The Yale Law Journal

March 2017

Why Have We Criminalized Aggressive War?

Criminal LawInternational Law

abstract. On the dominant view, accepted by both defenders and critics of the criminalization of aggression, the criminal wrong of aggressive war is inflicted on the attacked state. This view is mistaken. It is true that whether a war is criminally aggressive is determined ordinarily by whether it involves a particular form of interstate wrong. However, that is not why such wars are criminal. Aggressive war is a crime because it entails killing without justification. Five reasons explain why this is so. First, banning aggression restricted states from using force to protect their core sovereign rights, including even their rights of political independence and territorial integrity. Those core states’ rights cannot make sense of the move to ban aggression. Second, what distinguishes aggression from any other sovereignty violation—what makes it criminal, when no other sovereignty violation is—is not that it involves an especially egregious violation of territorial integrity or political independence, but that it involves killing without justification. Third, the unjustified killing account makes sense of aggression’s standing alongside genocide, war crimes, and crimes against humanity. The traditional notion that aggression is a crime against sovereignty instead isolates aggression as the inexplicably odd crime out. Fourth, the public reasons for restricting jus ad bellum rights in the early twentieth century focused not on infringements of states’ rights but on the infliction of death without justification. Finally, the importance of wrongful killing to the criminalization of aggression was apparent at the post-World War II tribunals at Nuremberg and Tokyo. Understanding the crime in this way matters doctrinally. This understanding clarifies the boundaries of the crime, resolving hard cases like unilateral humanitarian intervention and bloodless invasion. It also has implications for the legal rights of soldiers involved on either side of such wars.

author. Lecturer in Human Rights, University College London. For potent contributions ranging from conceptual discussions prior to the first draft to granular criticism of the fully drafted text, I am deeply indebted to Dapo Akande, Gary Bass, Charles Beitz, Kiel Brennan-Marquez, Amy Chua, David Crane, Oona Hathaway, Paul Kahn, Katerina Linos, Itamar Mann, Daniel Markovits, Frédéric Mégret, Marko Milanovic, Michael Reisman, Jed Rubenfeld, Kim Lane Scheppele, Scott Shapiro, John Witt, and participants in the Oxford Public International Law seminar and the American Society of International Law New Voices session. I am also very grateful to the Yale Law Journal editorial team, led by William Stone, for outstanding comments and suggestions, as well as for their work in shepherding this Article through to publication. As always, any errors are my sole responsibility.


Nearly seventy years have passed since an international tribunal has convicted a defendant of the crime of aggressive war. Nonetheless, beginning this year, aggression is likely to join the list of violations over which theInternational Criminal Court (ICC) has jurisdiction.1 To understand the implications of this development, and to grasp the controversies surrounding it, we need to be clear about what is criminally wrongful about aggressive war. This Article investigates that question.

Like most elements of international criminal law, aggression finds its foundation in the statutes and jurisprudence of the post-World War II tribunals at Nuremberg and Tokyo.2 However, it is an anomalous crime in several respects. The statutes of every post-Cold War international and hybrid criminal tribunal other than the ICC have ignored aggression, even as the General Assembly has repeatedly endorsed its status as an international crime.3 This marginalization stands in stark contrast to the central focus placed on war crimes, crimes against humanity, and genocide in those tribunals. Even aggression’s initial codification in the Rome Statute in 1998 was extraordinary. Despite being included as one of the four categories of crime, aggression was a placeholder bereft of content; the Statute required a subsequent definitional amendment before it would come into effect.4

The scope of criminal responsibility for aggression is also unique. At Nuremberg and Tokyo, the tribunals held that soldierly obedience was no excuse for participation in war crimes and crimes against humanity.5 In the case of aggression, however, they restricted criminal liability exclusively to members of the German and Japanese leadership cabals.6 The ICC amendment makes this leadership element explicit and buttresses it with a provision specific to aggression that narrows significantly the complicity doctrines on which almost every post-Cold War international conviction has hinged.7

Finally, and most significantly, aggression is widely understood to be rooted in a moral wrong “committed against a state” rather than in wrongs “against individuals.”8 In an open letter urging States Parties not to proceed with the incorporation of aggression, a coalition of pro-ICC human rights activists stressed precisely this normative contrast between the state-focused crime of aggression and the human-focused crimes of genocide, war crimes, and crimes against humanity.9 The distinction is arguably implicit in every relevant international criminal law provision, from Nuremberg and Tokyo to the ICC amendment.10

The most influential moral accounts of the crime of aggression also understand the wrong of aggression in these terms. For Michael Walzer and others defending its criminal wrongfulness, aggression is fundamentally a crime against the political collective, rooted in a “domestic analogy” in which states “possess rights more or less as individuals do.”11 Critics of the criminalization of aggression adopt the same understanding of the internal normative posture of the law, but object to its classification alongside the other international crimes precisely because it privileges sovereignty over humanity.12 For them, this feature of aggression contradicts what they take to be the defining moral thrust of international criminal law. Walzer and these critics disagree on whether the sovereignty violation that occurs in an aggressive war is a moral wrong worthy of criminalization. However, from the internallegal point of view, they agree that the crime as currently constituted is rooted in that claimed wrong.

Other international crimes, like genocide, also involve wrongs against a collective entity, but what is special about the putative wrong underpinning the crime of aggression on both of these competing views is that it occurs exclusively on the macro level. No one would deny that the individual men and boys killed at Srebrenica were the victims of genocide, even though it was also a crime against Bosnian Muslims as a group.13 In contrast, individual soldiers killed while fighting their states’ lawful wars against aggression are generally not thought to be victims of a crime.14 Or so the traditional normative account goes.

This Article rejects that understanding. Not only are individuals, including soldiers, wronged gravely in an aggressive war, the wrongfulness of their treatment as individuals is the very crux of what explains the criminalization of aggression. It is the normative core of the crime.

To be clear, the dominant view, shared by both defenders and critics of the criminalization of aggression, is correct in one respect: a war of aggression is an interstate breach and, typically, a violation of sovereignty. In other words, whethera war is criminal depends in large part on whether that interstate breach occurred. However, imputing personal criminal responsibility for an interstate breach is anomalous; that violation is not why aggressive war is criminal. The core moral problem with aggressive war is neither that it infringes sovereignty, nor even the extent to which it infringes sovereignty. Indeed, in at least one circumstance, it need not infringe sovereignty at all.15 Rather, the core moral issue is that aggressive war entails killing and maiming for reasons that are now considered unacceptable: reasons other than the protection and security of human life.

This Article establishes and defends this internal normative account of the crime of aggression. Achieving normative clarity in this respect is not merely an academic or theoretical exercise. Coherent interpretation of the law requires understanding its moral underpinnings. Excavating and clarifying those foundations can have profound doctrinal and structural effects. Understanding that the criminal wrong of aggressive war is its unjustified infliction of death and human suffering has three material consequences.

First, it clarifies how to interpret hard cases. Even if one accepts the dominant view that humanitarian intervention without Security Council authorization is illegal, the normative framework offered here weighs heavily against an interpretation on which such action would be criminal. By responding defensively to the internationally criminal, massive infliction of human harm, a genuine humanitarian intervention lacks the core wrong that makes aggressive war worthy of criminalization. Similarly, “bloodless” military invasions are best interpreted as noncriminal, despite their extremely effective and illegal usurpation of territory and political control. To be clear, this does not mean that such invasions cannot be countered lawfully by means such as defensive force and sanctions (unilateral or collective). Nor does it mean that the ensuing occupation is lawful or shielded from other legal remedies. It only means that such actions are best interpreted as falling below the demanding threshold of criminality.

Second, normative coherence requires international law to take seriously the human rights and refugee claims of soldiers who refuse to fight in aggressive war. Some adherents to the traditional account see soldiers who fight in criminal wars as contributory cogs, no more intimately involved than taxpayers in a macro wrong against a foreign state. This is a mistake; such soldiers perpetrate directly the constituent wrongs of the criminal action. There are good reasons not to hold them criminally or civilly liable for their participation. However, these are not reasons to deny them the right to refuse to fight in such wars. On the contrary, the best interpretation of refugee and human rights law would affirm such a right.

Third, recognizing that the core victims of the crime of aggression are individuals, rather than states, sheds light on how we ought to conceive of victim judicial participation and reparations in ICC aggression prosecutions. The crime of aggression is the core element of international criminal law that protects combatants’ and collateral civilians’ right to life. Unlike recent jus ad bellum reparations regimes, which have excluded combatant deaths from the wrong warranting remedy, the ICC regime of reparations for aggression must reflect the normative centrality of precisely those personal violations.

This Article provides the normative foundation for those doctrinal and operational implications. Part I explains what it is to offer a normative account of a law and why it is appropriate here. Part II identifies the dominant normative account of aggression, as understood by both advocates and critics of its criminalization. On this view, the core wrong of aggressive war is a wrong against the victim state and the political collective that it represents. Part III debunks that account, arguing that five aspects of aggression and its legal context show that its criminalization is primarily about wrongful killing: (1) sovereignty is indeterminate as a normative guide to the jus ad bellum and it cannot explain why aggression is criminal, but more severe sovereignty violations are not; (2) its infliction of unjustified human suffering can explain why illegal war is the only criminal sovereignty violation; (3) understanding aggression in this way reconciles it to the broader international criminal law project; (4) the claimed motivation for strengthening the jus ad bellum in the early- to mid-twentieth century was focused on human suffering, not states’ rights or sovereignty; and (5) the jurisprudence of Nuremberg and Tokyo affirms the place of unjustified killing at aggression’s normative core. Part IV addresses potential problems for this account posed by the legal status of bloodless aggression and humanitarian intervention. Part V introduces the key legal consequences of adopting this normative account.

i. the normative underpinnings of a crime

Stated most abstractly, a “normative account” aims to elaborate a notion of wrongfulness that would make sense of and underpin the law’s posture on a given issue. The premise of engaging in work of this kind is that laws do not merely serve to coordinate or to set incentives; they also instantiate what the community in question takes to be important values. The balance between the coordinative and moral expressive functions varies across domains, but criminal law in particular takes a stand on what is right and wrong and on who has acted culpably. It does not provide priced permissions; it prohibits and it condemns.16

This is nowhere clearer than in international criminal law. We cannot make sense of the criminalization of obedient participation in crimes against humanity or genocide as a way of coordinating behavior.17 Confronting individuals with contradictory obligations under domestic and international law may even muddy coordination and confuse expectations. The criminalization of such obedient participation is instead a moral expression of the wrongfulness of such participation and the culpability of those who engage in it.18

The criminalization of aggression, like that of other behaviors, is an expression of fundamental moral values. At the core of the contemporary jus ad bellum is the public moral claim that certain kinds of war are fundamentally wrongful.19 Signatories to the Kellogg-Briand Pact of 1928—the first treaty to ban war explicitly—not only committed to “renounce” the recourse to war, but also to “condemn” it.20 This followed the unabashedly moralized framing of the issue by the Pact’s intellectual forefathers and was accompanied by claims that war had finally been recognized as a global public wrong, rather than a state of bilateral dispute.21

The asserted connection between the criminalization of aggression and underlying moral values was even more explicit at Nuremberg. In his opening statement before the International Military Tribunal at Nuremberg (IMT), chief prosecutor Justice Robert Jackson stressed, “When I say that we do not ask for convictions unless we prove crime, I do not mean mere technical or incidental transgression of international conventions. We charge guilt on planned and intended conduct that involves moral as well as legal wrong.”22 He labeled the pre-Kellogg-Briand legal order “contrary to ethical principles” because it denied the maxim “that there are unjust wars and that unjust wars are illegal.”23

