The constitutional basis for Trump’s use of the military in LA has a long history but is ultimately unmoored from constitutional text.
On June 10, U.S. Northern Command (USNORTHCOM), the combatant command responsible for homeland defense and supporting civil authorities within the continental United States, announced that 700 Marines had been deployed to the greater Los Angeles area. Then, that evening and into June 11, those Marines began accompanying Immigration and Customs Enforcement (ICE) agents on deportation raids. A spokesperson for USNORTHCOM stated that these Marines would continue to accompany ICE agents during immigration raids in order to provide security.
California’s request for a temporary restraining order to enjoin President Donald Trump’s use of the National Guard contains further details of alleged Defense Department plans. In relevant part, the briefing notes that planned military activities include “holding a perimeter in communities around areas where immigration enforcement activities would take place, and securing routes over public streets where immigration enforcement officers would travel.”
The Department of Justice’s June 11 response to California’s suit addressed, in part, the theory of implied presidential authority to protect federal functions, persons, and property that underpins the military deployment to Los Angeles. Oddly, the briefing suggests that 10 U.S.C. § 12406 itself authorizes the National Guard to undertake law enforcement activities as an exception to the Posse Comitatus Act (PCA), the federal statute generally prohibiting the military from conducting domestic law enforcement activities. At a minimum, this would depart from the Defense Department’s existing understanding of exceptions to the PCA, which does not cite § 12406. It also does not comport with the statute’s context either at enactment (as noted in a prior Lawfare article, the statute was enacted in 1903 as part of a reorganization of the National Guard, not as a substantive authorization for federal duty) or its placement in Title 10, a positive law code, as part of a chapter concerning National Guard duty status and organization.
These assertions ultimately, however, yield to the theory of inherent constitutional power, which will be an important part of the June 12 oral argument in California’s suit. Accordingly, this piece sets out to do four things. First, I describe the limited but, at the margins, historically fuzzy practice of presidential invocations of the protective power. Second, I describe the executive branch’s shifting constitutional arguments supporting this implied power. Third, I lay out why I find these executive branch legal arguments to be unpersuasive. And, finally, I note unresolved tensions between the protective power and case law concerning the PCA.
This piece is drawn largely from a law review article published with the Harvard National Security Journal in 2023. That article, in turn, was based primarily on legal memoranda from the War Department, Defense Department, and Department of Justice from the mid-1800s to today. I obtained those memoranda from records at the National Archives and under the Freedom of Information Act (FOIA). You can find the memos obtained under FOIA at the Harvard National Security Journal’s website.
Historic Expressions of the Protective Power
The first fact to know about the protective power is that I could find no explicit assertions of it until after the Civil War. Indeed, until the 1830s every documented use of the military for a domestic law enforcement purpose was supported by express statutory authority (more on this below). It was only in 1850 that President Millard Fillmore articulated a theory of inherent presidential power, rooted in the president’s constitutional duty to “take care that the laws be faithfully executed,” that would ultimately form the legal basis for the protective power. As I noted in Lawfare over the weekend, Fillmore asserted this authority to suppress an uprising by Bostonians against the detention and extradition of an accused fugitive enslaved person, Shadrach (accounts differ on his last name), back to the South.
In fact, it was only starting with President Rutherford B. Hayes in 1877 that a relatively consistent pattern of executive branch practice began to emerge regarding the protective power. Many of these assertions of the protective power related to public protests. For example, the executive would assert his power to protect federal buildings and facilities during railroad strikes, to ensure that mail would be delivered, and—at least once—to protect immigration agents during a riot within a detention facility on Angel Island.
A trio of Department of Justice Office of Legal Counsel (OLC) memos lay out the modern interpretation of the protective power.
The first of these, from 1967, asserted that the protective power extended to using the military “to protect property and functions of the Federal Government from interference by anti-war demonstrators” who planned to protest at the Pentagon. Importantly, this memo ended by noting that “in the event that Federal troops are to be used not merely in a defensive operation to protect Government property and functions at the Pentagon or elsewhere in Washington, but for affirmative purposes of suppressing rioting or other general law enforcement activity in the city[,]” the president must rely on provisions of the Insurrection Act.
A 1970 memo concluded that the president could also rely on the protective power to use the military to protect foreign embassies in Washington, D.C. (interestingly, after an earlier OLC opinion came to the opposite conclusion). And finally, a 1971 memo, again concerning Vietnam War protests, asserted that the protective power included using the military to “protect the functioning of the government by assuring the availability of federal employees to carry out their assigned duties” and “prevent[ing] traffic obstructions designed to prevent the access of employees to their agencies.”
