The Impact of Military Justice Reform on Command Responsibility


| Jun 1, 2021

command responsibility. courts-martial-manual-1800


Over the past decade, the prevalence of sexual assault faced by women in the United States armed forces has prompted Congress to enact a series of reforms to the military justice system. The overarching objective of these reforms has been to reduce the influence of the chain of command in the disposition of sexual assault cases. Until now, all of these reforms have targeted various opportunities commanding officers have had to exercise leniency toward subordinates accused of serious crimes.

On April 28, 2021, however, Senator Kirsten Gillibrand, who has led this broader reform effort, introduced S.1520. Provisionally titled, the “Military Justice Improvement and Increasing Prevention Act,” S.1520 would fundamentally restructure the military justice system by removing the commander altogether from the prosecution of serious crimes. With sixty-three co-sponsors from all sides of the partisan divide, S.1520 is near certain to become law.

As most readers of Articles of War already know, the decision to prosecute a U.S. service member for a particular violation of military law (i.e., “convening authority”) has traditionally been an incident of command exercised by that service member’s superior officers. S.1520 will do away with that system. For specified serious crimes—which a current draft of the bill enumerates and which mirror those covered by UCMJ Article 60a(a)—convening authority will be exercised not by a commanding officer, but by what the press release accompanying the bill describes as “independent, trained, professional military prosecutors.” These military prosecutors will, in turn, work not within the chain of command but in the military equivalent of a civilian U.S. Attorney’s Office or District Attorney’s Office.

The precise details of how this system will operate in practice remain broadly drawn in the current draft of the bill. The one certainty, however, as reflected in one of the bill’s amendments to UCMJ Article 22, is that an officer “may not convene a court-martial under this section if the officer is in the chain of command of the accused or the victim.”

Over at Lawfare, I posted a detailed explainer of the background of S.1520 and what it attempts to do. I also offered some suggestions for how the current draft of the legislation should be clarified and revised to alleviate unintended legal problems. The first such issue I briefly addressed is the need for lawmakers to consider the ways in which removing the chain of command’s prosecutorial power over serious crimes applies differently to the special case of war crimes. Here, I explain that problem in greater detail and propose a few solutions for lawmakers to consider as they finalize the bill’s text over the coming months.

Not Prosecuted as War Criminals

The need to consider the problem arises from the fact that U.S. service members who commit war crimes are generally not prosecuted as war criminals. Service members who commit war crimes are as theoretically liable as anyone else to be prosecuted in the civilian federal courts under the War Crimes Act. But there are no examples of such prosecutions being mounted in the War Crimes Act’s quarter-century on the books. Rather, as the DOD Law of War Manual concisely states, the “principal way for the United States to punish members of the U.S. armed forces for violations of the law of war is through the Uniform Code of Military Justice.”

But other than Article 21 and the law of war jurisdiction clause in Article 18—which appear to have never been used to prosecute a service member since the UCMJ’s enactment—none of the UCMJ’s punitive articles confer jurisdiction over “war crimes” as such. American policy has instead been to prosecute U.S. service members who violate the law of war under the generally applicable punitive articles of the UCMJ that criminalize the conduct the law of war forbids.

Hence, when LT William Calley was prosecuted for the massacre at My Lai, he was not prosecuted with the war crimes he would hypothetically face at an international criminal tribunal—such as willfully killing protected persons or intentionally directing attacks against the civilian population as such. He was prosecuted for murder under UCMJ Article 118 and assault under Article 128. Likewise, when four soldiers raped an Iraqi girl, murdered her and her family, and then burned the bodies in the so-called “Black Hearts” case, they were court-martialed, not as war criminals, but for murder (Article 118), rape (Article 120), arson (Article 126), and stalking (Article 130) among other crimes.

Under S.1520, all of these crimes, and many others that cover conduct prohibited by the law of war, will no longer be under a field commander’s direct authority to punish. Indeed, as currently written, S.1520 forbids a commander from convening a court-martial to punish a subordinate who is credibly accused of committing of any these crimes, even if the underlying conduct would entail war criminality.

