Armed Conflict Executive Branch Foreign Relations & International Law

Using Force Against Mexican Drug Cartels: Domestic and International Law Issues

Ashley Deeks, Matthew Waxman
Wednesday, December 13, 2023, 1:58 PM

Executive branch lawyers could argue that using force against cartels would be consistent with past presidential uses of force, but it would be very difficult to defend under international law.

Former President Donald Trump with Mexican President Andrés Manuel López Obrador in the Rose Garden, July 2020. (Official White House Photo by Tia Dufour,; PDM 1.0 DEED,

Published by The Lawfare Institute
in Cooperation With

Over the past year, a range of presidential candidates and members of Congress have argued that the United States might benefit from conducting military strikes in Mexico against Mexican drug cartels and drug labs. The New York Times reported several months ago that former President Trump would seek to use military force in Mexico if he is reelected and that Trump’s “notion of a military intervention south of the border has swiftly evolved from an Oval Office fantasy to something approaching Republican Party doctrine.” Some Republican House members went even further and introduced an authorization to use military force (AUMF) against fentanyl-trafficking cartels, a proposal with more than 20 co-sponsors. 

While there is ample commentary about the wisdom of this idea as a policy matter, there is limited analysis about the legal questions that would arise if the United States sought to use force in Mexico against drug cartel members or drug labs without the Mexican government’s consent. This piece considers some of those issues. In sum, executive branch lawyers could argue that such action would be consistent with past presidential uses of force, even without an AUMF, but it would be very difficult to defend under international law and, as others have pointed out, would be seriously misguided as a policy matter. 

Would the U.S. Use of Force Against Mexican Drug Cartels Comply with International Law? 

Let’s start with the international law questions. The situations in which a state lawfully may use force against or inside another state are well rehearsed: when the UN Security Council has authorized that force, when the other state has given consent, or when the first state is acting in self-defense. In the situation at hand, it is safe to assume that the Security Council will not authorize the United States to use force against the Mexican cartels. Nor is Mexican government consent on the table: President Andres Manuel Lopez Obrador has asserted that his government will not permit a foreign government to intervene in Mexico’s territory and that U.S. military intervention would violate Mexico’s sovereignty. This is consistent with Mexico’s recent approach to U.S. agents on Mexican soil—one of great caution and skepticism.

That leaves self-defense (or anticipatory self-defense) as the only possible international law justification. As Article 51 of the UN Charter reflects, a state’s right to self-defense is triggered by an armed attack, which a range of states interpret to include a threat of an imminent armed attack. Further, the United States—and various other states—take the view that a state can use force inside another state in self-defense against non-state actors where the territorial state is unwilling or unable to suppress the threat posed by those actors. (Other states contest this idea, arguing that a state’s use of force in such a situation violates Article 2(4) of the Charter.) Finally, any U.S. use of force in self-defense would need to be both necessary and proportional.

Although there is no consensus among states about what precise activities constitute an armed attack, the United States has taken the view that there is no particular threshold of severity that a use of force must cross to constitute an armed attack. Former State Department Legal Adviser William Taft wrote in 2004, “The gravity of an attack may affect the proper scope of the defensive use of force ... , but it is not relevant to determining whether there is a right of self-defense in the first instance.” Other states (and the International Court of Justice in the Nicaragua case) believe that only the most grave forms of force constitute armed attacks and, thus, have set a higher bar for which acts trigger the right of self-defense.

Though kinetic operations such as missile strikes and ground invasions are easy to identify as armed attacks, technological developments have put the definition of “armed attack” to the test. For instance, states and commentators have considered in recent years whether and when to characterize certain hostile cyber operations as armed attacks. In 2012, Harold Koh set forth how the U.S. government thinks about this question:

In analyzing whether a cyber operation would constitute a use of force, most commentators focus on whether the direct physical injury and property damage resulting from the cyber event looks like that which would be considered a use of force if produced by kinetic weapons. Cyber activities that proximately result in death, injury, or significant destruction would likely be viewed as a use of force. 

Experts who drafted the 2013 Tallinn Manual 1.0, a nonbinding study that attempts to reflect how international law applies to cyber uses of force and cyber operations during armed conflict, agreed that whether “a cyber operation constitutes an armed attack depends on its scale and effects” and that “any use of force that injures or kills persons or damages or destroys property would satisfy the scale and effects requirement.” Those experts also analyzed which effects of a cyber operation count for purposes of assessing damage, concluding that “all reasonably foreseeable consequences” qualify.