The Tribunal itself went to some lengths to establish the positivist credentials for outlawing aggressive war.24 However, it combined these efforts with deeply moralized claims. The IMT held that, even in the absence of an explicit criminal law prohibition, the leader of an aggressive war “must know that he is doing wrong” and insisted that “it would be unjust if his wrong were allowed to go unpunished.”25 It emphasized in this respect such wars’ wrongfulness in “the conscience of the world,” terming aggression “essentially an evil thing” and the “supreme international crime.”26 The U.S.-administered Nuremberg Military Tribunals (NMT) took a similar line.27

A normative account seeks to make sense of such a legal posture. What is it that makes aggression a “high crime” rather than a “mere tort” in international law?28 What moral judgment underpins that status? To hold that such questions have answers is not to hold that any particular individual must accept the moral posture underpinning the substance of the law in all cases; it is merely to recognize the law’s internal claim to normativity.29

Thus, to give a normative account of a law is to endeavor to inhabit the internal legal point of view.30 This means starting not from first principles, but from an analysis of the current law. However, it also means seeking to account for that law’s underlying moral claim, rather than seeking merely to describe or interpret the rule. The premise of the argument that follows is that a normative account is superior to its alternatives in achieving that objective to the extent that it better satisfies three core criteria. First, it must offer an explanation for what the regime unambiguously permits and prohibits concerning the issue at hand. Second, it ought to reflect the law’s core purposes, as evinced in the law’s structure and in the articulations of its framers. Third, it should cohere with connected or related laws in domains adjacent to that which it explains. In forthcoming work, I argue that if multiple accounts pass the first three tests, the superior remaining account is that which is most morally plausible.31 However, that more controversial claim is unnecessary here, because the question that motivates this Article can be answered comprehensively with reference to the three primary criteria.

The first is a straightforward and inflexible threshold criterion. The essential feature of any normative account is an explanation of the morality of a given legal rule. It must explain what the law is, not what the law should be. The second and third criteria are more flexible. They do much of the work in distinguishing between better and worse accounts, but they are not minimum essential criteria. At their core is the basic observation that the law’s credibility as a normative system depends on the schedule of imperatives it issues cohering with the public purposes those imperatives serve in a structure of mutual support, rather than collapsing into discord or internal contradiction.32 When the normative messages of different laws conflict, this undermines both the law’s credibility in condemning violations and its authority in demanding action. Given the fragmented nature of international law’s enforcement and the rarity of credible coercive backing, this internal-coherence imperative is particularly acute at the global level.33

However, although coherence among laws and between legal rules and their stated objectives is an important trait of a good normative account of the law, it cannot be an essential criterion. Law is created and revised by the cumulative efforts of different agents, acting at different times with different objectives. It is the product of compromise and sometimes of deliberate efforts by competing lawmakers to create internal contradictions. As a result, it may be impossible in any given case to provide a normative account that both explains the moral core of a particular law andfits with the regime’s purpose or with adjacent rules. In that scenario, the best normative account of the law in question will be dissonant with the regime as a whole. When no coherent alternative is viable, work on normative underpinnings can be valuable in helping to identify such dissonance.

The core claim of this Article, substantiated primarily in Part III, is that judged against these three criteria, a normative account of the crime of aggression that locates the wrong in the constituent killings is superior to the traditionally dominant alternative. To set the stage for that argument, consider first the nature and the appeal of the latter view.

ii. the orthodox normative account of the crime of aggression

On the dominant normative account, the crime of aggression is a macro wrong against a foreign state (sometimes understood to draw its moral value from the self-determining collective it represents), not a compound of minor wrongs against a population of individual human persons. This sovereignty-focused understanding of aggression has long been the view of both critics and defenders of the criminalization of illegal war.

A. The Walzerian Account: The State as Victim and Perpetrator in the Law

The legal starting point for the orthodox account is clear. The judgments of the IMT and the International Military Tribunal for the Far East (IMFTE) described the crime of aggression as that of waging “aggressive wars” against other “countries” or “nations.”34 In addition, both tribunals arranged their findings regarding aggression into individual criminal wars, identified and separated by victim state.35 Similarly, Control Council Law No. 10, which governed the NMT, provided that the crime involved the “initiation of invasions of other countries,”36 and the ICC amendment follows the General Assembly’s 1974 articulation in defining an aggressive war as one that violates the “sovereignty, territorial integrity or political independence of another State” or otherwise runs contrary to the U.N. Charter.37

It is against this background that Michael Walzer describes international law on this issue as rooted morally in a “domestic analogy,” in which aggression is an infringement upon the attacked state akin to the infringement of burglary on a human person.38 Under this “legalist paradigm,” states “possess rights more or less as individuals do.”39 After all, he contends, it is“the state that claims against all other states the twin rights of territorial integrity and political sovereignty”—the rights that the crime of aggression seeks to protect.40 Understood in this way, aggression is a morally unique international crime. Whereas genocide, war crimes, and crimes against humanity are all rooted in wrongs against human beings, the normative heart of aggression is a wrong against states.41

Seeking to make moral sense of this idea,42 Walzer argues that it captures the genuine wrong an aggressive war inflicts on the attacked political collective—a collective that is defined imperfectly, but, all things considered, optimally, by state borders. For him, “[t]he state is constituted by the union of people and government,”43 and the jus ad bellum’s relationship to individual human beings is in its protection of their “communal rights” of self-determination.44 “Break into the [state] enclosures,” Walzer explains, “and you destroy the communities. And that destruction is a loss to the individual members . . . .”45 In other words, if individuals are wronged when their state is invaded illegally, that wrong is suffered only indirectly, through the wrong to the political collective of which they are part.

Emphasizing the priority status of the collective here, Walzer insists that the wrong of aggression obtains irrespective of the responsiveness of the community to its members. It obtains whether or not the state is democratic or respectful of human rights and the rights of minorities. Only “when a government turns savagely upon its own people” does Walzer think the morality of states that underpins the crime loses its force because the state ceases to be a viable forum for collective self-determination.46 Significantly, however, despite arguing that the morality of states’ rights finds its limit in this extreme context, Walzer is clear that his philosophical position on this point is a departure from the moral underpinnings of the law.47 In other words, for Walzer, the legal crime of aggression is rooted fundamentally in a moral wrong between states. On extraordinary occasion, that statist morality may be misguided, but for the most part, Walzer not only identifies it as the moral posture underpinning international law on aggression, but also defends it as morally appropriate.48 For him, the crime of aggression captures a real wrong—a wrong against the state, the moral value of which is derived from its function as a site of collective self-determination.

This assessment has been endorsed repeatedly. For example, Paul Kahn contends that the criminalization of aggression at Nuremberg represented and initiated a “new legal regime founded on protecting state sovereignty through the prohibition on the use of force.”49Kahn sees this prohibition as an effort to ban and condemn violations of “positive sovereignty, understood as the self-formation of a people” in a state whose boundaries allow that people to function as “a single, collective actor.”50 Similarly, Christopher Kutz argues that aggression’s wrong is in denying the target state’s people the chance to make “their politics on their own”—a denial that Kutz argues wrongs nondemocratic peoples no less than democratic ones.51 Unlike Walzer and Kahn, Kutz does not link this moral theory to the law. However, by locating the moral violation at the heart of aggression in its negation of collective autonomy through its infringement of states’ rights, he buttresses Walzer’s dominant account of the legal regime.52

Each of these theorists identifies the political collective as the sovereign that is wronged by aggression. The core point for each is that aggression is a wrong against that collective that cannot itself be “reduced,” in David Rodin’s words, to an aggregation of harms against individuals.53

Larry May arrives at a similar conclusion without relying on self-determination. He argues that “aggression is morally wrong because it destabilizes States that generally protect human rights more than they curtail them.”54 For May, then, the state’s value lies in its service to human beings, rather than in its approximation of the political collective. Nonetheless, on his theory, the wrong of aggression remains a wrong against the state. This, he argues, has implications not just for how we understand the victim, but also how we define the perpetrator. Precisely because the wrong at the core of the crime of aggression occurs on the macro level, “the acts of individuals that make up war are conceptually and normatively distinct from the State aggression.”55 This, he contends, raises a question about how we can hold even high-ranking individuals “criminally liable for invading Poland,” since the core interstate wrong is simply not something that they can commit.56

Of course, crimes against humanity and genocide have collective elements of their own: the former involves a widespread and systematic attack; the latter, an intent to destroy a group. But what distinguishes aggression for May is that the micro contributions of the participants in an aggressive war “are not themselves criminal,”57and, indeed, are “not themselves wrong independent of what is going on at the State level.”58 Accordingly, May contends that rooting the wrong of aggression in the component killings and destruction would provide “no relevant moral distinction between aggressive wars and defensive wars.”59 The result, May explains, is that aggression is not a wrong that individuals (“even lots of them”) can commit, unless they control the state itself.60 High-ranking officials can be morally connected to that interstate wrong only if they combine the right level of control over the state with the intention to shape state action.61

Just as Walzer’s, Kutz’s, and Kahn’s accounts of the state as victim of aggression might be thought to explain morally the fact that the crime of aggression involves an attack on the state, May’s exploration of what it means to perpetrate the macro wrong of aggression could be argued to illuminate another important aspect of the legal framework—namely, that criminal liability turns on acting through the state.62 At Nuremberg, that an individual made a substantial contribution to an aggressive war or organized massive lethal attacks on enemy troops was not enough to make that individual criminally liable.63 The same is true for an individual who had been aware of the criminal nature of the war and who acted under no duress.64 Thus, even senior members of the Nazi High Command were acquitted of crimes against peace.65 Only members of Hitler’s “inner circle,” who were able to “shape or influence” state policy, were found guilty.66 Similarly, under the ICC amendment, criminal liability for aggression is limited to persons with the capacity to “exercise control over or to direct the political or military action of a State.”67

From this perspective, the macro nature of the crime is two-dimensional. The moral wrong at the core of aggression is a wrong inflicted on a victim state (or at least the collective it represents) by individuals acting through a perpetrator state.68 Importantly, for Walzer and others adopting his view, this is not just a moral account of the legal posture on aggression; it is also a moral defense of that posture. From that perspective, in an aggressive war, a genuine wrong occurs at the interstate level and that wrong is worthy of criminal punishment and official condemnation.

B. Criticizing the Criminalization of Aggression on Walzerian Terms

Critics of the criminalization of aggression agree with the Walzerian normative account of the law. They agree, in other words, that the crime of aggression is rooted in a claimed moral wrong “against a state” as opposed to “violations against individuals.”69 The difference is that, for the critics, this account of the crime spotlights precisely why it should not be a crime at all.

