Chapeau elements of a crime against humanity in the Rome Statute

The Rome Statute defines the chapeau elements of crimes against humanity in Article 7, paragraph 1, as: ‘any of the following acts when committed as part of a widespread or systematic attack directed against any civilian population, with knowledge of the attack’.^{Footnote 4} I discuss in turn ‘widespread’, ‘systematic’, ‘attack’, ‘policy’, ‘civilian’, and later discuss ‘intent’, and ‘knowledge’, I use a variety of sources to expound the definitions of these key terms, including the Rome Statute, its Elements of Crimes, case law from the ICC and ad hoc tribunals, and scholars’ interpretations of these terms. I additionally cite the Oxford English Dictionary (OED) because some key terms in the Rome Statute differ substantially from everyday usage; it provides a useful comparison to these common words, and the ICC itself refers to it (Prosecutor v Jean-Pierre Bemba Gombo 2009: para 29).

Consider first the meanings of ‘widespread’ and ‘systematic’, The language in the Rome Statute is disjunctive: either ‘widespread or systematic’, Now consider some definitions from the OED and cases. According to the OED, widespread means ‘distributed over a wide region; occurring in many places or among many persons’.^{Footnote 5} Systematic refers to ‘arranged or conducted according to a system, plan, or organized method; involving or observing a system; (of a person) acting according to system, regular and methodical’.^{Footnote 6} These commonsensical meanings are consistent with interpretations of similar language in the statutes defining crimes against humanity offered by judges of the International Criminal Tribunal for the Former Yugoslavia (ICTY) and International Criminal Tribunal for Rwanda (ICTR). According to an ICTY judgment, widespread can be assessed by the ‘large-scale nature of the attack and the number of the victims’ (Prosecutor v Vidoje Blagojević and Dragan Jokić 2005: para 545). Other factors that can be taken into account to determine if an attack is systematic are ‘consequences of the attack upon the targeted population, the number of victims, the nature of the acts, the possible participation of officials or authorities or any identifiable patterns of crimes’ (Prosecutor v Vidoje Blagojević and Dragan Jokić 2005: para 546). ICTR judges in the Akayesu case wrote:

State actors need not carry out such policies in order for an act to qualify as widespread or systematic.

Given these definitions, many people must be harmed severely in order for an action to qualify as widespread or systematic. Being too specific about what this number is will not permit inclusion of cases that should qualify, but being too broad will allow the inclusion of too many crimes or be too vague to be helpful. For instance, if I suggest that at least 1,000 or 10,000 people must have their relevant rights violated by a leader’s actions to qualify as widespread or systematic, it is unclear how we can move from the language in the Rome Statute to such a specific claim. The problem with this approach, too, is that leaders might harm 999 or 9,999 and thereby avoid prosecution, but this intuitively seems wrong. There is little moral difference between harming 999 or 1,000 people. What may be helpful here is the former chief prosecutor’s suggestion regarding the gravity requirement that at least hundreds of people must be severely harmed (Moreno-Ocampo 2006: 8–9). This suggests, without requiring, that ‘widespread’ or ‘systematic’ typically means that at least hundreds or thousands of people must be severely harmed or killed. Such a number is consistent with the gravity requirement as well. All of my examples of nonviolent crimes against humanity easily meet this requirement.

Unlike widespread and systematic, the definition of ‘attack’ in the Rome Statute differs substantially from its common usage. It is normally taken to mean ‘the act of falling upon with force or arms, of commencing battle’,⁷ Rather, the ICC defines ‘attack’ in Article 7.2(a) as:

