Published by The Lawfare Institute
in Cooperation With
In May 2019, President Trump pardoned former 1st Lt. Michael Behenna for crimes committed while deployed in Iraq. Behenna, who had served less than five years in prison by that point, had executed an unarmed Iraqi he believed was responsible for an attack that killed two of his troops. In November, the president followed up his historic act of executive clemency with two more. He pardoned former 1st Lt. Clint Lorance (also convicted of murder in combat, this time in Afghanistan). And—after much public commentary—he pardoned former Special Forces Maj. Mathew Golsteyn, who had been charged with killing a detainee and associated offenses, but who had not yet faced trial by court-martial.
Contemporaneously, Trump reversed a Navy admiral’s decision to grant convicted Navy SEAL Edward Gallagher limited clemency: what had been a reduction in only one pay grade became a full restoration of his rank as a chief petty officer. Further sparking outrage was Trump’s order to foreclose the Navy Special Warfare Command’s administrative review process that could have stripped Gallagher of his SEAL status before his retirement. Defense Secretary Mark Esper fired Secretary of the Navy Richard Spencer for his behind-the-scenes attempt to block the president from interfering in this process.
The Behenna, Lorance and Golsteyn pardons and the Gallagher relief were the first instances of a president pardoning or granting extraordinary clemency to a soldier for an offense constituting a war crime. But Trump is not the first president to grant clemency to service members who have violated norms, codes of conduct and criminal law through their actions in combat. President Lincoln famously interjected his vision of justice in cases of Union soldiers accused or convicted of the grave wartime offense of desertion, stopping scheduled executions to the chagrin of commanding generals. President Andrew Johnson granted general amnesty and pardoned the vast majority of ex-Confederates during the Reconstruction era. President Nixon did not pardon Lt. William Calley after Calley’s conviction for the My Lai Massacre, but—with significant public support—Nixon moved Calley out of prison and into house arrest during his appeals, eventually resulting in an early release after a handful of years for the person chiefly responsible for the deadliest war crime in American history. President Obama controversially commuted the sentence of former Army private Chelsea Manning, who had been sentenced to 35 years in prison for a massive leak of classified and sensitive documents related to the global war on terror to Wikileaks.
Though Trump’s acts of judicial mercy on service members may not be wholly original, they have made him the first president to pardon soldiers—in this case, officers—already convicted of having committed offenses that violate the international law of war. Like most presidential pardons, his acts have garnered both applause and substantial criticism.
One notable source of criticism has been from within the current and former ranks, or those who know something of the traditional military ethos. Naval War College and Naval Postgraduate School ethics professors recently wrote: "The pardons of our war criminals by Trump, and his interference in and disrespect of our own military justice system is unprecedented and should trouble all Americans. We will not pull punches—they are shameful and a national disgrace."
Another representative example is retired Lt. Gen. David Barno, who argues that President Trump did not give sufficient consideration to the views of his advisers, the unambiguous results of due process under military law, the collateral consequences for our soldiers on the battlefield or obligations under the law of war. This argument is not overstating the facts. When the president’s defense amounts to sanctifying brutal acts of soldiers he thinks are trained as, and expected to be, “killing machines,” it is difficult to believe the president has given that due consideration. Rather, it seems to many as if he is “pushing the buttons of government indiscriminately,” a real-world illustration of his claims that Article II of the Constitution allows him to do “whatever” he wants.
The nature of these combat-incidental crime pardons should give caution to presidents claiming their moral duty or constitutional prerogative to grant clemency at will and without judicial or congressional review. First, war crimes (at the very least those of which Behenna, Lorance and Golsteyn were accused) implicate the well-known and universally accepted expectations, duties and rights of international law. Rules founded on basic principles of humanity, chivalry and honor regulate who may use force, against whom such force may be used, what places or things may be attacked, and what weapons may or shall not be used in those attacks and use of force. How Americans—as a nation through public discourse and as a military on the battlefield—address violations of these rules will signal something to the larger international community of current and potential allies, partners, competitors and enemies. It signals the tactical and political valuation by the United States of humanitarian practices and standards in conflict.
Second, war crimes implicate not just actions deemed criminal by customs, treaties and domestic military law. They also violate core customs, traditions, and standards of conduct and ethics that (when obeyed) further positive goals of self-regulation within the profession of arms.