David Luban, for example, accepts that “the crime of aggressive war is a crime of state against state.”70 However, for him, this is where the Nuremberg endeavor took a wrong normative turn. Luban insists that “the achievements at which the trial was aiming were compromised, rendered equivocal, by the trial itself,” with the framers “unwilling to question . . . the political system of nation-states.”71 The problem, he argues, was precisely the Walzerian effort to “confer moral rights” on the state.72 By developing a crime that depends on such moral rights, Nuremberg “erected a wall around state sovereignty.”73 Whereas crimes against humanity, together with the removal of official immunity and the imposition of liability for obeying orders that were legal according to the sovereign state, seemed to “perforate[] or even destroy[]” the classical doctrine of sovereignty “in the name of ‘humanity,’” Article 6(a) of the IMT Statute “fortified it by making aggressive war—a violation of sovereignty—an international crime.”74 The result, for Luban, is that Nuremberg’s “central achievement” of recognizing crimes against humanity was undermined, and that the trials left “a legacy that is at best equivocal and at worst immoral,” with the criminalization of aggression “a major moral enemy of the human rights movement.”75

Reacting to ICC’s pending amendment, Erin Creegan has put forth an updated version of this sentiment. She argues, “[A]ggression . . . is a crime committed by the leadership of one state, directed against the abstract interests of another state . . . . The harm of aggression is the insult to a state’s territorial independence and sovereignty . . . . [P]reventing wartime suffering is not the direct object of the crime of aggression . . . .”76

Like Luban, Creegan finds this to be an unacceptable normative core. She contends that, “[w]ithout adversely affected human victims, it is hard to put a crime like aggression in a category similar to war crimes or crimes against humanity or genocide. And it does not seem to belong next to them; it almost demeans them.”77 She goes on to describe human beings’ rights not to be subject to the wrongs of genocide, war crimes, and crimes against humanity as often “infinitely more” powerful than the right of states not to be the victims of aggression.78

The dispute between Walzer and his critics on this issue is complex. From the internal legal point of view, they are in close agreement as to the basic moral underpinnings of the crime. For both sides, the crime is premised on the idea that aggression is a wrong against states (or the political collectives they represent).79 Morally, however, they disagree about whether violating a state in that way is a wrong worthy of criminalization. Their agreement from the internal point of view is at the heart of their disagreement from the external point of view.

As discussed in Part V, if one accepts the widespread conception of aggression as an exclusively macro-level wrong, that understanding must underpin the “object and purpose” of the criminalization of aggression, with all of the entailed consequences for the proper interpretation of the crime and of connected rights and duties.80 Those implications would hold whether or not one endorses the law as morally well directed.

But the shared premise is mistaken. The next Part reframes how we ought to understand aggression from the internal legal point of view. The normative core of the crime of aggression is not a morality of states’ or political collectives’ rights, but a morality that condemns the unjustified killing of human persons.

iii. why aggressive war is an international crime

To be clear, the position advanced here is not that sovereignty is irrelevant to the crime of aggression. Under current international law, whether a war is criminal is determined in part by whether it violates the “sovereignty, territorial integrity or political independence of another State” or is otherwise “inconsistent with the Charter of the United Nations.”81 In other words, which wars are wrongful is a macro-level question that depends typically, although not exclusively, on which side has violated the other’s sovereignty.

But that interstate breach is not why waging such wars is criminal. Waging war in breach of those interstate rules is criminal because it entails widespread killing and the infliction of human suffering without justification. Seen in this way, aggression is a modified form of crime against humanity, perpetrated ordinarily through a violation of sovereignty.82

To understand why this must be the case, consider the point from five perspectives. First, the concept of sovereign rights is indeterminate as a normative guide on the issue of aggression. The criminalization of such wars is as great a restraint on state sovereignty as it is a protection of it. Moreover, interstate violations that more effectively and dramatically infringe core sovereign rights than does aggressive war are not criminal. Second, what distinguishes aggression from any other sovereignty violation is that it involves killing without the justification of responding to human violence or its immediate threat. Other forms of killing without justification are criminal in another form. By protecting combatants’ and collateral civilians’ right to life, the criminalization of aggression fills a crucial gap in that broader criminal-law approach to unjustified killing. Third, understanding aggression in this way reconciles its criminalization with the so-called “humanization” of international law—the rise of the human being as a normative focal point in legal interpretation and doctrinal development. This phenomenon is manifest especially clearly in international criminal law. Fourth, the claimed imperative to incorporate a restrictive jus ad bellum into twentieth-century international law was articulated not in terms of sovereignty, but in terms of human suffering. Finally, wrongful killing was normatively central to the reasoning of the judges and prosecutors at Nuremberg and Tokyo on the crime of aggression. Ultimately, states’ rights are important in structuring when the use of force is permitted, but aggression is a crime about the infliction of human death and suffering without justification.

Recognizing the criminal wrong of aggression to be the infliction of unjustified human violence suggests that legal interpretation and thinking in this domain might be better illuminated by a more recent, revisionist tranche of just war theory than has thus far been recognized. A growing cohort of moral theorists has countered the longstanding Walzerian orthodoxy that the jus ad bellum can be restricted morally to the macro-level, with no implications for the rights or wrongs of the constituent actions.83 These revisionist theorists, led by Jeff McMahan, have deliberately avoided engagement with international law, sometimes criticizing Walzer for what they consider to be a conflation of the law and the morality of war.84 One of the upshots of the arguments in this Part is that these philosophical critics may have more to say about the internal normativity of the existing legal regime than they or international lawyers have understood.

A. Sovereignty Cannot Explain the Jus Ad Bellum or the Criminalization of Aggression

When it comes to war, sovereignty cuts both ways. Although banning aggressive war has protected certain sovereign rights, it has limited the sovereign’s capacity to assert and protect many of its other rights. Indeed, despite his framing of Nuremberg and the U.N. Charter as moves toward protecting sovereignty,85 Kahn recognizes that on a maximal vision of sovereignty in international affairs, “there is no difference between protection and assertion: To protect the state is to assert its power to defend its ‘vital interests.’”86

Prior to Nuremberg, and certainly prior to Kellogg-Briand, international law reflected precisely that principle: states had the power and authority to use violence to punish or seek to end nonviolent infringements of their legal rights.87 They could wage war to settle a dispute, reverse a wrongful seizure, or otherwise retake what was rightfully theirs.88 Capturing this core doctrine, Vattel wrote simply, “Whatever strikes at [a sovereign state’s] rights is an injury, and a just cause of war.”89 The use of force was the key tool of law enforcement and sovereignty vindication in an international regime focused on interstate relations and state rights.90 Moreover, in the absence of a system of global institutions by which the “right” in any dispute could be verified, states also had the sovereign authority to determine whether such vindication was called for.91

In one sense, this system rendered states vulnerable to armed attack. But this was not because states’ rights were unimportant. Rather, the animating premise was that it was inconceivable that the state’s authority to determine whether force was necessary to protect its legal rights could be abrogated.92 The sovereign stood above international law.93 From that position of priority, what was truly defensive of its rights was necessarily something “that a state had to judge for itself.”94 Understood in this way, sovereignty “resists [the] universalization” upon which any genuinely restrictive jus ad bellum depends.95

Recognizing this assertive dimension of sovereignty is crucial. Precisely because it lacked any meaningful jus ad bellum restraint, the pre-Nuremberg, pre-Kellogg-Briand era was the high-water mark of sovereignty in international law. Thus, rather than asserting that banning war would erect “a wall around sovereignty,”96 Salmon Levinson—a vanguard advocate for what ultimately became the Kellogg-Briand Pact—felt compelled to rebut the allegation that allowing this “check upon [the state’s] original unlimited power” would unjustifiably “invade” and “impair” core sovereign rights.97 Accepting that adopting the prohibition would limit sovereignty, Levinson insisted that the retention of the sovereign right to use force at its own discretion was morally indefensible.98 Along similar lines, Frank Kellogg acknowledged in a public address after the signing of the Pact that the key obstacle facing the drafters had been the longstanding notion that waging war was simply “a nation’s legal right.”99

Of course, although the criminalization of aggression and the prohibition against the use of force have limited sovereignty in this way, the new regime has also enhanced states’ legal protection against armed attack.100 However, that newly protected dimension is just one elementof sovereignty. Precisely because of the elevation of that element, all other aspects of sovereignty lost the unilateral vindication mechanism upon which they had depended.101 The assertion and protection of those other rights have been taken away from the sovereign and transferred to the global collective.102

The obvious defense of the traditional account here would be to argue that the ban on aggression prohibited and criminalized the most severe violation of state sovereignty in exchange for eliminating lesser sovereign rights. Had that been the exchange, a sovereignty-focused account could explain why aggression, alone among sovereignty violations, is an international crime and a violation of jus cogens.103The explanation would be that aggression has this special status because it is the interstate violation most profoundly detrimental to the essential elements of sovereignty.

This, however, is not the case. Put to one side all of the sovereign rights given up in the ban on the use of force and consider in isolation the sovereign rights at the crux of today’s jus ad bellum: political independence and territorial integrity.104 On these issues alone, aggression is not uniquely harmful. A criminal use of force can be far more modest in its diminution, or even intended diminution, of these rights than would be nonbelligerent, noncriminal infringements of the same rights.

Consider, for example, the nonbelligerent installation of a puppet regime in a foreign state through the manipulation of its elections.105 Puppet regimes create principal-agent problems for the intervening state, and democratic manipulation will not always work flawlessly. Such illegal actions, however, would infringe on political independence and self-determination profoundly, and often to a greater extent than is even intended by an illegal military attack.106 And yet such nonviolent interventions are clearly not internationally criminal, since they lack “the use of armed force.”107

The same applies to nonviolent violations of territorial integrity. Failing to hand over territory to a lawfully seceding entity, holding another state’s territory following a misguided handover from a departing former colonial power, or illegally manipulating an independence referendum in a way that triggers an effective secession are all severe infringements of another state’s territorial integrity.108 However, without an armed attack, none would be internationally criminal.109

By contrast, examples of criminal aggression listed in the ICC amendment, such as aerial bombardment or an attack on a foreign state’s naval fleet, may do very little to undermine the victim state’s self-determination, territorial integrity, or political independence.110 The same is true of drone strike campaigns against nonstate actors in foreign territory, which are perfectly compatible with leaving intact the political independence and territorial integrity of the host state.111 The point is not just that such belligerent acts may prove unsuccessful in taking territory or overthrowing the government.112 More fundamentally, it is that criminal aggression need not even pursue the objective of undermining significantly those core elements of sovereignty.113

The fact that belligerent actions with minimal impact or intended impact on sovereignty are criminal, while nonviolent infringements of political independence and territorial integrity are not, cannot be explained with reference to the degree to which each action violates those core sovereignty rights. If that were the standard, the outcomes would be reversed. The crime as currently constituted must be explained by something else.

B. The Human Core of Aggression

Far more robust as a normative explanation of the crime than the fact or degree of aggression’s violation of sovereignty or states’ rights is the form of that violation, the means by which it is perpetrated.114 What is unique about illegal war among violations of states’ rights—what makes it criminal, when no other sovereignty violation is—is the fact that it entails the slaughter of human life, the infliction of human suffering, and the erosion of human security. To qualify as aggression, those harms must occur ordinarily (although not exclusively) in an unjustified attack in which one state infringes the core sovereign rights of another. However, it is the unjustified killing and infliction of human suffering, and not the violation of sovereignty, that are the wrongs at the heart of aggression. Consider three ways in which this is so.

First, even within the category of violent interstate interactions, the reason that a particular aggressive attack is unjustified is not that it infringes the target state’s territorial integrity or political independence. Assuming no Security Council authorization, what determines that one side of an international armed conflict is in violation of the jus ad bellum is that it does not respond to an armed attack.115 Therefore, the use of force seeking to remedy or prevent a severe but nonviolent infringement of political independence or self-determination, such as the manipulation of its elections through hacking, would be illegal and could qualify as an act of criminal aggression.116 Similarly, according to the Eritrea-Ethiopia Claims Commission, it is a jus ad bellum violation, and thus at least potentially criminal, for a state to use force to recover its own territory, if that retaking does not respond to an armed attack.117 Indeed, even when one state controls another’s territory following an armed attack by the former, if the latter’s forcible response is sufficiently delayed, it too would be illegal and potentially criminal.118

On the other hand, a defensive use of force that has a significant and intended impact on the aggressor’s internal structures of government or its territory would be lawful as long as that force were necessary and proportionate to stopping the armed attack. The paradigmatically lawful wars against Germany and Japan in World War II each led to long-term occupations and regime changes, not to mention Germany’s territorial and political fragmentation for half a century.