Responsible Command and International Law

The (apparently unintended) issue that stripping commanders of convening authority over serious crimes creates is that it undermines the United States’ international law obligation to conduct its military operations under “responsible command.” Since at least the 1907 Hague Convention, being “commanded by a person responsible for his subordinates” has been a definitional element of military organizations. While the United States’ armed forces will obviously continue to conduct operations under a chain of command, a commander’s power to punish subordinates for violations of the laws of armed conflict has been a necessary condition for any armed force to operate under a “responsible” command.

In the American military context, the principle that responsible command entails the commander’s both having and using their power to punish violations of the laws of armed conflict goes back to the Revolutionary War. In 1782, as George Washington and his British counterpart, James Robertson, negotiated the terms of battle, they were both “earnestly desirous,” in Washington’s phrase, to see that “the war may be carried on agreeable to the rules which humanity formed, and the example of the politest nations recommends.” To make that desire a reality, they mutually agreed “to prevent or punish every breach of the rules of war within the sphere of our respective commands.”

That some form of this duty fell to every commander of a modern military force as a matter of international law has been settled since at least the 1940s. The landmark case was the Army’s trial of General Tomoyuki Yamashita as a war criminal for atrocities committed during the Battle of Manila. There was no evidence that Yamashita ever ordered his subordinates to commit any of the war crimes perpetrated during that battle. In fact, the evidence showed that Yamashita issued a general order of retreat in advance of the battle and holed up in mountains with the bulk of his forces until the surrender was formally signed on the deck of the USS Missouri in September 1945. The atrocities in Manila had been committed by the 20,000 Japanese soldiers and sailors who had failed to evacuate in time. He was nevertheless sentenced to death under the doctrine that is now called “command responsibility.”

By abandoning the population of Manila, Yamashita abandoned his duty to prevent and punish atrocities that his subordinates might commit in his absence. And the Supreme Court upheld the judgment against him as a matter of international law because “the law of war presupposes that its violation is to be avoided through the control of the operations of war by commanders who are to some extent responsible for their subordinates.”

In the eighty years since, international law has consistently defined responsible command as requiring the power to “prevent or punish” violations of the law of war by subordinates. And international law has reciprocally embraced command responsibility as a legal basis on which to punish commanders who shirk that duty.

Under Article 87 of Additional Protocol I of the Geneva Conventions—which the United States considers reflective of customary international law and to which it deems itself bound—State parties are required to ensure that commanders have a ready means “to initiate disciplinary or penal action against violators” of the Geneva Conventions. The ICRC commentary to Article 87 describes “the role of commanders” as “decisive” to the effective enforcement of the law of armed conflict. And in the ICRC’s compendium of customary law, Rule 153 states that “Commanders and other superiors are criminally responsible for war crimes committed by their subordinates if they knew, or had reason to know, that the subordinates were about to commit or were committing such crimes and did not take all necessary and reasonable measures in their power to prevent their commission, or if such crimes had been committed, to punish the persons responsible.”

To be sure, neither the Geneva Conventions, nor the customary laws of armed conflict, impose a particular scheme that commanders must use to ensure that their subordinates are punished for violations of the law of armed conflict. Countries have different systems of military justice and some countries have no military justice system at all. As a group of prominent military law scholars, who were early advocates for the reforms that S.1520 aims to implement, have argued, none of the United States’ fifteen closest allies “permits non-lawyer commanders to order the trial of serious offenses by courts-martial. … All of them recognize the doctrine of command responsibility; none have found it necessary for commanders to retain disposition authority in order to accommodate that doctrine.”

Instead, as the commentaries to Article 87 attempt to clarify, a commander who discovers a subordinates’ war crimes must “propos[e] a sanction to a superior who has disciplinary power, or—in the case of someone who holds such power himself—exercis[e] it, within the limits of his competence, and finally, remit[] the case to the judicial authority where necessary with such factual evidence as it was possible to find.”