Proponents of military operations in Mexico have identified several different cartel activities as threats, but do they meet the “armed attack” threshold in light of this legal framework?

1.   Use of force against U.S. citizens or U.S. facilities in Mexico

The draft AUMF against cartels (H.J. Res. 18) refers to a 2022 grenade attack against the U.S. Consulate in Nuevo Laredo, Mexico. If Mexican cartel members staged an attack on a U.S. consulate and killed, injured, or held hostage U.S. employees, the United States readily could conclude that it had a right to rescue or defend its nationals, including by using force. The same result might hold if a cartel kidnapped or killed Drug Enforcement Administration agents or other U.S. officials working with the Mexican government in Mexico. However, any military operations would need to be both necessary and proportional, limited to the goal of rescuing those in danger and removing them from the threatening situation. Before acting, the United States would need to conclude that Mexico is unwilling or unable to address the threat to those U.S. nationals. Those favoring military intervention would argue that Mexico has been unable for years to stop the cartels’ drug operations and uses of violence. 

It would be harder to make a case for using force if cartels captured a group of U.S. citizens, but it was unclear whether they were targeted because of their nationality. For example, in March 2023, four Americans were kidnapped in northeastern Mexico, and two of them were killed. News reports indicate that a drug cartel kidnapped them because the cartel mistakenly believed they were Haitian drug smugglers. There was no indication that the United States perceived the kidnapping and murders as an “armed attack” on the United States. In general, the United States has been willing to use force inside other states to defend its nationals when they come under threat and the host state is unwilling or unable to act, but such operations receive the most international support when they are narrowly tailored to executing the rescue, cause no collateral damage, and do not appear to be a pretext for achieving a different U.S. foreign policy goal.

2.   Border shootings

H.J. Res. 18 states that “Mexican cartels have repeatedly fired at law enforcement and national guard posted along the United States-Mexico border.” It is difficult to fact-check this claim, and the resolution does not provide evidence to back it up. Let’s assume, however, that cartel members in Mexico shoot and kill several U.S. military and law enforcement officers who are stationed inside the U.S. border. In the same article cited above, Taft stated, “[I]f the United States is attacked with deadly force by the military personnel of another State, it reserves its inherent right preserved by the U.N. Charter to defend itself and its citizens.” He did not discuss whether the United States would take the same position if a non-state group used deadly force against U.S. military personnel, though the U.S. treated the Libya-sponsored terrorist attack on the La Belle nightclub in Berlin as an armed attack and responded with missile strikes in Libya. Even if the United States viewed the cartel shootings as an armed attack, it would have to assess that a forcible response was necessary and proportional to the threat. A range of other states (and the International Court of Justice) likely would construe a shooting by non-state actors of several U.S. officials on the border as a use of force that falls short of the gravity required to render something an armed attack. Accordingly, they would not view the United States as having a right of national self-defense in this setting.

3.   “Kinetic actions against law enforcement, military, or other governmental personnel” of any country in the Western Hemisphere

H.J. Res. 18 would authorize the United States to use force in response to “kinetic actions against law enforcement, military, or other governmental personnel” of any country in the Western Hemisphere, including countries “with a common border with the United States.” That is, if a cartel used force against Mexican government officials—which cartels do on a regular basis—the United States could use force. There is no support in international law for this proposition, because attacks on Western Hemisphere government officials do not constitute attacks on the United States. If Mexican cartels engaged in armed attacks on government officials in, say, Colombia, Colombia conceivably would have a right to use force in Mexico against the cartels in self-defense (along the lines of scenario 2 above), and conceivably could ask the United States to participate in Colombia’s collective self-defense.

However, cartel attacks on Mexican officials are not “armed attacks” under the UN Charter; Mexico’s response to this domestic threat would not violate Article 2(4) and would not require a resort to Article 51 as justification. Therefore, unless Mexican officials sought forcible assistance from the United States to address cartel violence, a U.S. use of force in Mexico under this proposed domestic authorization would lack a basis in international law.