Unlike earlier formulations of crimes against humanity that required a nexus to armed conflict such as those in the Nuremberg Charter, following the 1954 Draft Code of Offenses against Peace and Security of Mankind the Rome Statute’s definition of attack thus does not require the use of armed forces or a nexus to armed conflict (International Law Commission: Article 2(11); Schabas Reference Schabas2007: 99–102). The Rome Statute’s absence of a requirement of an attack being committed by armed forces is consistent with the Tadić, Kunarac, Akayesu, Musema, and Vasiljević decisions from the ICTY and ICTR wherein the tribunals considering those cases concluded that a crime against humanity can be effected without force or military action (Prosecutor v Dusko Tadić a/k/a ‘Dule’ 1995: para 141; Prosecutor v Dragoljub Kunarac, Radomir Kovač and Zoran Vuković (Appeal Judgment) 2002: para 86; The Prosecutor v Jean-Paul Akayesu (Trial Judgment) 1998: para 581; Prosecutor v Mitar Vasiljević 2002: para 29; The Prosecutor v Alfred Musema 2000: para 205; The Prosecutor v George Anderson Nderubumwe Rutaganda 1999: para 70). Thus, ‘attack’ means simply conduct involving the multiple commission of acts that are part of or constitute a policy, which includes almost any state policy and many nonstate actors’ deeds (Schabas Reference Schabas2007, 101–2; The Prosecutor v Jean-Paul Akayesu (Trial Judgment) 1998: para 580). ICC judges wrote that an attack does ‘not necessarily equate with a ‘‘military attack’’’ (Prosecutor v Jean-Pierre Bemba Gombo 2009: para 75). They continued, ‘the commission of the acts referred to in Article 7(1) of the Statute constitute the ‘‘attack’’ itself and, beside the commission of the acts, no additional requirement for the existence of an ‘‘attack’’ should be proven’ (Prosecutor v Jean-Pierre Bemba Gombo 2009: para 75). The broad definition of attack in the Rome Statute allows nonviolent crimes against humanity to meet the attack criterion.

Third, a ‘policy to commit such attack’ requires that the State or organization actively promote or encourage such an attack against a civilian population’ (‘Elements of Crimes of the International Criminal Court’ 2002: 5). Although the policy requires encouraging the attack (‘the multiple commission of acts’), it does not require the intention that these actions result in harms or deaths. (Like the broad definition of attack, this definition of intention will be important for my argument.) In fact, there is no malice requirement for either ‘attack’ or ‘policy’ (Starr Reference Starr2007: 1304). It is important to note what is not included in the policy requirement. Actions need not be part of a larger state or organizational policy – although they can be – because the language in the Elements of Crimes allows any person in power to make policy, even if it diverges from the official policy or orders of superiors. Those actually committing multiple harms of the relevant types could also be convicted of a crime against humanity if their actions were part of the relevant policy. A serial killer does not commit crimes against humanity because the decision to kill is taken alone (Ambos Reference Ambos and Sadat2011: 285). But those same actions might be prosecutable if the killings were part of an organizational policy (assuming the other elements of a crime against humanity were met). ICC judges held that the policy requirement ‘implies that the attack follows a regular pattern’ (Prosecutor v Jean-Pierre Bemba Gombo 2009: para 81). Any ‘attack which is planned, directed or organized . . . will satisfy this criterion’ (Prosecutor v Jean-Pierre Bemba Gombo 2009: para 81). Proving intent and knowledge may seem especially damaging to my argument, but below I show this is not the case because of the way the Rome Statute defines the two terms.

Civilians are defined in international law as not combatants (‘Protocol Additional to the Geneva Conventions of 12 August 1949’ 1977, section 50). Four conditions are necessary and sufficient to be a combatant: wearing a symbol recognizable at a distance that identifies one as part of a military force, carrying arms openly, following the laws of war, and being subject to a command structure (May 2007: 103). Nonstate rebels may qualify as combatants as well as regular military personnel. Controversial cases of who qualifies as a combatant are not important for my argument because all of the examples I focus on clearly affect large numbers of civilians and for this reason I will not focus on debates about who qualifies as a civilian.

Having clarified the chapeau elements, I now turn to the enumerated elements of a crime against humanity. Instead of reproducing all of the enumerated elements in Article 7.1, which includes 11 main and many more clarifying definitions of what constitutes the relevant acts, I focus on only Article 7.1(k) here. I do this because in order to qualify as a crime against humanity only one of the 11 transgressions included in Article 7 must be met, in additional to the chapeau elements. Article 7.1(k) reads as follows:

Paragraph 2 of Article 7 clarifies possibly controversial terms in paragraph 1, but the Rome Statute does not elaborate on (k). Acts that cause great suffering or serious injury to one’s body, physical, or mental health that are widespread or systematic could constitute part of the actus reus of a crime against humanity.