Does S.1520 Dilute the Power to Punish War Crimes?

There is nothing, therefore, per se unlawful about the world that S.1520 envisions, where military prosecutor’s offices are ultimately responsible for the prosecution of all serious military crimes, including war crimes. It may even be preferable. The overarching motivation for S.1520 has been the perceived conflict of interest that exists in the current military justice system, under which commanders’ focus on their operational objectives gives them a perverse incentive to show leniency toward subordinates who are operationally valuable. That conflict of interest is at least as significant a problem when it comes to the prosecution of war crimes as it is in cases of sexual assault perpetrated against other service members. One need only consider the difficulties in bringing the perpetrators of My Lai to justice or some of the pardons issued by former-President Trump, to doubt the seriousness with which some in the chain of command will view war crimes perpetrated against members of the enemy population.

S.1520, however, is not intended or designed to address the special problem of war crimes. And how its reforms will operate in war crimes cases deserves some consideration, particularly insofar as they might inadvertently muddle operational commanders’ sense of accountability.

It is certainly true that other allied nations have not depended upon commanders having the direct power to punish subordinates for war crimes. But the United States remains unique in terms of the scale and duration of its overseas military commitments. There is also a considerable amount of U.S. military history, tradition, regulation, and education that has built up around the premise that commanders both have and must use this power to ensure law of war compliance. In fact, in 1983, the Court of Military Appeals reaffirmed the power of superior convening authorities to increase the seriousness of charges brought against subordinates by inferior convening authorities precisely because weakening “the commander’s power to assure that crimes are referred to tribunals that can mete out adequate punishment,” would weaken the commander’s responsibility under the law of war “to control their troops and to maintain discipline.”

As written, S. 1520’s revision to Article 22 takes commanders out of the prosecutorial decision-making process entirely over most UCMJ offenses that implicate conduct forbidden by the law of war. And though the bill’s current text is ambiguous, the presumption that military prosecutors will act with “independence” certainly casts doubt on what authority commanders will have to even encourage prosecutors to refer charges against war criminals without raising the specter of unlawful influence. In the law of war context, such constraints could pose unique problems if a commander believes that their subordinates’ particular war crimes were unusually detrimental to the broader war effort and therefore deserving of conspicuous punishment. What then is left of commanders’ power to punish war crimes?

Arguably, S.1520 leaves commanders’ authority to convene courts-martial under Article 18 or other ad hoc law of war tribunals under Article 21 intact. As a historical matter, both articles trace their origins to an analogous gap in command authority that surfaced during the Mexican War. When General Winfield Scott was confronted by atrocities and abuses committed by his subordinates and camp followers, his hands were arguably tied by the existing jurisdictional constraints on the use of courts-martial. As a result, he created what we would now call military commissions to punish war crimes in the exercise of what the Supreme Court would later describe an “important incident” of a commander’s general power to prosecute a war. Doing so was not simply a means by which to maintain good order and discipline in his ranks, but also to demonstrate to the Mexican people that the U.S. military was different from the armed forces of its international peers by virtue of being an accountable, rule-of-law bound organization.

There are various problems with looking to Articles 18 and 21 as loopholes, though. For one thing, these powers have rarely been used. For another, as the decades of litigation surrounding the military commissions in Guantanamo demonstrate, the elements of common law of war offenses are far from certain even with Congressional guidance and therefore rife with litigation risks. And, as the newly revised Commander’s Handbook on the Law of Land Warfare warns, the use of Article 18 is liable to preclusion challenges, whenever service members find themselves charged with war crimes whose elements mirror offenses contained in the punitive articles. If a modern-day William Calley found himself charged under Article 18 for attacking civilians, in other words, he would have a strong legal argument that he can only be charged for the specified offenses that S.1520 will now put exclusively into the hands of military prosecutor’s offices.