It is this final OLC memo that provides the most direct analogue to the protective activities Marines undertook Tuesday night and into Wednesday. It is also, as I will discuss briefly below and as Elizabeth Goitein has argued more fully, the element of the protective power most liable to abuse and the most likely to be in tension with prevailing judicial interpretations of the PCA.
Doctrinal Basis for the Protective Power
This pattern of practice belies significant shifts in the executive branch’s justification for the protective power. Twentieth-century War Department memoranda, for example, characterize the protective power as a manifestation of the “government’s right of self-defense,” without citing any provision of the Constitution. Indeed, a 1943 War Department memorandum rooted the protective power in the “‘law of necessity’ inherent in every sovereign to preserve its existence. Thus, the President may call on the Army to protect government buildings without issuing a proclamation, where he deems it expedient.”
This sovereignty-based justification of the protective power was most pointedly framed in the 1967 OLC memo about Vietnam War protests at the Pentagon. OLC characterized the protective power as “a particular expression of the martial law principle that when unlawful force threatens the order of the State and the appropriate civil authorities are unable to preserve order, the sovereign may use military force to whatever degree is necessary for that purpose.” OLC traces this justification for the protective power to a line of prior War Department opinions and “powerful dicta” from In re Neagle (more on this in a moment). Nowhere does the opinion cite a constitutional provision on which this “martial law principle” rests.
Over the next two years, OLC refashioned its legal argument to sound more explicitly in the Take Care Clause, which provides that the president “shall take care that the law be faithfully executed.” This shift is most plainly visible in the 1971 OLC opinion, which cites In re Neagle and In re Debs as standing for Supreme Court recognition of the protective power based in the Take Care Clause. This is, at best, an overstatement of the Supreme Court’s holdings. In re Neagle, for example, affirmed a U.S. marshal’s authority to use force without explicit statutory authority to protect Justice Stephen Field during an attack by two armed assailants on a train in California. Only in dicta—that is, not in direct support of its decision—did the Supreme Court state, in incredibly broad terms and again without reference to particular constitutional text, that:
the government of the United States may, by means of physical force, exercised through its official agents, execute on every foot of American soil the powers and functions that belong to it. This necessarily involves the power to command obedience to its laws, and hence the power to keep the peace to that extent.
Similarly, In re Debs at best provides support in dicta for the government’s theory of inherent authority. In this case, the Supreme Court upheld a circuit court’s authority to issue an injunction against those involved in the 1894 Pullman labor strikes. More particularly, it affirmed the circuit court’s decision that Eugene Debs, as one of the principal organizers of the strike, was in contempt of the court’s injunction. Once again in dicta cited by OLC, the Supreme Court asserted that:
The strong arm of the national government may be put forth to brush away all obstructions to the freedom of interstate commerce or the transportation of the mails. If the emergency arises, the army of the Nation, and all its militia, are at the service of the Nation to compel obedience to its laws.
Aside from these two judicial decisions, OLC inexplicably cites no other case law regarding implied presidential powers. At least two Supreme Court opinions, rendered much closer in time to 1971 than either Nealge or Debs, are of foundational importance. First, in Curtiss-Wright, Justice George Sutherland’s majority opinion affirmed “[t]he broad statement that the federal government can exercise no powers except those specifically enumerated in the Constitution, and such implied powers as are necessary and proper to carry into effect the enumerated powers.” Likewise, OLC did not grapple with the Supreme Court’s holding in Youngstown Steel, in which—to summarize—the Court much more parsimoniously calibrated the president’s inherent powers against congressional action and the clarity of constitutional text.
Moving beyond judicial precedent, the 1971 OLC memo also relies on Edward Corwin’s 1950s study of presidential power as evidence of support for this expression of inherent presidential power during the early Republic. The last basis for the protective power asserted in the 1971 memo was War Department practice, summarized above.
Defects in the Executive Branch’s Legal Justifications
There are many reasons why the executive branch’s constitutional justification for the protective power is weak.
First, as noted above, no president so much as intimated an understanding of Article II authority similar to that which underpins the protective power until the mid-1800s. Corwin’s treatise, for example, misses at least three instances in which presidents George Washington, John Adams, and Thomas Jefferson relied on explicit statutory authority for domestic military deployments. Corwin also mistakenly asserts that Jefferson’s 1808 use of the military to enforce the Embargo Act must have been taken without statutory authority (there were at least four plausible statutory bases in effect at the time).
Even when President Fillmore first asserted this theory of inherent authority in 1850, the Senate Committee on the Judiciary responded that “[w]hen occasions arise (which must necessarily be rare) where the civil power proves inadequate to maintain the law, the President is authorized [by antecedents to the Insurrection Act] to call forth and employ, in the manner prescribed by these acts, the militia of the States … to suppress insurrections, and to enforce the due execution of the laws.”