During the Rome Statute’s formative debates, some pressed for crimes such as human-caused mass starvation to be included as a crime against humanity, but that language was ultimately rejected (Schabas Reference Schabas2007: 105; von Hebel and Robinson 1999: 103). Given this rejection, is it legally permissible to argue that some acts can constitute nonviolent crimes against humanity? The standard way to interpret treaties comes itself from a 1969 treaty called the ‘Vienna Convention on the Law of Treaties’ (‘Vienna Convention on the Law of Treaties’ 1969). The relevant clause says that a ‘treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose’ (‘Vienna Convention on the Law of Treaties’ 1969, section 31(1); Schabas Reference Schabas2007: 200–1). This highlights a possible tension in how to interpret international law. On the one hand, it is valid to read and interpret the language of an international treaty straightforwardly. On the other hand, there is a long tradition of looking to the intent of the founders by referring to the travaux preparatoires to assess how any given clause of an international treaty should be interpreted if the language is unclear or controversial. The strategy I take in this paper is to use a straightforward reading of the Rome Statute and to support this claim through multiple sources of jurisprudence and secondary literature. Thus, it is legally permissible to argue that those who aimed to have human-caused famine included as a crime against humanity did not ultimately fail.^{Footnote 8}

What sorts of harms qualify as ‘other inhumane acts of a similar character intentionally causing great suffering, or serious injury to body or to mental or physical health’ (Article 7.1(k))? Causing someone to die unnecessarily is a serious injury to physical health. Severe malnutrition is another example of serious injury to physical and mental health, even if one can sometimes recover from it. Such harms violate the international law codified in the ICESCR (‘International Covenant on Economic, Social and Cultural Rights’ 1976). Because some might object to the mental injury component, I limit my argument to physical harm. I assume it is uncontroversial and therefore provide no further argument that causing death by starvation or severe malnutrition is a severe physical harm. Of course, not all death or malnutrition is or should be criminal. Death and malnutrition can be caused in numerous ways including through poverty that no one is responsible for causing. But death and severe malnutrition can result from the active diversion of food and aid, as was the case in parts of Ethiopia in the mid-1980s or in Sudan after president Bashir was indicted by the ICC in 2009, by stealing money that would otherwise go to life-sustaining programmes, or by other policies. The Elements explicitly allow that a crime against humanity can occur through extermination by ‘deprivation of access to food and medicine’ (‘Elements of Crimes of the International Criminal Court’ 2002, Article 7(1)(b) 1. note 9) so such deprivation meets the ‘inhumane acts’ criteria.

Intent and knowledge

I will now show that even if a policy maker does not intend the outcomes of some acts that result in widespread or systematic harms, she can still be legally liable for a (nonviolent) crime against humanity. What I do not have to show for my argument to be convincing is equally important. A key point of the above exegesis is to show that neither ‘attack’ nor ‘intent’, as we normally understand them, needs to be present because of the broad way the Rome Statute defines those two terms. What I must show is that the ICC has the authority to try such types of crimes (as I did above) and that there is good reason to think that some actions do qualify as nonviolent crimes against humanity under Article 7.

The definition of intent in the Rome Statute is broad. According to the OED, ‘intent’ means ‘that which is willed’,⁹ When one intends something, one tries to realize an outcome that one has in mind. The Rome Statute’s definition differs substantially from this commonsensical usage. The Elements of the Crimes defines intent regarding ‘other inhumane acts’ as ‘The perpetrator was aware of the factual circumstances that established the character of the act’ (‘Elements of Crimes of the International Criminal Court’ 2002: Article 7(1)(k) 3). Starr takes this to ‘refer to those circumstances that render the consequences the ordinarily expected result of the act’ (Starr Reference Starr2007: 1257).

This coincides with what the ICC’s own judges have decided the intent requirement of Article 30 requires. Before discussing this, I quote Article 30, which discusses the mental element (mens rea). It reads as follows:

To be legally responsible, a person must have both intent and knowledge, which seems at first to significantly raise the bar for criminal liability (see May 2005: 124–32). However, ‘intent’ and ‘knowledge’ are defined broadly. A person must mean to engage in some conduct, but the second half of Article 30(2)(b) reads as follows. Someone must simply be ‘aware that it [a consequence] will occur in the ordinary course of events’ as a result of their action. Intent, here, just means being aware that an action will cause some effect ‘in the ordinary course of events’ – one need not want or hope to cause the consequences. This radically broadens what intent means when compared with a common-sense understanding of the term. ICC judges used this reasoning in the conviction of Thomas Lubanga Dyilo (Prosecutor v Thomas Lubanga Dyilo 2012: paras 1273–9 and 1351).