4.   The flow of fentanyl into the United States

A final scenario—one that has attracted serious attention from Republican presidential candidates—is the idea that the volume of fentanyl sent into the United States, which has led to the deaths of thousands of Americans, rises to the level of an attack on the United States. Proposals to address this threat include bombing fentanyl labs or sending in U.S. special forces to capture or kill cartel members and destroy their labs.

It’s true that a significant number of Mexican drug cartels engage in a financially lucrative criminal enterprise that results in a substantial number of deaths inside the United States. It’s also true that fentanyl is an inherently unsafe product, and those selling it surely can foresee a range of downstream harms, including addiction, violence that inevitably surrounds the drug trade, and death from overdoses. There is little evidence, however, that the cartel members actively desire that result. Nevertheless, could the deliberate introduction of a large quantity of harmful and frequently lethal substances—in other words, poison—into the United States constitute an armed attack?

The “cyber armed attack” analysis offers an instructive analogy here. If a state were to use a cyber operation to cause destructive kinetic effects, including the loss of life, at a nuclear reactor, the clear intentionality and proximate causation between the cyber tool and the harm might well lead states to conclude that the cyber operation constituted an armed attack. But if a state were to use cyber tools to shut down another state’s electric grid, causing significant economic cost to the latter state and, through a chain of events, the loss of life due to power outages at hospitals and traffic light failures, states would be less likely to conclude that the operation constituted an armed attack. The intentionality and foreseeability of the harm, while perhaps significant, is far less than in the nuclear reactor scenario.

The fentanyl case looks much more like the latter cyber case. The causal link between the cartel actions and the harm is far more attenuated than the typical armed attack scenario. Indeed, if the United States construed the cartels’ actions—which are motivated by economics but produce some level of foreseeable harm after several steps in a causal chain—as an armed attack, the United States would weaken its long-standing position that economic sanctions (which also are economic tools that can produce some foreseeable harm to the population of a target state through an attenuated causal chain) cannot be viewed as uses of force.

The facts as they exist today make it difficult to identify the existence of an actual or imminent armed attack. Any use of force by the United States in Mexico against the cartels without Mexican consent would thus not be consistent with international law.

Would the U.S. Use of Force Against Mexican Drug Cartels Comply with U.S. Constitutional Law?

Aside from the international law objections, could the president take military action without congressional authorization? The constitutional hurdles, though controversial, are lower than the international law ones. If one accepts the descriptions of threats or attacks mentioned above and views the issue in terms of the executive branch’s broad interpretation and past practice of Article II presidential power to use force, the constitutional justification begins to take shape. 

For starters, there’s some relevant U.S.-Mexico history here, beyond the Mexican-American War that established the main border that we’re discussing. More than a century ago, President Wilson claimed constitutional power to bomb and seize the port of Veracruz when a civil war ripped through Mexico. In 1914, Wilson thought that U.S. intervention would protect American nationals and commercial interests from violence south of the border and tip the balance against Mexican leadership that Wilson viewed as illegitimate. Using a minor incident—the Mexican government’s refusal to apologize or offer U.S. naval forces a 21-gun salute after having arrested a group of American sailors—Wilson asked Congress for authority to use force. Notably, he did so while claiming that he didn’t need Congress’s blessing: 

No doubt I could do what is necessary in the circumstances to enforce respect for our Government without recourse to the congress, and yet not exceed my constitutional powers as President; but I do not wish to act in a manner possibly of so grave consequence except in close conference and cooperation with both the Senate and the House. I therefore come to ask your approval that I should use the armed forces of the United States.

Congress obliged, and passed what looks something like an AUMF, declaring that “[t]he President of the United States is justified in the employment of armed forces of the United States to enforce the demands made upon Victoriano Huerta for unequivocable amends to the Government of the United States for affronts and indignities committed against this Government by Gen. Huerta and his representatives.” The chauvinistic resolution says that the president was “justified,” though, without affirmatively “authorizing” anything, perhaps a nod to the president’s independent power to use force. The resolution also included language that the United States “disclaims any hostility to the Mexican people or any purpose to make war upon Mexico.” In any event, Congress didn’t pass this resolution until after the president had also acted militarily—a military expedition that turned out to be a big strategic blunder. Two years later, Wilson would blunder again when he sent a U.S. military expedition across the border against Pancho Villa, whose forces had conducted deadly cross-border raids into the United States. This unsuccessful intervention lacked congressional authorization and further damaged bilateral relations despite Wilson’s flimsy claim that the military actions somehow respected Mexico’s sovereignty. Besides their relevance to U.S.-Mexico relations, we mention these prior cases because, in the modern era, the executive branch draws on historical practice in justifying presidential power to use force. 