Third, the president as the military’s commander-in-chief has a different moral, legal and practical standing in relation to both the military offender and the war crime offense itself. The president is the ultimate superior in the chain of command. The service member convicted of war crimes could not have acted when and where he did but for the president’s express order or tacit acceptance of the military operation within which the service member dutifully executes a mission. While the president is clearly not legally complicit in the wrongful act, his constitutional duties imply a moral responsibility for the enabling context of the wrongful act. In that sense, pardoning a war criminal of one’s own military appears to be a conflict of interest, broadly understood. President George H.W. Bush was condemned roundly for pardoning his former White House colleagues for their role in the Iran-contra affair—actions that likely happened with Bush’s situational awareness while he was serving as vice president, and which he defended as being motivated by their “patriotism.”
A possible consequence of failing to properly balance these legal and practical realities and authorities is a strong disagreement between the military agent and the civilian political principal who pardons. When this disagreement reflects fundamental differences over what is morally-permissible conduct on the battlefield, the civilian political leader’s “right to be wrong”—if exercised over the objections of four-star service chiefs of staff and the civilian service secretaries—risks at least four considerable penalties and costs. First, it risks alienating those in uniform or who have been in uniform who believe such conduct was immoral or illegal and therefore beneath them, damaging the institution and its professional reputation. Second, it risks undermining the confidence the military agent has in the civilian principal’s knowledge, intentions and good faith. Third, it risks signaling civilian disregard for the very military due process for which the commander-in-chief is responsible for managing as a specialized criminal justice system. And fourth, it risks signaling preapproved permission to engage in similar acts with similar intentions to those in uniform who are facing or who may face circumstances risking moral injury—what a recent New York Times editorial called the signature “defining injury of the global war on terrorism.” This combination of risks is too strong for a civilian principal to ignore.
For all these reasons, presidents ought to restrain their impulses when certain conditions are present. Call it the combat-incidental crime pardon exception.
But how to deny a president the power to pardon service members accused or convicted of war crimes by courts-martial? The most obvious route would be to amend the Constitution—but this is, of course, highly improbable. Instead, the method most respectful to the principle of separation of powers is to amend not the Constitution, but the Uniform Code of Military Justice (UCMJ), which could be done by Congress. This course of action recognizes that criminal activity by service members and their combat actions (especially when they arise from the same facts) raise the interests of both Congress and the commander-in-chief, both of whom have constitutionally prescribed responsibilities for the regulation and use of those service members. The Supreme Court has repeatedly acknowledged that military justice’s unique elements—primarily the role of the commander, exercising both prosecutorial and judicial functions—justify different applications of civil liberties and generally do not violate constitutional protections. With this in mind, it is not impossible to imagine a UCMJ amendment that imposes a limited restraint on presidential discretion and passes the Supreme Court’s muster.
Nevertheless, the probability of amending the UCMJ over a likely presidential objection is near zero. With a formal legal mechanism implausible, the better way to understand this restraint is as a presidential self-denial of otherwise unilateral discretion.
But this denial is conditional and triggered by various presumptions. As Immanuel Kant wrote, the right to pardon “is certainly the most slippery of all rights of the sovereign .... [It] can demonstrate the splendor of his majesty and yet thereby wreak injustice to a high degree.” Presumptions in favor of granting a pardon (or not to) should depend on the timing of the possible pardon decision relative to where it would occur in the military justice process. This conditional framework accounts for the fact that pardons for individuals’ actions during combat implicate not only the president’s Article II pardon power but also a president’s Article II role as commander-in-chief, the president’s role in executing the congressionally enacted and regulated military justice system under Article I, the principal-agent character of the civil-military fiduciary relationship, and—critically—the duties and rights the United States has subscribed to under the international law of war. Pardoning a civilian for something like obstruction of justice, tax evasion, illegal campaign contributions or even murder is categorically and normatively different from pardoning a war criminal. Assuming there is a rational, nonarbitrary calculus, this decision ought not to be colored predominantly by the moral judgment of the president as chief magistrate of the laws—much less by the judgment of how such a pardon might be politically advantageous.