The meaning of ‘knowledge’ is similarly broad. To have knowledge according to the Rome Statute, one needs only ‘awareness that a circumstance exists or a consequence will occur in the ordinary course of events’. The ICC’s judges have determined on several occasions that the knowledge requirement is met if the perpetrator ‘knew this [an outcome] would occur, or was aware that there was a substantial likelihood that the crimes would occur’(Prosecutor v Thomas Lubanga Dyilo 2011: para 40). In the Bemba case, the judges ruled that ‘the perpetrator must be aware that a widespread attack directed against a civilian population is taking place and that his action is part of the attack’ (Prosecutor v Jean-Pierre Bemba Gombo 2009: para 88). To be convicted, the accused need not know the details of the attack (Prosecutor v Jean-Pierre Bemba Gombo 2009: para 88). He simply must be aware only that the relevant wrongs will occur in the ordinary course of events (Prosecutor v Jean-Pierre Bemba Gombo 2009: paras 87–88).

The standard of knowledge the ICC uses requires a minimal capacity to reason, knowledge of how the world functions, and responsibility for learning the probable outcomes of one’s actions. There need not be actual evidence of this knowledge. The Elements hold that ‘existence of intent and knowledge can be inferred from relevant facts and circumstances’ (‘Elements of Crimes of the International Criminal Court’ 2002, General Introduction section 3; The Prosecutor v Thomas Lubanga Dyilo 2007: paras 350–354). The ICC can reasonably assume that leaders generally possess the requisite knowledge of first-order effects of their actions because any reasonable person would who is an adult of average intellectual ability. Moreover, leaders typically keep themselves informed. Even Mao’s fawning Communist party underlings informed him of the famine he was causing through his collectivist agricultural policies (Dikötter Reference Dikötter2010; Waal 1997: 18–19). Peng Duhai confronted Mao with evidence of the great famine he was causing. Instead of taking this seriously and changing his policy, Mao purged Duhai (Waal 1997: 18–19).

The Rome Statute’s definition of intent additionally undermines a widespread intuition that intentionally wronging others or the state is worse morally than unintentionally causing identical harms. This intuition is reflected in many legal systems. Individuals who commit a crime through negligence or recklessness are generally treated less harshly than those who premeditate and intend to inflict a severe harm. But the Rome Statute’s definition of crimes against humanity blurs that typical distinction because of its definitions of knowledge, intent, and policy. Individuals can be held criminally responsible for some outcomes they do not mean to cause, but which will occur in the normal course of events. The policy requirement of crimes against humanity holds that someone must mean to make the relevant policy, but the intent and knowledge requirements do not require that the suspect intend (in the everyday definition of the term) the outcomes which result. The basic point I wish to make here is that doing certain things in a position of power will produce lots of consequences in the normal course of events just because one is in that position of power. The Rome Statute assigns international legal responsibility for some acts even if the power wielder does not intend the outcome. (I address in the Objections section whether this is a design flaw of the Rome Statute.)

The Rome Statute thus stretches our commonly held beliefs about the limits of legal responsibility for political leaders and many powerful nonstate actors. My interpretation of the Rome Statute would probably make many more leaders criminally liable because many set avoidable policies that they know will result in widespread or systematic harms of the relevant types to civilians. This has a parallel in Michael Walzer’s argument regarding political action and the problem of ‘dirty hands’ (Walzer Reference Walzer1973). Walzer claims that the structures in which politicians and some individuals must act put them in positions in which all of their available choices are morally wrong. Even if Walzer is correct that extremely powerful people sometimes face a moral dilemma, the powerful do not face a legal dilemma. The powerful can avoid committing crimes against humanity, as I discuss in the Objections section below.

In sum, to be convicted of a crime against humanity, someone can intend to commit multiple acts as part of a policy (7.2(a)) that result in widespread or systematic harms of the relevant types (7.1) which cause great suffering or serious injury to bodily or mental health (7.1(k)) to civilians (7.1) with the knowledge that these harms will occur in the normal course of events (30).

Unintended perverse incentives

Prosecuting nonviolent crimes against humanity might create perverse incentives by making current leaders clutch to power ever more tenaciously and ruthlessly in order to avoid punishment. These leaders may reason they have everything to lose if they fall from power because the ICC could prosecute them, and that killing more people violently or nonviolently to stay in power would not increase their punishment much if at all. This may be the true. There are several responses to this objection.

First, alleged international criminals already have strong incentives to remain in power. It is uncertain whether the threat of prosecution would change leaders’ incentives. The prosecutor has the power to not indict individuals under Article 53, if not prosecuting an alleged perpetrator were in the interest of justice, taking into account the interests of victims. Because the Rome Statute does not define the interests of justice or victims, the prosecutor has wide leeway in making this assessment. Therefore, the prosecutor can claim credibly to not prosecute a leader if he changes course and refrains from or stops committing nonviolent crimes against humanity, undercutting the perverse incentive.