Moreover, using force in response to a foreign armed attack is lawful, even if that attack does not threaten the victim’s territorial integrity or political independence.119 Indeed, according to the Ethiopia-Eritrea ruling, such defensive uses of force are lawful even if the objective is merely to defend the attacked state’s personnel while they are exercising peaceful control over the aggressor’s own sovereign territory.120

What unifies these cases is that a lawful use of force by a state must respond to an international attack on, or threat to, the lives of its human subjects; that is what it means for an attack to be armed.121 Of course, territorial integrity and political independence are often protected by defensive war and violated by aggressive war. But the legal realities discussed above indicate the overriding importance of human life and physical integrity. Severe violations of territorial integrity or political independence are not internationally criminal, and do not trigger a right to use remedial force, unless they also involve a severe threat of violence to human beings. At the same time, uses of force that do little or nothing to protect core sovereign rights are lawful as long as they respond to an armed attack. Conversely, it is criminally aggressive to inflict violence on human beings in a foreign state even if there is no intention or significant prospect of harm to territorial integrity or political independence as a result of the attack. Indeed, this holds even if that infliction of violence aims merely to recover the aggressor’s own sovereign territory.

There are, of course, marginal jus ad bellum cases in which human life seems to play a less central role and core sovereignty rights appear to come to the fore. The most notable such case is that of the so-called “bloodless invasion.”I address this case at greater length in Part IV, below. However, by way of brief preview, whether bloodless invasions are criminal at all is at least debatable. Moreover, even if they are criminal, the key to understanding their criminality is not the fact that such invasions infringe sovereignty, but rather that they do so by imposing the immediate threat of lethal violence against anyone who might resist.

The second reason in favor of the unjustified killing account of aggression is that it fits with the broader legal posture on violence to human beings. Whereas aggression is uniqueamong sovereignty violations in its criminal status, it is decidedly notunique in that respect among forms of large-scale killing and human harm that do not respond defensively to the threat or infliction of such killing and harm. On the contrary, such nondefensive killing is generally criminal in one form or another.

Perhaps the most important element of sovereignty is the state’s monopoly on the legitimate use of internal force. However, that authority is not unlimited. When state agents inflict widespread killing or human harm internally, without responding to some form of attack or imminent internal threat, they commit a crime against humanity.122 State militaries may use force against insurgents in a noninternational armed conflict, but until such insurgents have formed (and thus have begun to pose an internal belligerent threat) the state cannot initiatesuch an attack. To do so would be to engage in a widespread and systematic attack on a civilian population.123 If nonstate actors coalesce into a military response to such an attack by the state, then, from that point on, such actors would become combatants and state forces may target them without committing a crime.124 But that transformation would not change the criminality of state actors’ initial attack on human life.

Similarly, when nonstate actors inflict widespread death or human harm against civilians, they too commit crimes against humanity.125 When they inflict harms of this kind on state armed forces or other combatants, they commit murder and other universally applicable domestic crimes without the international law privileges of belligerency.126 Although the latter harms are not internationally criminal, they are domestically criminal in all states, and international law offers no immunity to those who perpetrate such crimes on behalf of nonstate actors.

On a smaller scale, other forms of killing not justified on defensive grounds are war crimes.127 Classic examples are killing civilians or prisoners of war.128 And, outside the context of armed conflict, killings that do not protect persons or the broader community are criminalized domestically as murder; international human rights law demands as much.129

Thus, whether perpetrated by the state, a nonstate group, or an individual, killing cannot be used to change even an illegal or unjust status quo unless it responds to human violence or the threat thereof. Of course, the point at which such violence or threat of violence becomes criminal varies depending on the actors and contexts involved. There is a presumption in favor of the legitimacy of internal uses (and especially threats) of force by the state and a strong presumption against the use or threat of force by nonstate actors. Similarly, the context of war changes the flexibility of what constitutes defensive action and the attribution of responsibility for unjustified killing.

In that sense, the differentiation across these crimes is important. Indeed, it helps to make sense of the interstate element of aggression (which typically involves a sovereignty violation). Rather than being the core criminal wrong, as the traditional account would have it, that interstate infringement identifies a particular form of unjustified massive killing and enables the thresholds of criminality to reflect the presumptions of legitimacy and contextual considerations appropriate to that form. The state has a lower presumption of legitimacy using (or threatening) force internationally than it does domestically, and a higher presumption of legitimacy than do nonstate actors. But this differentiation among crimes ought not obscure the common trait that makes them all crimes. The nondefensive use of lethal force by any actor is generally criminally prohibited, even if its purpose is to remedy an unjust status quo. Aggression extends that fundamental principle to the interstate context.130

Viewing aggression in this way spotlights why it is its own crime. The other provisions described above prohibit rebels’ killing of civilians and combatants; they prohibit a state’s forces from killing civilians or groups yet to form a lethal rebellious threat; and they prohibit nondefensive killing outside of armed conflict. What no provision other than the crime of aggression prohibits is a state’s forces’ unjustified killing of a foreign states’ combatants and collateral civilians. In the absence of a crime of aggression, the unjustified infliction of death and suffering in such interstate contexts would have been the normative anomaly in which such harms could be inflicted without any prospect of criminal liability. Aggression fills a crucial gap, providing otherwise missing criminal law protection to the right to life of combatants and collateral civilians.131

The third aspect of the unjustified killing account’s explanatory superiority is that at least one type of criminally aggressive war involves the infliction of human violence, but noviolation of sovereign rights. Article 8 bis includes as a category of criminal war alternative to uses of armed force “against the sovereignty, territorial integrity or political independence of another State,” the “use of armed force by a State . . . in any other manner inconsistent with the Charter of the United Nations.”132 As in the U.N. Charter, the more specific categories serve as points of emphasis.133 However, for the broader category of armed force used “in any other manner inconsistent with the Charter” to make sense, it must include wrongful wars that do not infringe another state’s “sovereignty, territorial integrity or political independence.”

One obvious example of a use of force fitting that description would be a war waged by a state in its own territory against a U.N.-authorized force deployed to prevent the host state from engaging in atrocities. A recent example was the use of force by Libya in its own territory against the U.N.-authorized North American Treaty Organization (NATO) coalition in 2011.134

In that particular war, NATO engaged only aerially and suffered little damage and zero casualties, so the Libyan action against NATO would almost certainly not surpass the character, gravity, and scale threshold requirements of Article 8 bis (1).135 However, for the reasons outlined below, the little force that Libya did muster was plainly “inconsistent with” the U.N. Charter, as required by 8 bis (2). Counterfactually, had Libya caused sufficient NATO casualties to exceed the 8 bis (1) threshold, its action in doing so would have been criminal aggression. Because the conflict took place exclusively within universally recognized Libyan territory and the subject of dispute was Libyan domestic policy, this is explicable only in light of the lack of justification for the violence against coalition troops, not in light of any infringement of sovereignty.

The inconsistency of such action with the U.N. Charter is established by three elements. First, the Charter prohibits states from using force in their “international relations”—including on their own territory—in any way contrary to the Charter’s “purposes.”136 Interactions between a host state like Libya and a U.N.-authorized force clearly fall in the category of international relations, and the furtherance of the Security Council’s work under Chapter VII is clearly a U.N. purpose.137 Second, the only exception to the Charter’s prohibition on the use of force (other than acting pursuant to a Security Council authorization) is self-defense. Under Article 51 of the Charter, however, self-defense applies only “until the Security Council has taken the measures necessary to maintain international peace and security” and self-defensive measures “shall not in any way affect the authority and responsibility of the Security Council” to act under Chapter VII.138 Third, the authorization of forces to prevent domestic atrocity is now widely recognized as part of the Security Council’s authority, and indeed responsibility, under Chapter VII, as part of its activity in maintaining international peace and security.139 As such, it overrides any affected state’s right of self-defense.

The upshot is clear. When states fight back against U.N.-authorized forces on their territories, their actions meet the core requirement of Article 8 bis (2). As long as the action is of sufficient character, gravity, and scale, it is a criminal aggression.140

Thus understood, “aggression” is a term of art; its meaning is defined not by which state struck first, but by which engaged in an illegal and grave use of force.141 As the Institut de Droit International resolved in 1971, “the party opposing the United Nations Forces has committed aggression,” irrespective of whether it acted first or crossed a border to do so.142

This definition bears on the question at hand. Reference to core states’ rights cannot explain the criminality of waging war against a U.N.-authorized humanitarian intervention force operating exclusively in the attacking state’s own territory.143 But the criminality of killing U.N. personnel in order to defend an ongoing atrocity can be explained with reference to killing and inflicting human suffering without justification.

Of course, humanitarian intervention without Security Council authorization is generally thought to be illegal. The notion that fighting against a humanitarian intervention within one’s own borders could itself be criminal therefore applies most plausibly only when the intervention has Security Council backing.144 Nonetheless, even if limited to that scenario, this argument further supports an account of aggression as a crime rooted in wrongful killing, not the severe violation of states’ rights. The special case of unauthorized humanitarian interventions is addressed in Parts IV and V.

For each of the reasons discussed in this Section, the nondefensive killing in an illegal war is why waging such a war is criminal. This explanation best accounts for why the state’s right to protect itself from armed attack survived Nuremberg and the U.N. Charter, whereas its right to use force to vindicate eachof its other legal rights was discarded comprehensively. It is consistent with the criminalization of unjustified killing more generally. And it is what clarifies why using force against an invading U.N.-authorized humanitarian intervention would itself involve waging a criminally aggressive war.

C. The Humanization of International Law

An additional virtue of the unjustified killing account is that it reconciles the criminalization of aggression to the broader normative context. In recent decades, the very foundations of international law have shifted away from a regime rooted exclusively in state sovereignty and toward a regime that privileges human rights and human values—a transformation often termed the “humanization” of international law.145 This phenomenon is most explicit in the contemporary notion that sovereignty is at some level conditional on the state discharging its “responsibility to protect” the basic human rights of those within its control.146 The account presented here identifies the initial criminalization of aggression as an early step in this trajectory and makes sense of its ongoing customary criminality and its incorporation into the Rome Statute in the more deeply humanized contemporary international legal context.

The humanization process has been especially prominent and consequential in international criminal law. At the vanguard of that regime’s revival, the International Criminal Tribunal for the former Yugoslavia (ICTY) wasted no time in framing its interpretive approach in precisely such terms, reasoning that “[a] State-sovereignty-oriented approach has been gradually supplanted by a human-being-oriented approach. Gradually the maxim of Roman law hominum causa omne jus constitutum est (all law is created for the benefit of human beings) has gained a firm foothold in the international community as well.”147

This notion is manifest throughout the framework of international criminal law. Crimes against humanity, war crimes, and genocide focus on the most severe harms to human beings.148 The Office of the Prosecutor at the ICC selects cases with a view to surpassing the gravity threshold for admissibility based on the severity of the human harms they involve.149 And, in perhaps the most obvious privileging of humanity over sovereignty, international criminal law requires subjects to disobey domestically authoritative sovereign commands in order to protect other human beings.150

The traditional account suggests that the criminalization of aggression is in tension with, or even opposition to, this “humanization,” and thus with international criminal law itself.151 The account presented here, by contrast, locates aggression at the heart of that process.152 Criminalizing aggression constrains the sovereign’s previously unlimited authority to vindicate its rights with force, preserving that authority only when necessary to protect human beings from the illegal infliction of violence or to respond to the illegal threat of such human violence.153 Seen in this light, the classification of aggression (and no other sovereignty violation) alongside genocide, crimes against humanity, and war crimes makes sense.