Under normal conditions, pardons may be used by a president to provide clemency to a person who might yet be tried for, or has already been convicted of, a federal offense. It is a plenary and totally discretionary act, shielded from any external review other than the opinion of the electorate. A pardon deletes the conviction from the record or prevents the possibility of conviction, but the facts of the crime do not disappear; there is simply no longer a legal consequence for them. It is an act of “official forgiveness” from the senior representative of the entire U.S. population. That absolution or legal forgiveness is not meant to, nor does it in practice, reverse the “general deterrence” purpose behind that criminal prohibition or its specific prosecution in that pardon petitioner’s case. Rather, according to the seminal 1866 Supreme Court decision Ex parte Garland, the pardon “releases the punishment and blots out of existence the guilt, so that, in the eye of the law, the offender is as innocent as if he had never committed the offense.”
Military crimes and their sanctions, established by Congress in the UCMJ and managed by the president down through individual judicial and prosecutorial discretion of subordinate commanders, are fundamentally different from other crimes and sanctions. They apply to a specific and narrow community employed for specific purposes. These criminal proscriptions are only constitutionally lawful, even when they seemingly breach constitutional norms or otherwise sacrosanct civil liberties, to the extent that they ensure that this specialized community is able to accomplish its mission on behalf of the nation.
The Framers gave the power to regulate conduct by its military members to Congress but provided the power to use the military members as an armed force to accomplish security and defense objectives to the president. The resulting justice system reflects Congress’s judgment about what conduct is criminal, and Congress by law delegates to the president the authority to control the court-martial procedure and prescribe punishments for violations, and even act as a court-martial convening authority as if he were a senior uniformed commander. (This is not a role the president plays in any other criminal justice system.) Congress has provided the military chain-of-command discretion in individual cases to determine whether some conduct is “service-discrediting” or “prejudicial to good order and discipline” to the extent that it should be criminally prosecuted, even if it could never have been in an ordinary criminal court. The Supreme Court has validated (with some exceptions) the military’s determination about what sort of conduct might be so proscribed. Congress and the president have given military commanders a role in investigating potential crime, in deciding when to prosecute certain crimes and how, and in approving some sentences.
This criminal law system also deliberately and self-consciously incorporates international humanitarian law (also known as the law of war, or the law of armed conflict). Under the “combatant’s privilege,” international humanitarian law permits nations and individuals to engage in some conduct that would be impermissible and criminalized even under the UCMJ. But this body of law imposes additional layers of duties and prohibitions that apply only in the context of armed conflict.
For this reason, what constitutes a war crime under military justice is highly contextual. Because of the general default presumption of combatant immunity, that context is even more relevant than the context associated with typical issues of “excuse” and “justification” that shape criminal prosecutions. The actions that could be labeled as “criminal” are often taken under extraordinary pressures of time, responsibility for the lives of others, a mission dictated by a superior chain-of-command, and possibly being engaged by a hostile force at the time the decision in question is made or in the immediate aftermath of such decisions. Moreover, the action (or failure to act) that could constitute a crime could not have been committed but for the fact that the service member was in a particular place, doing a particular job, under the lawful authority, responsibility and direction of the president.
If law holds individual agency to be a key factor in determining a person’s criminal culpability, behavior in combat reflects a kind of shared agency. This shared agency does not diminish the soldier’s culpability. Rather, it accentuates the political leader’s role for the express purpose of diminishing that leader’s unilateral discretion to forgive and remove the stain of that culpability.
As a nation, the United States has by law and policy subscribed to these duties and prohibitions under international law, and the U.S. holds its military to those standards and rules. They are cemented in doctrine and reinforced through training. One of those duties is to investigate and hold accountable those responsible for violations of international humanitarian law. The process and substance of the criminal law system in the military is fundamentally distinct from any other criminal justice system.
All this means that the decisions for pardoning war crimes should be distinct from those that guide pardons of any other crime in a civilian justice system, or even pardons for another non-war crime offense under the UCMJ. One of the main purposes of pardons is to erase or preclude punishment when the criminalization of behavior is unjust—and in the modern context, this has proved especially true when that behavior carries mandatory minimum punishment. But in the context of a war crime, virtually nobody can say that it is unjust to criminalize the killing of unarmed detainees without due process, and most military crimes carry no mandatory minimum sentence. Such a categorical distinction is reasonable because the pardon’s consequence signals that the president as commander-in-chief has validated, excused or justified the particular conduct that was so highly contextualized in the military combat domain. And that conduct was possible only because the president ordered the service member’s participation in that context.