Second, the opposite side of such an incentive is deterrence. Individuals may choose not to commit nonviolent crimes against humanity if they could be prosecuted for such offences. Although the research on deterrence for international crimes and associated harms is mixed (Drumbl Reference Drumbl2007: 169; Mennecke Reference Mennecke2007: 323; Snyder and Vinjamuri Reference Snyder and Vinjamuri2004), there is some evidence that prosecutions deter others in similar cases (Kim and Sikkink Reference Kim and Sikkink2010). I do not advocate fiat justitia, et pereat mundus, however. Elsewhere I argue that the ICC should not indict and prosecute any alleged international criminals if it were likely to result in many more people being harmed severely over the long term. In general, the prosecutor can and should prosecute nonviolent suspects, but she should assess carefully the trade-offs for each case and may sometimes decide to not prosecute – just as she would with violent international criminals.

Third, the creation of perverse incentives should be weighed against the intrinsic importance of holding leaders to account and the instrumental importance of prosecuting in order to deter others. Another important aspect of international criminal law may be norm enforcement (Damrosch Reference Damrosch1994; Drumbl Reference Drumbl2007). Mark Drumbl calls this ‘expressivism’ and Luban calls this ‘norm projection’ which roughly hold that an important function and justification of international law is bolstering respect for international norms (Drumbl Reference Drumbl2007: 11–12, 16–17, 60–3, Chapter 6; Luban Reference Luban2006: 354–5). Even if some leaders have marginally more incentive to remain in power, the moral value of holding individuals to account may outweigh such negative factors in some cases.

Fourth, trying nonviolent offenders may prevent perversely incentivizing sly political leaders from shifting their strategies from violent ones to nonviolent ones to avoid ICC prosecution. Expanding the scope of who can be prosecuted pressures those in office to refrain from and avoid committing nonviolent crimes against humanity. Thus, on intrinsic and instrumental grounds – and because the prosecutor has the power to make exceptions – the perverse incentives are outweighed by the positive ones.

Given limited resources, the ICC should only prosecute violent international criminals

A related, second, objection is that my argument undercuts the stated purpose of the ICC, namely, to hold to account those who perpetrate the most egregious international crimes. If the ICC prosecuted all nonviolent crimes against humanity, its docket could swell to an unwieldy size, individuals could languish in legal limbo, and states might withdraw their support for the young court. This criticism is weak for several reasons.

First, notice that this objection does not undermine any of my interpretation of the Rome Statute, or the view that it is morally important to prosecute nonviolent international criminals. Second, even as a practical suggestion, it fails because there is no reason to assume that all nonviolent crimes are less deserving of punishment than are violent ones. Why would someone like Qaddafi, who violently killed a few thousand people in 2011, be more deserving of punishment than someone who intentionally starved to death a million people? Because the prosecutor has the power to decide when not to indict alleged international crimes, she can strategically decide who to prosecute and who to not prosecute. Instead of trying all possible international criminals, the prosecutor could indict only those who perpetrate especially egregious violent or nonviolent crimes against humanity, at least for now. Indeed, rather than undermining the purpose of the ICC, my argument bolsters it. Now the ICC lets nonviolent international criminals walk free. If the ICC’s prosecutor adopted my argument, she would be able to expand who she decides to prosecute and indict and try the worst of the worst whether they were violent or nonviolent perpetrators. I acknowledge that the ICC and the prosecutor have scarce resources that must be used carefully, but there are moral reasons to include nonviolent offences in the calculation of who is most deserving of prosecution, and I have shown that the ICC has the legal authority to do this.

Is some of the language in the Rome Statute a design flaw?

A third objection is that the language in Articles 7 and 30 contains a design flaw and should thus be changed. The language is a design flaw in the sense that the authors of the Statute did not intend for it to be interpreted in the way I do, so far as I can tell. The drafters of the Rome Statute, for example, rejected the idea that mass starvation should be included as a crime against humanity (Schabas Reference Schabas2007: 105; von Hebel and Robinson 1999: 103). But the authors of the Rome Statute may have intended such actions as I term nonviolent crimes against humanity to be included under the section on ‘other inhumane acts’. The language is also not a moral flaw because nonviolent crimes against humanity can be as severe and kill as many people as violent ones, and individuals who commit such crimes deserve punishment.