Without the power to prosecute subordinates directly for their war crimes, all that is arguably left of a commander’s duties in this sphere are the fairly general reporting requirements put in place by the DOD Law of War Program. As both the floor and ceiling of commanders’ remaining responsibility for their subordinates’ war crimes, one can doubt what level of responsibility commanders can be truly said to exercise over their subordinates if their power to “prevent and punish” in a forward deployment is so constrained and indirect.

Perhaps this issue is more theoretical than real. As many commentators have noted, criminal liability via command responsibility remains conspicuously absent from the UCMJ. Hence, diluting commanders’ responsibility for their subordinates’ war crimes is unlikely to matter much in practice.

It is not obviously true, however, that the UCMJ does not make commanders liable to prosecution under a theory of command responsibility. In addition to prosecutions for dereliction of duty and under the general article (which have happened), Article 77’s extension of principal liability to those who “command” UCMJ offenses arguably encompasses command responsibility, as does Article 18’s general provision for law of war offenses.

While such prosecutions are rare to the point of non-existent, the criminal law is chock-a-block with underused laws that are directed at the conduct of senior government officials. Those laws are retained in the belief that their very existence has a hortatory deterrent effect, if only to remind officials of what is expected. Diluting the responsibility of commanders for their subordinates’ war crimes, therefore, risks inadvertently making the application of command responsibility more ambiguous and therefore weakening the ownership commanders should feel over their subordinates’ actions.

Possible Solutions for Lawmakers

As the text of S.1520 is finalized over the next few months, there are a variety of steps lawmakers could take to ensure that their broader effort to restructure the military justice system will not adversely affect the prosecution of war crimes.

The simplest solution would be to expressly grant commanders’ concurrent convening authority over war crimes. This could be done by expressly amending Article 18 to state that commanders retain convening authority over law of war offenses and that the trial of such offenses is not precluded by the specification of other punitive articles covering the same conduct. Or, Congress could include a general exception to Article 22, giving commanding officers concurrent convening authority along with the stand-alone military prosecutors for any offense for which the conduct alleged constitutes a violation of the law of war.

Neither solution would resolve the difficulties that arise from the indeterminacy of common law of war offenses. And adding an exception to Article 22 that hypothetically said, “except where the conduct alleged constitutes a violation of the law of war,” would effectively add a jurisdictional element to any such prosecution. But both reforms would alleviate any doubt that commanders retain the power and legal duty to punish war crimes to the same extent as other derelictions of duty over which they will still retain convening authority.

Alternatively, or in addition, Congress should strengthen commanders’ affirmative duty to report the perpetration of war crimes to the stand-alone military prosecutor’s offices that S.1520 now puts in charge of prosecutorial decisions. Such reporting requirements should be backed by criminal penalties, such as by making a commander’s “failure to report” itself a category of principal liability under Article 77. The law should also clarify that commanders have the power to make recommendations respecting the disposition of individual cases and that military prosecutors are permitted to base prosecutorial decisions on such recommendations without running afoul of the UCMJ’s unlawful influence prohibitions.


With S.1520, Congress’ overriding goal is to remedy what it has determined are undue obstacles to ensuring accountability for the victims of sexual assault and other serious crimes. Its focus is understandably the routine use of the court-martial system and its effectiveness as a substitute for the civilian criminal courts in serving that purpose. But lawmakers should not forget that the existence of the court-martial system has in no small part been justified on the need for a ready means of seeking justice on the battlefield.

There is no indication that Senator Gillibrand or the laws’ scores of co-sponsors have yet considered the impact of their reforms on this somewhat obscure and niche aspect of the military justice system. They should, however, because it affects the United States’ compliance with international law and because the U.S. military’s readiness to punish its own war criminals has a profound effect on its reputation and effectiveness around the world.


Michel Paradis is a leading human rights lawyer and national security law scholar. He has won high-profile cases around the globe, including some of the landmark cases to arise out of Guantanamo Bay for the U.S. Department of Defense, Military Commission Defense Organization. He is a Lecturer at Columbia Law School, where he teaches courses on national security law, international law, and the constitution, an adjunct professor at Georgetown, where he teaches the law of war, and a fellow at the Center on National Security.