The risk of such a pardon is that other service members facing similar contextual facts, operating under similar pressures and constraints, may view that validation as permissive precedent. The community may not view it as a singularly unique instance in which a president, as chief law enforcement official, decides in the interest of justice to undo a civilian conviction or preclude a civilian trial. Rather, the pardon communicates an unexpected endorsement of behavior that is expected to be deemed morally and criminally wrong. Not only that, but such pardons go beyond merely disrespecting the law of war. They actively undermine well-known customary practice and treaty law-imposed duties to investigate and hold responsible parties accountable for violations.
The following set of additional factors ought to guide presidents in exercising their Article II power to pardon when the beneficiary of that clemency is a service member accused or convicted under the UCMJ for crimes proscribed by the law of war:
If the service member has been convicted by court-martial and the appellate process through the Court of Appeals for the Armed Forces (CAAF) is complete (any remedy for the soldier has been granted or denied by the judicial process), do not pardon.
If the service member has been convicted, but the appellate process is not yet complete, presume no pardon. Grant only if an objective and prudent person, knowing the relevant facts, would likely not think that the United States tolerates conduct that constitutes a war crime under international humanitarian law; and the rationale for clemency outweighs the recommendations of the relevant civilian and military chain-of-command; and if an objective and prudent military commander would agree that an enemy belligerent, under similar circumstances, would deserve a pardon from his or her own government for conduct committed against a U.S. civilian or service member.
If the service member has been charged, but court-martial adjudication at trial is not yet complete, presume no pardon. Grant only if doing so satisfies any of the three conditions above.
If not yet charged, do not grant a pardon, and do not engage in or seek to influence the UCMJ disposition decision. Doing so raises the specter of undue influence, if not “unlawful command influence” that unjustifiably taints the public’s perception of the system’s fairness and due process.
Per Congress’s direction in Article 33 of the UCMJ, the president and the secretary of defense have published “disposition guidance” to military commanders exercising their prosecutorial discretion under the UCMJ, which includes a handful of “inappropriate considerations” not to be taken into account when weighing a potential prosecution. Likewise, there are inappropriate pardon considerations: rank of the service member; character of the service member’s combat experience; previous professional awards or recognition for performance of duties; results of the combat incident that served as context for the offense; collateral misconduct by the service member unconnected with the conduct constituting a war crime; the range of potential punishments available to the court if convicted; the actual punishment adjudged by the trial court or affirmed by the appellate court; and probability or promise of partisan political support from the military at large or specific individuals. The presumptions against granting combat-incidental crime pardons, unless certain conditions exist, recognize but also balance the president’s plenary power to pardon, his duty to faithfully execute the laws, his role as a civilian commander-in-chief for the military, and the nation’s ever-present duty under international law of armed conflict.
Ultimately, the difficulty of pardoning war crimes lies in that most people do not consider it a difficult question at all—the president clearly has plenary, unilateral discretion to grant a pardon. Alexander Hamilton, writing in Federalist #74, called clemency a “benign prerogative” and argued a wise president would wield this authority as a matter of case-by-case compassion to mitigate “unfortunate guilt,” or as a means to put the cork back in a potentially explosive public passion. The irony is that President Trump’s most recent pardons were not meant to provide compassionate relief for unjust prosecutions or to douse the fires of public outcry. The pardoned conduct has long been criminalized under the law of war because it lacked the necessary humanitarian compassion international law requires as a constraint on the means and methods of war. And the pardons actually inflamed public passion.
In light of implications raised by international law of war, norms of professionalism and health of stable civil-military relations, and the standing of the president relative to the offense and the offender, the president’s use of his Article II power to pardon war crimes raises fundamental issues of the rule of law. The exercise of power should be limited by laws, not enabled by the whims of individuals. These laws require citizens and those in power to obey certain rules, but they also create mechanisms for deliberate, transparent rule- and decision-making; they ought to account for a diversity of interests, not just opinions. If there is no possibility of a constitutional amendment or a restraint imposed via the UCMJ, then presidents ought to engage in deliberate self-restraint when it comes to pardoning American war criminals in a way that acknowledges the plurality of interests involved.
War crimes are faulty mirrors of “normal” crimes. Just as images are distorted by physical gravity, we see and understand war crimes as shaped by their severity and their weighty implications for professionalism, civil-military relations and obligations under international law. Following the rule of law means acknowledging the distinction between typical criminal offenses and war crimes: Their differences also demand that the decision to pardon those crimes be justified or denied in a different way.