Now, Republican presidential candidates who have said they were prepared to use military force in Mexico have not clearly stated whether they would do so without express authorization from Congress. Over at Just Security, Brian Finucane summarizes discussion of this issue at a recent House hearing on AUMF reform, during which Biden administration officials were asked about the president’s authority to take military action in Mexico. In response, Defense Department General Counsel Caroline Krass recited the familiar two-part test often employed by the Justice Department’s Office of Legal Counsel (OLC), namely, that the president may unilaterally order the use of force when it would serve an important national interest and if the anticipated nature, scope, and duration of the military operation does not amount to “war” within the meaning of Article I’s Declare War Clause. 

This is not a stringent test. If a president were to assess that using force was in the national interest (a very low bar) and was reasonably likely not to escalate into a major, extended military conflict (an easy prong to meet, too, and one factually possible here, notwithstanding our own concerns that any initial, limited U.S. use of force could well become protracted), then according to prior OLC positions, the president probably has power to act on his own. This would trigger the War Powers Resolution and questions of what would happen beyond 60 days of military action. But the ease with which the constitutional test is met at the front end is one reason that OLC’s expansive interpretation of Article II is widely criticized

There might be another constitutional argument altogether, though. If the premise described above is correct—that for the purposes of international law the drug flow from Mexico really is an ongoing attack or invasion—then even under a much more restrictive view of Article II powers to use force, the executive branch might take the position that the president is exercising his implied power to repel direct attacks. As with the international law argument, such an argument would blow a giant hole in a narrow implied power by essentially saying that nonmilitary cross-border harms can be functionally equivalent to direct, and perhaps continuing, military attacks for constitutional purposes.

One might take the view that if international law bars the use of force in Mexico, then the president is also legally disabled from exercising Article II powers in this way. But it was  William Barr—who recently endorsed the proposed cartels AUMF and argued that the United States should “go after [the cartels] as we would a terrorist group”—who wrote as head of OLC in 1989 that the president may order cross-border law enforcement activities notwithstanding any customary international law or non-self-executing treaty prohibition. That memo concerned law enforcement actions without host-nation consent, but its logic would seem to apply also to military force: “[E]ven if [UN Charter] Article 2(4) were construed as prohibiting certain forcible abductions, we believe that the President has authority to order such actions in contravention of the Charter.” The current status of that 1989 memo is unclear, but it is not hard to imagine that an administration willing to use force in Mexico would also adopt this constitutional position.


Legal issues aside, a president’s use of force against drug cartels inside Mexico without Mexico’s consent—or even the threat of force—would pose huge policy risks. 

One of the main questions in both the international law and constitutional law analyses is whether the use of force would be based on direct “kinetic” harm to U.S. persons or based on indirect harms flowing from drug trafficking. A use of force in the first type of case, while still subject to strong objections, at least fits more neatly into existing doctrinal formulas. A use of force in the second, more causally attenuated case would create a radically broad precedent, one that would be subject to future misuse by states (and other administrations) in ways that one can only start to imagine.

Ashley Deeks is the Class of 1948 Professor of Scholarly Research in Law at the University of Virginia Law School and a Faculty Senior Fellow at the Miller Center. She serves on the State Department’s Advisory Committee on International Law. In 2021-22 she worked as the Deputy Legal Advisor at the National Security Council. She graduated from the University of Chicago Law School and clerked on the Third Circuit.
Matthew Waxman is a law professor at Columbia Law School, where he chairs the National Security Law Program. He also previously co-chaired the Cybersecurity Center at Columbia University's Data Science Institute, and he is Adjunct Senior Fellow for Law and Foreign Policy at the Council on Foreign Relations. He previously served in senior policy positions at the State Department, Defense Department, and National Security Council. After graduating from Yale Law School, he clerked for Judge Joel M. Flaum of the U.S. Court of Appeals and Supreme Court Justice David H. Souter.

Subscribe to Lawfare