A legal regime’s core claim to normativity is strengthened when it coheres internally.154 The traditional account renders international criminal law “equivocal” and self-undermining—simultaneously a major step forward in human rights and “a major moral enemy of the human rights movement.”155 This tension weighs heavily against it. Conversely, on the unjustified killing account, aggression is of a normative piece with the rest of international criminal law, and is an important element of the human rights movement. This militates powerfully in its favor. Therefore, even if, contrary to the arguments in Sections III.A and III.B, the traditional account and the unjustified killing account offered equally plausible explanations of the crime, the latter would still be the better normative account.

D. Revisiting the History of the Move to a Restrictive

This focus on the human being is not merely a matter of internal normative logic. It has historical plausibility as part of the regime’s purpose. Following the devastation of World War I, human life was the overriding public concern of those who led the intellectual and political movement that ultimately produced the Kellogg-Briand Pact—the key legal hook for the aggression convictions at Nuremberg.156

In his 1921 pamphlet laying the foundations for this transition, Levinson described war as “inhuman” and compared legal toleration of the practice to the toleration of dueling, which had since become “plain murder under our laws.”157 Levinson recognized that the ban would require states to forswear the sovereign authority to vindicate their rights through the “legal device of violence.” 158 Rather than bolstering this argument with a claim that sovereignty would also be augmented by the ban on interstate armed attacks, he insisted that the longstanding maximalist vision of sovereign authority was no longer tenable, given its infliction of “the worst form of violence and crime existing among men.”159 Colonel Raymond Robins, a fellow proponent of the outlawry of war, wrote of the movement as an effort to “declare[] [war] in international law to be what it is in fact, the supreme enemy of the human race.”160

Leading political figures also adopted this framing. President Harding spoke of the move toward a restrictive jus ad bellum as the call “of humanity crying for relief.”161 Former British Prime Minister David Lloyd George reflected in his memoirs on the growing sense in this period that war was a “crime against humanity” the “perpetrators and instigators” of which ought to be punished.162 Secretary of State Kellogg wrote to the French Ambassador in the months before the Pact’s conclusion, “From the broad standpoint of humanity and civilization, all war is an assault upon the stability of human society, and should be suppressed in the common interest.”163 Following the Pact, and with the carnage of World War I still fresh in memory, he described war as an “assault on human existence” and noted that the first thought in eliminating it was “the millions of wounded and dead” that result from armed conflict.164

The various efforts to ban the use of force reached their first global legal fruition in 1928 in the Kellogg-Briand Pact. Notably, the treaty’s preamble emphasized not sovereignty, but states’ “solemn duty to promote the welfare of mankind” through the “humane endeavor” of ending war.165 Picking up on this idea, Ruti Teitel has described the Pact as a treaty whose focus on humanity forged a “connection . . . between the two historical strands of jus ad bellum and jus in bello (humanitarian law).”166

None of this is to say that there would not be sovereignty benefits to banning aggressive war. Nor is it to say that these benefits were unrecognized by those involved in the drafting of the Pact. But those sovereignty benefits (exchanged for sovereignty sacrifices) were not the focus of the public reasons given for this very fundamental legal restructuring. Instead, the focus was on the human wrongs inflicted by those who initiate war without justification. If anything, advocates of the Kellogg-Briand Pact, including Kellogg himself, were forced to defendit against the charge that it ceded too much in the way of sovereign authority.167

E. Aggression and the Human Dimension in the Courtroom

This focus on the unjustified infliction of death and human suffering as the normative core of the issue was also palpable at Nuremberg and at Tokyo. Justice Jackson, the lead American prosecutor at the IMT, argued, “[W]hat appeals to men of good will and common sense as the crime which comprehends all lesser crimes, is the crime of making unjustifiable war. War necessarily is a calculated series of killings, of destructions of property, of oppressions.”168 In his retrospective on the trials, Jackson’s Nuremberg colleague, Whitney Harris, adopted a similar line, reasoning:

Hitler ordered the killing of men called to the defense of countries which German armies invaded at Hitler’s command. Insofar as these invasions were acts of pure aggression, and wholly without legal justification, the resultant killings offended the conscience of mankind just as the slaughter of persons in concentration camps offended universal conscience. . . . The killing of innocent human beings by order of heads of states is subject to substantially the same moral blame whether it is the killing of civilian populations in connection with war or the killing of troops resisting unlawful aggression. . . . It is not the fact of initiating and waging aggressive war which reprehends it, but that it is necessarily a course of killings and of brutality which is attained in no other relationship of man or nation.169

Nuremberg prosecutor and subsequent University of Chicago law professor Bernard Meltzer argued that “the Kellogg-Briand Pact and similar agreements are important, not because they directly made aggressive war a crime, but because, by destroying it as a defense, they made the instigators of aggression subject to the universal laws against murder.”170 Hartley Shawcross, the chief British prosecutor at the IMT, brought this framing into the courtroom, emphasizing in his closing statement that “where a war is illegal . . . there is nothing to justify the killing, and these murders are not to be distinguished from those of any other lawless robber bands.”171

These were not mere prosecutorial flourishes. The IMT judgment described aggression as “the supreme international crime differing only from other war crimes in that it contains within itself the accumulated evil of the whole.”172 Similarly, the NMT described aggression as the “pinnacle of criminality” due to its infliction of “horror, suffering, and loss.”173 The wrong the tribunals recognized in these statements is not the macro violation of sovereignty, but the aggregation of harms that that macro policy entails.174 It is the death and destruction internal to aggression that is evil, and it is the accumulation of that evil that warrants criminalizing the war.175

The IMTFE was more direct, holding that waging illegal war is the gravest crime because it entails “that death and suffering will be inflicted on countless human beings.”176 Going even further than their IMT counterparts, the prosecutors in Tokyo had supplemented the charge of waging aggressive war with the charge of murder as a crime against the peace, including in the latter the killings of enemy soldiers in the course of an illegal invasion.177 The Tribunal adopted this reasoning, declining to consider the murder charge only because it was deemed redundant to the charge of waging aggressive war. Specifically, the IMFTE held:

If, in any case, the finding be that the war was not unlawful then the charge of murder will fall with the charge of waging unlawful war. If, on the other hand, the war, in any particular case, is held to have been unlawful, then this involves unlawful killings . . . at all places in the theater of war and at all times throughout the period of the war. No good purpose is to be served, in our view, in dealing with these parts of the offences by way of counts for murder when the whole offence of waging those wars unlawfully is put in issue upon the counts charging the waging of such wars.178

More recently, a committee reporting to the ICTY on NATO’s aerial campaign against Yugoslavia in 1999 found that “a person convicted of a crime against peace may, potentially, be held criminally responsible for all of the activities causing death, injury or destruction during a conflict.”179 Similarly, British Attorney General Lord Peter Goldsmith warned Prime Minister Tony Blair of the outside possibility that he could be charged with murder for killings by British soldiers in the 2003 invasion of Iraq, should the war be deemed illegal.180 Meanwhile, Security Council condemnations of aggressions since World War II have expressed concern and indignation not only at the infringement of sovereignty, but also at the killing and human suffering entailed.181

F. The Wrong of Aggressive War

For the reasons provided in this Part, an account of aggression that locates its core wrong in unjustified killing better fulfills the three criteria of a good normative account of the law than does an approach that focuses instead on the harm to sovereignty.

First, identifying the wrong as the unjustified killing in an aggressive war explains the legal contours of the crime in a way that a focus on sovereignty or even self-determination cannot. The jus ad bellum restrains the sovereign capacity to vindicate legal rights using force as much as it protects the sovereign rights to territorial integrity, political independence, and self-determination. Moreover, nonbelligerent acts that more severely infringe territorial integrity, political independence, or self-determination than does aggressive war are notcriminal. Conversely, other forms of unjustified killing are criminalized in some other form. And illegal uses of interstate force that involve no sovereignty violation, or that vindicate territorial integrity or political independence rights without responding to an armed attack, are criminal aggressions. In each of these respects, neither sovereignty nor self-determination can make sense of the jus ad bellum and its role in international criminal law. In contrast, a focus on killing and human suffering is illuminating. Were it not criminal, aggressive war would be the key form in which such harms are inflicted on a massive scale without either being justified as a necessary and proportionate response to the threat or infliction of such harms by another, orbeing criminalized in some other way.

Second, the humanity-based explanation of aggression comports better with the overall purpose of the law in this domain, aligning both with the motivations for the initial outlawry of warand with the overall purposes of international criminal law as a key dimension of the “humanization” of international law. A connection between the killings in aggressive war and the wrong of murder ran through the heart of the prosecutions and convictions at Nuremberg and Tokyo, and has been reiterated in the work of several legal authorities since.

Third, a normative account rooted in humanity coheres with adjacent legal rules: explaining why aggression is the only criminal sovereignty violation and accounting for aggression’s fit within the broader legal approach to the unjustified infliction of human suffering. By contrast, as Luban and others emphasize, a crime that privileges sovereignty over humanity would sit extremely uneasily—and possibly in conflict—with the general humanizing posture of international criminal law and the noncriminal status of all other sovereignty violations.182

Aggression typically involves a sovereignty violation, but it is fundamentally a crime against human beings. It is a “law that has as its purpose protecting the fundamental right to life of millions of people.”183 What makes it special in international criminal law is not that it protects sovereignty, rather than humanity, but that it alone protects the right to life of combatants and proportionate collateral civilians against the wrongful violence of foreign states.

iv. two problem cases: bloodless aggression and humanitarian intervention

Before introducing the doctrinal implications of this normative reconceptualization, two potential problem cases should be addressed: bloodless invasion and humanitarian intervention. A “bloodless invasion” is an illegal military taking of territory or usurpation of governmental power without the infliction of casualties. Such actions are, by definition, unlawful. Although the legal status of humanitarian interventions lacking Security Council authorization is contested, the dominant view is that they, too, are unlawful. As illegal uses of force, a case can be made that both actions are also criminal.

The putative problem is this. Both actions involve a clear infringement of sovereignty. But neither appears to fit the unjustified killing paradigm. Bloodless invasions involve no killing at all, and the killing in a humanitarian intervention is responsive to the illegal human violence and killing of an atrocity crime in much the same way as the killing in a defensive war responds to the unjustified killing of an armed attack. Therefore, if such actions are criminal, the moral explanation would seem to be rooted not in human life, but in state sovereignty. Or so the objection would go.

This Part rebuts that concern. The criminal status of bloodless invasion at Nuremberg was dubious and marginal, and little has occurred to change this reality since. Indeed, there is space in Article 8 bis to exclude bloodless invasion from the crime altogether. This is telling, because if the sovereignty-focused, traditional account were true, bloodless invasions ought to be the paradigmatic example of the crime. The fact that they are not militates strongly againstthe traditional account. Moreover, the reason bloodless invasions are even potentially criminal under current law is rooted not in their impact on sovereignty, but in their latent killing. This is what underpinned the limited criminal status of bloodless invasions at Nuremberg and it is what distinguishes bloodless invasion from the noncriminal sovereignty violations discussed in Section III.A. Thus, if bloodless invasions are criminal, the reason why would only serve to add further support to the unjustified killing account.

The humanitarian intervention objection is weaker. There is no precedent for the prosecution of a leader of such an action and the contemporary illegality of such interventions is most plausibly rooted not in the principle that such wars infringe sovereignty, but in the worry that legalization would encourage wars waged on a humanitarian pretext, and thus unjustified killing. Here too there is both interpretive space and good reason to exclude humanitarian intervention from the crime. Even if it were to be included, it would be with reference to encouraging unjustified killing.