Further, as noted above, OLC (and the Justice Department’s briefing in the California litigation) fails to reckon with Supreme Court case law in tension with that articulated by In re Neagle and In re Debs. There is a vast body of scholarship on this matter, but excellent pieces documenting this shift away from broad, inchoate assertions of the prerogatives inherent in federal sovereignty, in the national security context, include a 2015 article by Ganesh Sitaraman and Ingrid Wuerth, another 2015 contribution by Jack Goldsmith, as well as pieces on the Take Care Clause and the Executive Vesting clause by Jack Goldsmith and John Manning and Julian Mortenson. I don’t want to overstate Supreme Court precedent here. Sarah Cleveland, in a masterful 2002 piece, showed the wide range of circumstances in which constitutional arguments based in sovereignty continue. And David Driesen and Bill Banks have similarly argued that this retreat from theories of inherent presidential authority is overstated. But even if current jurisprudence has not entirely abandoned sovereignty-based theories of presidential power, these theories have at least been significantly tempered in ways that the executive branch has simply never grappled with.
Finally, the executive branch has failed to account for the radically different statutory environment that currently prevails. At the time that presidents first asserted the protective power, there was little in the way of federal law enforcement capacity. Today, there are a multitude of federal law enforcement agencies—from the National Park Police to the Bureau of Indian Affairs Police, to the Bureau of Reclamation Security Response Force, to police forces for the Bureau of Engraving and Printing and the U.S. Mint, to name a small selection. Moreover, Congress has specifically empowered the secretary of homeland security to “protect the buildings, grounds, and property that are owned, occupied, or secured by the Federal Government” and to detail law enforcement officers to other agencies to more effectively protect their properties.
There are, moreover, a laundry list of statutory authorizations for the president or secretary of defense to aid law enforcement activities far beyond the broad delegations in the Insurrection Act. These extend from matters small (e.g., protection of Yosemite, Sequoia, and General Grant National Park) to large (e.g., assisting with response to weapons of mass destruction attacks). The executive branch appears to have ignored these results of decades of congressional action.
Folks can reasonably differ as to whether these judicial and statutory developments suggest that there is no basis for the protective power, that the protective power should be more minimally construed, or that the protective power in its current form is justified. But they cannot simply be ignored.
Unresolved Tensions Between the Posse Comitatus Act and the Protective Power
Similarly, these executive branch opinions do not grapple with federal case law regarding the PCA. As OLC recognized in an opinion in the final days of the first Trump administration—itself reflecting many decades of a separate line of OLC opinions—there are at least three tests for what conduct violates the PCA’s prohibitions. Some courts believe the PCA prohibits conduct that “actually regulates, forbids, or compels some conduct on the part of those claiming relief.” Others believe it criminalizes activity that involves “the direct active participation of federal military troops in law enforcement activities.”
Importantly for present purposes because it prevails in California, the U.S. Court of Appeals for the Ninth Circuit in United States v. Dreyer has held that conduct that “pervades the activities of civilian officials” violates the PCA. In Dreyer, the Naval Criminal Investigative Service (NCIS) used a software program that scraped the internet for child pornography shared anywhere in Washington state. Although the aim was to catch any military members in possession of child pornography, the software caught many civilians as well. Dreyer was one such civilian. NCIS shared the information it collected with state law enforcement, who then based their warrant for and subsequent prosecution of Dreyer on that information. It was this enmeshing of the military’s activities and the civilian law enforcement process that the Dreyer Court found to violate the PCA.
In part because the facts continue to develop so quickly, it’s unclear how close to the standard established in Dreyer the Pentagon is coming in using the military to accompany ICE agents on immigration enforcement raids. Cordoning off a street so that ICE agents can conduct an ICE raid, for example, seems to fall short of the military involvement in law enforcement at issue in Dreyer. But the more that the military’s mission expands in LA—including temporary detention of civilians, to take just one example—the closer the facts get to impermissible law enforcement activities. As Joseph Nunn has noted in detail, the PCA is unhelpfully vague in what it prohibits. And the executive branch has implied a plethora of exceptions in the act’s text. This makes policing the line between defensive activities—which the executive believes to be within the scope of the protective power—and “pervasive” involvement in law enforcement incredibly difficult.
* * *
The protective power is a theory of inherent presidential power encased in 19th-century legal amber. It has a long history, stretching at least to the 1800s. But this long history masks a shifting legal justification that fails to account for numerous developments in constitutional and statutory law since the turn of the 20th century.
– Chris Mirasola is an Assistant Professor of Law at the University of Houston Law Center. Previously, he was a Climenko Fellow and lecturer on law at Harvard Law School and an attorney-advisor at the Department of Defense Office of General Counsel. Published courtesy of Lawfare,
