally funded mission. In either situation, deploying National Guard members requires both a mobilization authority (calling them to service from their civilian status) and a mission authority (authorizing what they’ll do).

Title 32

Under 32 U.S.C. § 502(f), National Guard troops remain under the command and control of their state governor but are paid by the federal government and are deployed to “perform training or other duty” as directed by the federal government. This is known as “Title 32 status.” The troops remain in state status while conducting a federal mission. The law allows activation for  “support of operations or missions . . . at the request of the President or Secretary of Defense” or “[s]upport of training operations and training missions” in limited circumstances.

Title 32 was originally created to provide federal funding to train National Guard troops so that they could perform at similar levels to active duty troops. Traditionally, § 502(f)’s purpose was likewise limited to training; the provision is housed within a section of the U.S. Code setting requirements for “drills and field exercises” to ensure that the National Guard, otherwise serving their state militias, were properly prepared for their dual purpose as federal reserves. Over time, however, 502(f)’s scope grew: Congress added drug interdiction missions in 1989, and certain “homeland defense activities.” Both of these statutes cite 502(f) as the mobilization authority for their missions, but the new statutes, 32 U.S.C. § 112 and 32 U.S.C. § 904, provided the mission authority for such operations. In 2006, Congress amended § 502(f) to specify that “training or other duty” included National Guard support of federal operations or missions “at the request of the President or the Secretary of Defense.”

The current Trump administration leaned on § 502(f) heavily—Lawfare has documented at least 20 instances—and its reach is now being tested in court. In litigation challenging National Guard deployments to Washington, D.C., the Justice Department argues that the 2006 amendment lets the president or the defense secretary not just mobilize the National Guard under 502(f), but order it to perform any mission the president or secretary may choose. Critics argue that this interpretation contravenes common sense and legislative history.

What does this mean for the president’s § 502(f) authorities when it comes to protecting elections from military interference?

First, the Posse Comitatus Act doesn’t apply: Title 32 troops remain under gubernatorial command, and the Act only restricts “any part of the Army, the Navy, the Marine Corps, the Air Force, or the Space Force”—services that don’t include state-controlled Guard units, unless and until they’re federalized under Title 10.

Second, the election-protective statutes at §§ 592-593 have a similar problem, since both target the “Armed Forces of the United States.” If National Guard members serving in Title 32 status are considered members of their state militias simply supporting the federal government—and not part of the “Armed Forces of the United States” while in that status—then §§ 592-593 would not apply by their own terms.

But § 592 reaches people in broader “civil, military, or naval service of the United States” with authority over “troops or armed men” at election sites—which means that whoever ordered the Guard to the polls could still be liable, even if the troops themselves escape liability. Notably, the Defense Department Directive on Defense Support of Civil Authorities forbids “DoD personnel and National Guard in Title 32 status” to “conduct operations at polling places and strictly refrain from activities similar to those prohibited by” 18 U.S.C. §§ 592-594.

Title 10

The president can also fully federalize the Guard. Most Title 10 mobilization authorities—unlike Title 32—do not require the governor’s consent (for instance, §§ 12302, 12304, 12406).

Under Title 10 status, Guard members are part of the Army or Air Force, so the Posse Comitatus Act and §§ 592-593 apply. By extension, the analysis in our companion piece and in our discussion of the Insurrection Act above applies, too.

Title 10 § 12406 provides that the president may call National Guard members into federal service if

1. (1) the United States, or any of the Commonwealths or possessions, is invaded or is in danger of invasion by a foreign nation;
2. (2) there is a rebellion or danger of a rebellion against the authority of the Government of the United States; or
3. (3) the President is unable with the regular forces to execute the laws of the United States.

The Supreme Court has weighed in on only the third subsection. Trump v. Illinois, decided six months ago, arose after the current administration relied on § 12406 to deploy National Guard troops to cities across the United States. The president claimed that the Illinois deployment was “to protect federal personnel and property from violent resistance against the enforcement of federal immigration laws.”

The Court held in an interim order that § 12406 didn’t authorize this use, because the Posse Comitatus Act bars the military from executing the laws—which is what the Guard would be doing here. In other words, for § 12406 to work as backup authority, the federal military must have the authority to carry out the mission itself. Whether the president can use § 12406(3) for elections, then, turns on the same question as whether he can use the federal military for elections.

The second subsection—“rebellion or danger of rebellion against the authority of the Government of the United States”—is murkier. Lower courts have touched on it in the Northern District of California (Newsom v. Trump; later stayed on appeal), the District of Oregon (State of Oregon v. Trump; also stayed on appeal), and the Seventh Circuit (Illinois v. Trump; which the Supreme Court later avoided).

The most important takeaway for domestic military action is that both the Ninth and Seventh Circuits held that the courts must give at least a tentative level of deference to the president’s determination that the requirements of § 12406 have been met. This assumption of deference has implications for other provisions of § 12406 and other statutes authorizing domestic deployments. The Ninth Circuit said courts may review the president’s determination to confirm it reflects “a colorable assessment of the facts and law within a ‘range of honest judgment.’” The Seventh Circuit agreed deference was due but declined to say how much.

Even amid this uncertainty, a troubling pattern has emerged: deployment first, legal challenge and litigation after. Given the slow pace of litigation, and considering that deference to the government may make injunctions harder to get, a president’s legal theory may not need to be sound to keep troops on American streets for months while courts sort it out.