A. Bloodless Invasion

Under the pending Article 8 bis of the Rome Statute, a war is criminal when it infringes the “territorial integrity or political independence” of a state.184 Rodin argues that “this condition is both logically and factually independent of the question of whether the lives of individual citizens within the state are threatened.”185 Rodin here invokes the specter of so-called “bloodless,” or at least near-bloodless, aggression.186

The category is far from a null set. Most would recognize Russia’s 2014 invasion and annexation of Crimea as a prominent recent example. Following the Maidan protests, the flight and de facto abdication of President Yanukovych, and the pro-European takeover in Kiev in late February 2014, Russia took decisive military and paramilitary action in Crimea, purportedly to protect Russian nationals in the autonomous region.187 Far exceeding the terms of a 1997 agreement with Ukraine regarding Russian troops’ presence, the deployment enabled the armed takeover of the regional parliament, public buildings, and infrastructure.188 In a hastily arranged referendum, conducted under the shadow of that dominant military presence, Crimeans voted to secede from Ukraine and to accede to Russia, paving the way for further troop deployments and a full annexation. Moscow now claims Crimea as its own, near-global nonrecognition of the transfer notwithstanding.189 Most important for the discussion here, despite the clear impact of Russian military action in effecting the annexation, there were no significant casualties.190

Uses of force such as that in Crimea undermine what Walzer terms the “highest values of international society”namely, “the survival and freedom of political communities”curtailing the political independence and self-determination rights that he insists “are worth dying for,” even when no one actually dies for them.191 Even if the attacked state’s central politics are left intact and the population of the annexed region has little objection to their new sovereign, the territorial taking anyway violates the other core sovereign value protected by the jus ad bellum: territorial integrity.192 In short, such invasions involve severe and illegal infringements of sovereignty, but no human death or violence. If they are indeed criminal, this might seem to militate in favor of the traditional account and against the unjustified killing account of aggression.

In fact, the reverse is true. The criminal status of an illegal invasion that takes territory or usurps a foreign state’s government without spilling a drop of blood is ambiguous and marginal at best—a point that weighs heavily against the traditional sovereignty-focused account.

Unlike each of its other invasions, Germany’s relatively bloodless annexations of Austria and Bohemia and Moravia were excludedfrom the criminal indictment before the IMT, which then distinguished them explicitly from wars of aggression in its final judgment.193 This exclusion is telling. If the core wrong of aggression were its infringement of sovereignty or nullification of a people’s self-determination, bloodless aggression would be the paradigmatic form of the crime. Invasions that lack violent military confrontation eviscerate the territorial integrity and political independence of the victim states with far greater efficacy than do ordinary wars of aggression, many of which are unsuccessful in unseating the target government or holding territory.194 The Nazi annexations of Austria and Bohemia and Moravia exemplify this perfectly, so the point would not have been lost on those at Nuremberg.195

On the traditional normative account, then, it is difficult to make sense of either the failure to charge those invasions before the IMT, or of their separate normative status in the final judgment. On the account presented here, however, this marginalization makes perfect sense. For, while bloodless invasions are no less harmful to political independence and territorial integrity, they lack the same human violence as aggressive wars involving conflict and casualties.

Of course, although it was the most significant aggression prosecution to date, the IMT’s was not the only word on this issue. The NMT subsequently adopted a broader definition of the crime, which included the invasions of Austria and Bohemia and Moravia.196 It is in this sense that the Nuremberg legacy on bloodless invasions is ambiguous.

However, the reasoning for the NMT’s more expansive approach itself weighs heavily in favor of the unjustified killing account. In explaining the inclusion of these invasions as criminal aggressions, the NMT emphasized the fact that they were achieved against the backdrop of threatened slaughter by an overwhelmingly militarily superior Nazi force.197 What made these actions criminal, the tribunal emphasized in the High Command Case, was “the exerting of violence” by German forces, even though that exertion was met with no resistance and thus the latent killing it entailed was not ultimately consummated.198

Moral theorists critical of the Walzerian approach have started to bring precisely this notion of latent violence to the fore in related areas of philosophical discussion. Thus, as Cécile Fabre argues, a key normative link between bloodless invasions and traditional aggressions is that the former involve at their core “individuals posing a lethal threat, either ongoing or imminent, to other individuals.”199 This is what distinguishes both kinds of invasion from the kind of nonbelligerent, and noncriminal, sovereignty infringements discussed above.200 Whereas the latter violate political independence and territorial integrity without exerting violence, the former achieve that end by inflicting latent lethal harm.

This distinction is crucial. If bloodless invasions are criminal (as the NMT suggests they are), the key trait that defines them as such, just as in the ordinary aggression case, is the means by which sovereignty is violated, not the fact or degree of the sovereignty violation. The difference between an aggressive military invasion that achieves its sovereignty-infringing end bloodlessly and the illegal foreign manipulation of elections is akin to the interpersonal difference between mugging someone for her wallet with a lethal weapon and pickpocketing the same wallet from her without any physical threat.201 One inflicts latent violence; the other does not. If aggression is criminal, its criminality as compared to the mere illegality of otherwise similar sovereignty violations would parallel armed mugging’s felony status as compared to the misdemeanor status of nonviolently stealing the same item.202

How the ICC or domestic criminal courts will approach bloodless invasions is unclear. No legal authority has assessed the criminality of such actions since World War II. Indeed, the analysis and debate regarding the precise legal status of Russia’s 2014 annexation of Crimea has not even considered whether it might have been criminal (presumably due to the lack of any viable means of prosecution).203 Nonetheless, it is notable that in various deliberations on the matter at the Security Council and elsewhere, the concept of latent Russian violence to human life was raised frequently. Those condemning the operation as an act of aggression described Crimean voters as acting under “the barrel of a gun” and “in the shadow of Russian bayonets.”204 If the annexation of Crimea was criminal, this feature of the invasion was surely the crucial one.

Seen in this light, bloodless aggression does not provide a counterexample to the humanity-based account of the crime of aggression. On the contrary, in combination with the arguments in Part III, it fills out a context in which that is the only viable account of aggression. First, nonviolent but severe infringements of sovereignty are not criminal. Second, large-scale, lethal uses of international force not responsive to armed attacks are criminal, even when they vindicate the core sovereign rights related to territorial integrity, political independence, or self-determination. Third, bloodless invasions that threatensignificant human harm are at the margin of criminality—of mixed status at Nuremberg, and, as elaborated in Part V, of dubious criminal status under the ICC regime. Fourth, when including such invasions in the crime, NMT judges emphasized the latent violence of those actions.

This analysis deals with the core objection, but before turning to humanitarian interventions, it is worth addressing a second dimension of the worry about bloodless invasions—namely, that the lawfulness of wars waged defensively against such invasions might be thought to contradict the argument in Section III.B that the jus ad bellum allows for the use of force only in response to an attack on, or threat to, human life.205 This challenge differs from that of the purported criminality of bloodless invasions, because the right to self-defense is triggered not by subjection to criminal aggression, but by subjection to an “armed attack”—a distinct and autonomous legal concept.206 Nonetheless, the crux of the response is the same.

To warrant defensive force, an invasion must carry the immediate threat of lethal force; it must be an armed attack.207 A bloodless deployment of border officials and national flags to foreign territory would satisfy this criterion no better than the kind of electoral manipulation discussed above.208 Of course, the arrest or eviction of such an unarmed deployment could trigger an (illegal) use of armed force on the part of the invading state. Under existing doctrine, however, the right to self-defense arises only following the application of such force, or at least its certain imminent application, not merely because it may loom as the potential mechanism for preserving the effects of an unarmed intervention.209 Moreover, even when the immediate lethal threat to human beings is present, the use of defensive force is lawful only to the extent it is proportionate to the wrongful armed threat to which it responds.210

Just as it is essential to the potentially criminal status of a bloodless invasion, this element of latent violence to human persons is also vital to grounding the right to self-defense. Fabre argues, “The crucial question [in this context] is not whether sovereignty-rights themselves warrant defending by force; rather, it is whether [combatants of the attacked state] are under a duty to surrender those rights as a means to save their life and as an alternative to killing” combatants of the aggressor state.211 She contends that they are under no such duty becausetheir rights to defensive force are engaged by the “actual threat of future lethal harm.”212 Again drawing an analogy to violent mugging, those making this moral argument invoke the parallel disparity between the quantum of force appropriate in resisting an armed mugger demanding “your money or your life” and that which would be appropriate to prevent a nonviolent attempt to steal the same item.213

Only if we understand the right to defensive force as a right that responds to the aggressive infliction of latentviolence can we explain why bloodless military invasions trigger a legal right to the use of defensive force, while the nonviolent infringements of territorial integrity and political independence discussed in Part III do not.214 Here too the status of bloodless invasions coheres better with the unjustified killing account than with the sovereignty-focused, traditional account.

B. Humanitarian Intervention

The second apparent problem case is that of humanitarian intervention lacking Security Council authorization. A small minority of states and other authorities have argued that such interventions can be lawful.215 If this is right, humanitarian intervention poses no difficulty for the account presented here. However, most experts and states reject that position.216 If the latter view is correct, unauthorized humanitarian intervention might be thought to pose a difficulty for the unjustified killing account. Without taking a view on the legality of humanitarian intervention, the following discussion assumes such actions to be illegal for the sake of argument.

The putative difficulty posed by the presumed illegality of humanitarian intervention is as follows. Article 8 bis criminalizes all “manifest violation[s]” of the U.N. Charter’s rules on the use of force, as defined by their “character, gravity and scale.”217 Efforts during the amendment process to attach an explicit understanding excluding humanitarian intervention from the crime failed.218 As such, the argument goes, manifestly illegal humanitarian interventions are necessarily criminal.219 The problem is that genuine and proportionate humanitarian interventions use lethal violence only to defend against criminally wrongful killing and violence.220 In other words, they lack the wrong of aggression defined above—namely, the infliction of death and violence that is not justified by its response to the threat or infliction of the same.221

To be clear, humanitarian interventions do involve killing in an illegal action. In other words, they do involve legally unjustified killing.222However, the lack of legal justification for the killing in such an intervention is derivative of the illegality of the intervention. And the latter is not itself explicable with reference to the deeper wrong of killing without the justification of responding to the threat or infliction of illegal killing or analogous violence.

The problem posed by the purported criminality of humanitarian intervention, then, is that the lack of legal justification for killing in a humanitarian intervention might be thought to reflect a privileging of sovereign rights over the value of human life.223 Such interventions do, after all, tend to have significant impact on the political independence and territorial integrity of the target state.224 On that basis, if humanitarian intervention is criminal, this might seem to militate in favor of the traditional sovereignty-focused account, and against the unjustified killing account, of aggression.

As with the concern about bloodless invasion, however, the objection fails. First, it misrepresents the normative status of humanitarian intervention in the contemporary order. The illegality of such interventions is not about protecting sovereignty over human life; it is about banning “good” wars so as not to encourage or facilitate “bad” wars. If genuine humanitarian interventions are criminal, it is so as to prevent the unjustified killing of disproportionate or pretextual interventions. Second, as discussed in Part V, there is ample interpretive space, and good reason, to exclude presumptively illegal humanitarian intervention from the crime.

In any domain, there is an inevitable gap between even optimally crafted law and the internal normative foundations of that law.225 Put another way, there are always hard cases, where legal status does not reflect moral status, even when judged against the very moral standards on which the law is premised. This is for the familiar reason that optimal legal rules, unlike the underlying moral principles of right and wrong, must take account of moral hazard, the risks of abuse, the danger of slippery slopes, the collateral impact of rules on broader normative culture, and the “migration” of regulated behaviors out of the intended domain.226

On a normative account that has gained particular prominence since NATO’s 1999 intervention in Kosovo, and which offers the most coherent explanation of contemporary international law in this area, humanitarian interventions fall precisely into this gap between the law and its own underlying moral posture.227 Along these lines, an influential independent report on the Kosovo intervention declared NATO’s actions to have been “illegal but legitimate.”228 Prominent voices have endorsed this as the appropriate, long-term normative equilibrium for humanitarian intervention.229 And several of the states that participated in the campaign claimed moral justification, while insisting that the intervention ought not be understood as a legal precedent.230

Underpinning this idea is the view that an explicit ex ante legal permission for even genuinehumanitarian wars would facilitate the waging of both nonhumanitarian interventions on a humanitarian pretext and disproportionate interventions that far exceed their humanitarian purpose, and that these twin dangersoutweigh the benefit of enabling genuine, proportionate humanitarian interventions.231 In short, the aggregate risk of false positives under a permissive regime is thought to outweigh the risk of false negatives under a restrictive regime.

Pretext is a heightened danger here because the jus ad bellum is typically evaluated and enforced not by a judicial authority, capable of engaging in fine-grained, case-by-case analysis, but by states acting collectively to sanction and ostracize lawbreakers.232 In that sense the effective enforcement of the jus ad bellum depends on broad consensus in each case. Because the precise threshold for humanitarian intervention is less easily identified than is the armed attack threshold of self-defense, it would be difficult, were genuine humanitarian intervention lawful, for an uncoordinated population of states to engage in such collective enforcement against states that wage illegal war on a humanitarian pretext. Or so the theory goes.

Understanding the illegality of unauthorized humanitarian intervention in this way makes better sense of the contemporary legal structure than does the sovereignty-based account. It fits with the notion that the Security Council is thought to have a responsibility to authorize (and thus render lawful) precisely the same substantive action.233 The procedural requirement of Council authorization imposes a political and epistemic check that is likely to block the vast majority of pretextual interventions.234 Although some genuine humanitarian interventions will be prohibited, the presumption is that this is a defensible tradeoff, especially if the genuine interventions can be excused as legitimate post hoc.235 This understanding also makes sense of the debate regarding humanitarian intervention’s illegality. As a hard case on the law’s own normative terms, it is not surprising that different states and commentators have come to different legal views as to its permissibility.236

Most significantly here, this account of humanitarian intervention’s illegality is compatible with the normative account of aggression described above. Humanitarian intervention, on this understanding, is illegal not because it inflicts the wrong of massive killing without responding defensively to the threat or infliction of the same, but because permitting it would encourage wars that do inflict that wrong—that is, pretextual or disproportionate interventions. If humanitarian interventions are criminal, therefore, it is with a view to the prevention of unjustified killing, even though those interventions do not themselves inflict that wrong.

This reconciles the illegality (and possible criminality) of humanitarian intervention with the unjustified killing account. However, it also highlights the unusual nature of humanitarian intervention among illegal uses of force—namely, that it lacks the core criminal wrong of aggression. As discussed in the next Part, even assuming the illegality of humanitarian intervention, this militates against its criminality, and there is ample interpretive space to exclude it from the crime.

v. why it matters that aggression is a crime of unjustified killing

Getting the normative underpinnings of the crime of aggression right is not a purely theoretical exercise. Doing so has practical consequences for the law and its application. In forthcoming work, I elaborate three such consequences, which are introduced here.237 The first, raised tangentially in Part IV, emphasizes how understanding the moral meaning of a crime can help to guide interpretation of ambiguous aspects of that very criminal prohibition. The second and third identify implications for related rules governing how we treat soldiers on either side of an aggressive war.

A. The Utility of Normative Clarity in Resolving Ambiguous Cases

The first internal implication of identifying the wrong of aggressive war as the infliction of massive killing not in response to the same is that this clarifies the object and purpose of the crime, and thus sheds light on how the ICC should interpret ambiguous cases.238 As indicated above, both bloodless invasion and humanitarian intervention fall into this category. The first, although clearly illegal, has a mixed criminal law legacy; the second is of disputed legality and has not been evaluated in criminal case law.

The focus here is different from that in Part IV. There, the task was to explain why the (presumed) illegality of humanitarian intervention and the dubious (but presumed) criminality of bloodless invasions are compatible with the unjustified killing account of aggression. This Part turns to the separate and distinct question of what the unjustified killing account indicates regarding the optimal legal assessment of either action by the ICC. In contrast to the traditional account, the unjustified killing account of aggression weighs against the criminality of either action.

As argued above, if humanitarian intervention is currently illegal, it is “illegal but legitimate” by the law’s own normative lights. Two factors explain why this means that the best interpretation of Article 8 bis would exclude such wars. First, the condemnation and punishment inherent in criminal conviction heighten the imperative to narrow the gap between laws and the moral principles that underpin them.239 This is particularly so in a domain focused exclusively on “the most serious crimes.”240 The normative authority of international criminal law would suffer if the leader of a genuine humanitarian intervention were convicted of the “supreme international crime” of aggression, despite not perpetrating its core “accumulated evil.”241 In other words, the justificatory threshold for a regime under which an action is “supremely criminal but legitimate” is significantly higher than is that for a regime under which an action is merely “illegal but legitimate.”

Second, whereas the jus ad bellum is ordinarily enforced collectively by states, international crimes are enforced by a single judicial authority capable of reaching nuanced case-by-case judgments with authoritative effect.242 As such, unlike the broader jus ad bellum, international criminal law operates in an institutional context in which a rule permitting genuine humanitarian interventions would not thereby encourage large numbers of bad wars.243 In that sense, the key justification for humanitarian intervention’s illegality is inapplicable in the criminal context.

Consider these points together. The justification for criminalizing humanitarian intervention despite it lacking the core wrong is higher than is the threshold for justifying its illegality. And a core aspect of the justification for its illegality is anyway absent. Combining this with the observation above that humanitarian intervention lacks the core criminal wrong, there is good internal reason to exclude humanitarian intervention from the crime.

This frames how the ICC ought to interpret aggression. Article 8 bis (1) provides that only those illegal uses of force with the “character” and “gravity” to constitute a “manifest” violation of the Charter are criminal.244 As the attached understandings emphasize, “aggression is the most serious and dangerous form of the illegal use of force; . . . a determination whether an act of aggression has been committed requires consideration of all the circumstances of each particular case, including the gravity of the acts concerned and their consequences.”245

The normative account put forward in this Article demands measuring the “seriousness,” “gravity,” “character,” and “danger” of an illegal use of force by its unjustified infliction of human death and suffering. Pursuant to that reading, genuine humanitarian interventions would fall short of the criminal threshold. In that sense, Harold Koh and Todd Buchwald are right to insist that, “[w]hatever one’s legal views on whether humanitarian intervention is a permissible basis under international law for resorting to force, a true humanitarian intervention . . . should not entail the risk of international criminal prosecution.”246

For similar reasons, the normative account presented here can guide the interpretation of bloodless invasions. On the dominant, sovereignty-focused account, the proper interpretation of Article 8 bis would surely include future actions along the lines of Russia’s 2014 annexation of Crimea.247 The account presented here insists instead that such cases are not at all clear-cut. If anything, the stronger argument is for their exclusion from the crime.

This position may seem to be out of sync with the widespread condemnation of Russia’s annexation of Crimea, which a number of states labeled “aggression.”248 Indeed, there was good reason to use the term. As Tancredi observes, the Russian intervention exemplified several of the acts of aggression listed in Article 3 of the General Assembly’s definition.249 Those same acts are included in the identical list in Article 8 bis (2) of the Rome Statute amendment.

However, there is a crucial distinction. Public condemnations of Russia’s actions have not claimed criminality, and criminal punishment of the Russian orchestrators has never been on the table. This is important, because one can hold that the Crimean intervention was an act of aggression without endorsing the notion that it exemplified a criminal use of force. The “character” and “gravity” thresholds internal to Article 8 bis qualify “acts of aggression,” including precisely those listed in 8 bis (2).250 In other words, by the very structure of the provision, not all instances of the listed acts of aggression are criminal, only those of the requisite character and gravity.251

Measured with reference to unjustified human harm, the latent violence that could underpin the criminality of bloodless invasions is plainly of lesser “seriousness,” “danger,” and “gravity” and of a different “character” than consummated violence.252 As such, bloodless invasions fall somewhere between clearly noncriminal acts—like illegally, but nonviolently, manipulating foreign elections—and clearly criminal, high-casualty invasions. This raises the real possibility that even highly effective bloodless invasions are not criminal under Article 8 bis.

There is an obviously troubling aspect to letting the leaders of a bloodless invasion off the criminal hook. Had their presence been resisted by Ukraine, Russian troops stationed in Crimea would almost certainly have responded with lethal force. That was surely why they were stationed throughout the territory, and it is how Russian forces reacted to resistance in Donbass subsequently.253 In that sense, the bloodlessness of Crimea was due not to restraint in Moscow, but to supererogatory restraint in Kiev.254 To condemn and punish leaders for one of these invasion types and not the other might appear morally arbitrary.

However, such divergent treatment is commonly accepted in morality and law. Persons are judged morally not on what they would have done had circumstances outside of their control been different, but on what they actually did.255 Similarly, criminal law gives the perpetrator the strong benefit of the doubt, recognizing, however unlikely it may be, that a threat may never have been consummated, and that those who make a threat, even if it succeeds, should not be treated equivalently to those who make and consummate that threat. Notably, most other international crimes require more than latent unjustified violence: they require its direct infliction. There are rare exceptions, such as the war crime of “declaring that no quarter will be given” or the crime of inviting unjustified killing through shielding one’s own troops with civilians.256 But those crimes are anomalous and aggression is the “supreme” crime, defined by an internal gravity threshold.257 There is good reason to hold that bloodless invasion falls short of that threshold.

To be clear, the noncriminality of such actions is compatible both with the dominant view that Russia engaged in a serious violation of public international law, and with the collective and severe sanctions imposed on Russia in response, including nonrecognition of the annexation.258 Bloodless invasions are serious violations of international law that warrant significant consequences. The argument above goes only to the question of whether the leaders of states that engage in such actions should be prosecuted as the world’s supreme criminals, alongside perpetrators of genocide, crimes against humanity, and war crimes.

This is not to say that the normative account presented here disposes unequivocally with the question of bloodless invasion’s criminality. As explained in Part IV, the unjustified killing account is compatible with a lower threshold that would include bloodless aggressions as criminal.259 However, in recognizing that such invasions are less grave on the core measure of the crime’s wrongfulness, and in stark contrast to the dominant account, the argument weighs against including such invasions among the “most serious and dangerous” forms of illegal force that count as instances of the “supreme crime.” The best interpretation of Article 8 bis is that bloodless invasions are not criminal.

Even if it were thought appropriate, after reflection and debate, to adopt a lower threshold and count some bloodless invasions as criminal aggressions, the account presented here would identify a criminality threshold within the category of bloodless invasions. Specifically, criminality would turn on the scope of the action’s latent human violence.

B. The Right To Disobey

The second practical implication of the normative account presented here goes to the legal rights of soldiers ordered to fight in criminal wars. Viewed on the traditional account as remote, and thus morally detached, from the macro wrong against the victim state, such soldiers have been denied an internationally protected right to refuse to participate. On the account presented here, that is a mistake. Although typically nonculpable, soldiers are involved intimately in the unjustified killing and violence that are the core criminal wrongs of aggressive war. A better interpretation of existing law would recognize their right to refuse to perform those wrongs.