Constitutional Theories

Beyond statute, a president intent on election-related deployments could turn to the Constitution itself. These legal theories may prove the most difficult to counter. Read expansively, they could override the statutory restrictions discussed in our companion piece and erode the limitations in the already-permissive statutes described above. Some of these interpretations seem unlikely to survive judicial scrutiny—but none have been tested.

The Protective Power

The first theory, known as the “protective power,” asserts that the president possesses inherent Article II authority to deploy troops to protect federal property, persons, and functions. The theory surfaced in the mid-19th century but did not take hold until 1877, around the same time as the Posse Comitatus Act.

OLC laid out the modern interpretation of the protective power in three memos: In 1967, authorizing military protection of "property and functions of the Federal government” against Vietnam War protestors; in 1970, extending it to foreign embassies in Washington, D.C; and in 1971, extending it to “assuring the availability of federal employees to carry out their assigned duties,” including through traffic control. All three memos cite constitutional principles but no case law, relying exclusively on dicta for judicial support.

The theory stayed confined to internal OLC opinions until last year, when the Trump administration tested it in court for the first time. In Newsom v. Trump, the Justice Department defended the president’s Los Angeles deployment of National Guard troops—when nationalized under 10 U.S.C. § 12406 in response to anti-Immigration and Customs Enforcement protests—as consistent with an “inherent protective power to provide for the safety of federal property and personnel.” In Oregon v. Trump, the government argued that the “protective function does not constitute law execution” under the Posse Comitatus Act and the president’s authority was an “‘inherent’ protective and emergency power derived from the Take Care Clause.” And in Illinois v. Trump, the government told the Supreme Court that protecting “those who enforce the laws” doesn’t constitute “execut[ing] the laws” under the Posse Comitatus Act, and, in any event, these protective functions were “expressly authorized by the Constitution.”

The results have been mixed and inconclusive. In Newsom v. Trump, District Court Judge Charles Breyer ruled that the protective power is “not grounded in the history of the Posse Comitatus Act, Supreme Court jurisprudence on executive authority, or common sense,” but he left open whether he was rejecting the protective power outright or merely as a constitutional exception to the Posse Comitatus Act; that piece of the appeal is now in abeyance

In Oregon v. Trump, District Court Judge Karin Immergut addressed the government’s protective power argument only in a footnote: “the President’s inherent authority under the Take Care Clause does not support the conclusion that he has inherent authority to federalize the state National Guard to accomplish this purpose.” Her opinion was also appealed to the Ninth Circuit, later remanded for trial, and appealed again, but none of these later opinions analyzed the protective power argument.

In Illinois v. Trump, neither the district court nor the court of appeals directly treated the protective power argument, instead focusing on arguments under § 12406 and the Tenth Amendment. The Supreme Court’s per curiam opinion in that case addresses the protective power to some degree, albeit without deciding to what extent it could serve as a proper justification for deploying troops. Instead, the Supreme Court held that the protective power is incompatible with deployments under § 12406 because the former assumes that the military’s function is not “executing the laws,” whereas the latter relies on the premise that the president needs the National Guard because he is “unable with the regular forces to execute the laws of the United States.”

In short, courts haven’t resolved whether the president can deploy the military domestically under the protective power theory—only that he can’t use it to trigger § 12406.  Whether it could support Title 32 deployments, or deployment of regular forces, remains untested. If read as broadly as the Trump administration has argued, it could justify sending troops to “protect” federal property near polling locations, or to “protect” federal employees like postal workers handling mail-in ballots.

Article II Power

A second theory rests on the widely accepted view that the president, as commander in chief, has some inherent authority to use the military domestically to respond to an attack on the United States.

If a president claimed that such an attack was underway at a polling place or an elections hub, he could argue this power overrides the Posse Comitatus Act, §§ 592-593, and any other statutory limitations—framing the domestic deployment as serving a military purpose rather than a law enforcement one. This idea was debated extensively in the aftermath of 9/11, albeit usually alongside the Authorization for the Use of Military Force, which the president could use as a statutory hook to justify using the military.

Under this reasoning, a president invoking commander-in-chief powers could use troops to block or close polling sites, treating any electoral impact as merely incidental—or to seize election infrastructure he claimed was the target of a cyberattack or foreign interference.

A president relying on inherent constitutional authority is unlikely to be meaningfully constrained by the Posse Comitatus Act or the election-protective statutes. All are criminal statutes that his own Justice Department would presumably decline to prosecute, or that he could pardon away. The Supreme Court’s opinion in Trump v. United States also suggests that the threat of prosecution by a future administration would have little deterrent effect. 

*      *     *

The Constitution and Congress are both clear on the sanctity of elections in American democracy. The Constitution gives the president no role in regulating when, where, or how citizens vote. Congress has consistently passed and preserved statutes shielding elections from military involvement even in wartime—unless needed to “repel armed enemies of the United States.”

But the president also holds other broad military powers, and how those interact with the election-protective legal regime remains largely untested. The question is whether centuries of statutory protection are enough, given that the president’s authority is at its broadest precisely when responding to threats to the country.  The system depends on the executive wielding that power in good faith. If he doesn’t, it risks breaking down.

What happens if a president sends troops to polling sites to provide security? What happens if he orders them to seize ballots over alleged voting fraud he claims are causing certain people to be “deprived of a right”? And what happens if this or any other intervention takes place before litigation can be brought, much less resolved, to challenge its lawfulness? Assume that, after years of relative calm, the Constitution, Congress, and the president clash over the place of the military in U.S. elections.