It is legally clear that only those who control, or at least influence, the policy to wage illegal war can be criminally liable for aggression.260 On the Walzerian normative account, this reflects the deeper moral truth that soldiers are responsible only for what falls within “their own sphere of activity.”261 They are detached morally from any macro interstate wrong, because they lack influence at that level.262 This idea of a tightly defined sphere of soldierly activity and responsibility underpins the so-called “moral equality” of combatants—the idea that soldiers waging an aggressive war are no less licensed morally to kill combatants (and proportionate collateral civilians) than are their opponents.263

The idea that trivial or remote contributions to a macro wrong can be subject so easily to moral detachment ought to be controversial from the legal point of view, particularly in light of international criminal law’s broad and oft-used doctrines of shared responsibility.264 But even accepting that premise, on the unjustified killing account, soldiers are notremote from the wrong of aggression. Quite the opposite: the interstate violation is criminal because the killings of the soldiers on the aggressor side, unlike those of their enemies, cannot be justified. The moral foundations of that legal posture are best explained by Walzer’s revisionist critics, who have insisted that it is untenable to hold that soldiers participating in an aggressive war have a moral license to “kill people who have done nothing other than to defend themselves and other morally innocent people from an unjust attack.”265

To identify the legal relevance of this insight is notto hold that soldiers should be criminally liable for participating in aggressive wars.266 On the contrary, there are two cumulative reasons to grant soldiers blanket jus ad bellum immunity. First, their typical uncertainty about the war’s legality, their associative sympathies, and the heightened culpability threshold for international criminal liability mean that very few soldiers are sufficiently culpable for their wrongful killings to warrant prosecution.267 Moreover, it is institutionally infeasible to identify the few that surpass that threshold.268 Second, guaranteeing soldiers jus ad bellum immunity arguably helps to frame a set of incentives maximally conducive to jus in bello compliance and thus mitigates the horrors of war, once underway.269 Given soldiers’ typical nonculpability, sharpening the impact of the in bello regime in this way sets the best incentives for good conduct.

However, these good reasons not to punish soldiers for aggression cannot reverse the wrongfulness of their jus-in-bello-compliant killing in a criminal war.270This is not merely a point about philosophical accuracy. It goes to the proper interpretation of soldiers’ rights.

Soldiers have a clear right to disobey orders that violate jus in bello.271 Those punished for such refusal are eligible for refugee status.272 Neither the core right, nor the refugee claim, requires that the soldier would have been personally criminally liable had he obeyed the illegal orders; it is enough that, if obedient, he would have been sufficiently intimately associated with the wrong to be unable to “wash his hands of guilt.”273

Similar rights have not been extended to those who refuse to fight in criminal wars of aggression. When not relying on the political question doctrine to ignore such claims, domestic and foreign courts have reasoned that there is no right to disobey such orders because there is no legal duty to do so—because, in other words, the individual soldier’s contribution to a criminal war is not itself wrongful from the legal point of view.274 This position was applied even in Germany after the IMT judgment to those who had refused to fight in Nazi aggressions.275

If the normative core of aggression were a macro interstate wrong, treating jus ad bellum resisters in this way might be normatively comprehensible given how far removed they are from that macro violation. However, if soldiers’ contributions to an aggressive war are the wrongs that make the war worthy of criminalization, soldiers that refuse to fight on those grounds sound their claims in the law’s own normative register. They cannot be told coherently that they ought to be able to wash their hands of guilt.

As I explore in forthcoming work, perhaps the most plausible alternative explanation for the denial of soldiers’ rights to disobey rests on the damage that recognizing such rights would do to the functioning of military institutions in lawful wars, and thus to global security.276 This justification, however, is contingent on both the nature of war and the nature of military institutions. That contingency underpins an imperative internal to international law’s own normative posture to carve out disobedience rights for soldiers refusing to fight in criminal wars whenever doing so would be compatible with preserving military functionality in lawful wars.

The results of two isolated cases offer hope for a more coherent jurisprudence in this respect. Mohammed Al-Maisri, a Yemeni deserter from the Iraqi invasion of Kuwait in 1990, was granted asylum in Canada in 1995 on the grounds that “the invasion and occupation of Kuwait was condemned by the United Nations.”277 And Florian Pfaff, a German Federal Army Major who refused to provide support to the U.S.-led invasion of Iraq in 2003, had his right to disobey upheld domestically by the German Federal Administrative Court on the grounds that he identified “objectively serious legal reservations” to the war.278 In neither case did the court explore the nature of the soldier’s relationship to the wrong in question. However, the holdings protected the right of those individuals not to violate the rights of others.279

C. Reparations and Participation at the ICC

The third and final implication of the normative account offered here goes to the rights of soldiers fighting against aggression. Two of the ICC’s landmark innovations in international criminal law are that it provides crime victims the opportunity to participate (with legal representation) in various ways in the criminal proceedings and that it renders such victims eligible to receive reparations following a conviction.280 The upshot of the account articulated in Part III is that, in an aggression prosecution, those participatory and reparative rights and privileges must attach primarily to combatants killed or injured fighting an aggressor force and to civilians harmed in proportionate collateral damage. These are the constituencies whose rights the criminalization of aggression protects.281

Certain aspects of the ICC’s developing approach to defining the scope of victims in nonaggression prosecutions are worthy of note here. In the first decision on reparations in 2012 (following Thomas Lubanga’s conviction for child conscription), the Trial Chamber articulated a potentially broad standard, holding that the Court could issue reparations to all whose harms were proximately caused by an international crime, including both direct and indirect victims.282 Although affirming these points in the abstract, the Appeals Chamber interpreted the terms narrowly.283 Declining to include the full spectrum of persons harmed as a foreseeable consequence of the crime, it identified “direct victims” with reference to the rationale for the criminal prohibition—namely the protection of children from the fear and violence of combat and from the trauma of separation from family and school.284 On this reading, the direct victims of Lubanga’s crime were the conscripted children, and the harms relevant to the reparations proceedings were the physical injury, psychological trauma, and loss of schooling associated with that criminal wrong.285 In other words, the direct victims were those the criminal prohibition was “clearly framed to protect”—terminology used by the Trial Chamber in an earlier effort to follow Appeals Chamber guidance on the victims eligible to participate in the criminal proceedings.286 The Appeals Chamber limited “indirect victims” for the purposes of both participation and reparations to those who suffered due to a “close personal relationship[]” to direct victims or who were harmed in the course of protecting direct victims.287

The focus on core victims of the crime, rather than all persons foreseeably harmed by the crime, is probably inevitable given the wide scope of international crimes and the limited capacity of the Court.288 It is also arguably more in line with the context of a criminal prosecution, which might be thought to require a focus on repairing the criminal wrong, rather than providing comprehensive civil compensation.289

If this approach to defining victims holds in the context of aggression prosecutions, the implications of the traditional normative account are clear. For those who subscribe to that view, “the typical victim [of the crime of aggression] is a ‘state,’” and the natural implication is to accord victims’ rights and privileges to state representatives, with the ICC channeling reparations to the attacked state following a conviction.290

On the unjustified killing account, this would be a mistake. The state is not the person (and its sovereignty is not the value) that the criminalization of aggression was “clearly framed to protect.” The criminalization of aggression protects combatants’ and collateral civilians’ right to life.291 The harm by virtue of which aggression is criminally wrongful is that inflicted through unjustified violence against human beings.

Seen in that light, those granted participatory rights and privileges in aggression prosecutions must be combatants who fought against aggression and civilians harmed as collateral damage, probably as two or more broad classes with collective representation.292 Reparations should focus on the same groups, funding veterans’ care or reintegration programs, assisting the families of dead soldiers and collaterally harmed civilians, and otherwise acknowledging the true wrong at the heart of the crime.293

In addition to diverging radically from a state-focused participation and reparations regime at the ICC, this upshot also stands in stark contrast to recent jus ad bellum reparations regimes in the compensation commission context. Consider here the U.N. Compensation Commission (UNCC), set up after Iraq’s 1990 invasion of Kuwait, and the Ethiopia-Eritrea Claims Commission (EECC), set up following Eritrea’s armed attack on Ethiopia at the turn of the millennium. Both commissionsissued a wide array of awards relating to the losses suffered by Kuwaiti and Ethiopian businesses and civilians as a result of the respective wars.294 Among the many claims deemed eligible by the more expansive UNCC were traffic accidents that occurred due to the general breakdown of civil order in Kuwait, individual and corporate property losses caused by that breakdown, and losses caused when the continuation of contracts became impossible for one or more of the parties due to the conflict.295However, despite their generally broad approaches to defining eligible victims, both commissions excluded almost all harms suffered by combatants fighting against the aggressor force, other than those caused by jus in bello violations.296

As indicated above, that a compensation commission would take a broader reparative approach than a criminal court is quite appropriate. However, in light of what is most fundamentally wrong with aggressive war from the internal point of view, the marginalization of soldiers’ deaths and suffering was a grave mistake that the International Criminal Court must not replicate. Together with collaterally harmed civilians, those combatants should in fact be at the very heart of participatory and reparative rights at the ICC.

These three implications—clarifying the marginal status of humanitarian intervention and bloodless aggression, the soldier’s right not to fight in an illegal war, and the centrality of combatant death and suffering to victim status and reparations at the ICC—are only sketched in introductory form here. Plainly, each demands more detailed elaboration and consideration. However, this sketch illustrates why defining the normative core of aggression matters doctrinally. With the Court poised to take jurisdiction over the crime, clarity on these issues is a matter of urgency.


This Article has debunked a common misconception—that aggression is a normative anomaly in international criminal law, uniquely rooted in a wrong inflicted on the attacked state, rather than in an accumulation of wrongs inflicted on individual human beings. Not only are individuals wronged gravely in an aggressive war, the wrongfulness of their treatment as individuals is at the very crux of what explains the criminalization of such wars. For five reasons, wrongful killing—and not aggression’s typical infringement of sovereignty—is the normative core of the crime.

First, sovereignty is indeterminate as a normative guide on the issue of aggression. The criminalization of such wars is at least as great a restraint on state sovereignty as it is a protection of it. Moreover, other sovereignty violations that more effectively and dramatically infringe the core sovereign rights of territorial integrity or political independence, or the collective right of self-determination, are not criminalized. Second, what distinguishes aggression from other sovereignty violations is that it involves unjustified, direct attacks on the lives and physical integrity of human beings. Indeed, some criminal wars involve the infliction of unjustified death and human suffering without infringing sovereign rights at all. And, crucially, if aggression were not a crime, it would be the anomalous context in which unjustified killings were noncriminal. Third, understanding aggression in this way reconciles it to the broader moral project of international criminal lawnamely, the protection of individuals and groups from the most egregious violations of their rights and dignity. Fourth, the claimed imperative to incorporate a restrictive jus ad bellum into twentieth century international law was articulated not in terms of sovereignty, but in terms of humanity. Finally, the reasoning of the judges and prosecutors at Nuremberg and Tokyo reveals a shared understanding that the prohibition of wrongful killing is the normative core of the crime.

For all of these reasons, the account of aggression advanced here better explains the contours of the crime, comports more closely with the purposes of international criminal law, and better aligns with adjacent areas of international law. Recognizing this ought to inform how we interpret the criminal law status of bloodless invasions and humanitarian intervention, how we understand the claims of soldiers who refuse to fight in criminal wars, and how we define the expressive focus of victim participation and reparations in ICC aggression prosecutions.