Those who want to remain faithful to the founders’ intentions might suggest that because the Rome Statute can be interpreted in the way I do, it should be edited to exclude nonviolent crimes against humanity. One way to achieve this aim would be to delete section 7.1(k). That would preserve all of the language of the chapeau elements, as well as the majority of underlying acts that constitute a crime against humanity. This and any other modification that would preclude prosecution of nonviolent crimes against humanity should be rejected on moral grounds. It is more important to improve the global accountability of individuals who cause the deaths and harms of great numbers of people than it is to remain faithful to the possible original intent of the state representatives and others who composed the Rome Statute when doing so can remain within the law.

Another legal reason why the language does not need to change is because the current language does not violate the nullum crimen sine lege principle because it does not expand the scope of one of the core crimes of the ICC to an extent that it is unclear how to avoid committing a crime against humanity. The principle of nullum crimen sine lege requires that laws have sufficient specificity to allow any reasonable person to know how to avoid violating them. Although given my argument it is more difficult for powerful individuals to avoid committing a crime against humanity, it is not impossible. That there is an expansion of a law and that it is more difficult to avoid violating it does not mean that it is impossible to avoid committing a crime or that it is impossible to know how to not commit the crime. Legally, leaders can avoid prosecution by refraining from committing the acts that I discuss above. Practically, if democratic leaders act in good faith to avoid committing nonviolent crimes against humanity, the ICC’s prosecutor should avoid indicting these individuals because there are numerous more deserving cases.

The ICC is itself unaccountable

A fourth objection is that because the ICC itself is not sufficiently accountable, it should not have its mandate expanded, lest it more easily abuse its power. John Bolton used a similar argument in an attempt to derail realization of the ICC (Paris Reference Paris2009: 61 and passim). A purportedly powerful institution that is undemocratic and unaccountable to state governments is dangerous because it might indict people for political purposes, some might claim. Another way the objection might be framed is to query who can hold the surrogate accountability holders to account (Rubenstein Reference Rubenstein2007).

There are a few responses to this objection. The main point is that my argument does not alter any of the institutional checks and accountability mechanisms already in place. The ICC has internal accountability mechanisms. The prosecutor can proceed with an investigation only if the Pre-Trial Chamber confirms that there are ‘substantial grounds’ for the charges (Articles 15; 61.5). The Pre-Trial Chamber exercises their power. As of late January 2012, the Pre-Trial Chamber disallowed the prosecutor to proceed with over 25 per cent of investigations (Schabas Reference Schabas2012). Just as the Pre-Trial Chamber checks the power of the prosecutor for violent crimes, so would they vet the prosecutor’s attempt to prosecute nonviolent offences. Another accountability mechanism is that state parties can remove the prosecutor with a simple majority vote for ‘serious misconduct or a serious breach of his or her duties’ (Rome Statute 1998: Article 46; Rodman Reference Rodman2006: 33). A third, external accountability mechanism rests with the UN Security Council, which can postpone indefinitely, subject to yearly renewals, any case for any reason (Article 16).

Consider the problems if the ICC were more responsive and accountable to member states. Being more directly accountable to state governments would make the court even more political, and would not eliminate the criticism of problematic accountability. Look to the critics of the IMF and World Bank for objections to this sort of accountability. Recall the ground-breaking arguments of the Federalists who argue that judicial independence is important for fair trials and the rule of law (Hamilton, Madison, and Jay [1787], sections 78–9). Similarly, some judicial independence is vital for the ICC. Expanding the court’s jurisdiction would not undermine the extant checks and accountability mechanisms already in place. Finally, it is better to try some from a whole class of heretofore mostly unprosecuted miscreants than to try none. The small possibility of the ICC abusing its power does not trump the importance of bringing nonviolent international criminals to justice.

If the ICC is not performing well, why should it expand its powers?

Fifth, someone might object that despite these checks and accountability mechanisms, Ocampo did a poor job and therefore there should be no expansion of the prosecutor’s mandate until Fatou Bensouda proves her merit. Why should we expand the powers of a poorly running institution? First, notice that this does not undermine my moral or legal arguments, but rather suggests the ICC should postpone prosecuting nonviolent crimes against humanity. This objection addresses a whole class of arguments, including mine, that would expand the ICC’s mandate. Individuals who raise such an objection make two assumptions and one implicit argument. First, they assume that the ICC is performing poorly. Second, they assume that the ICC will perform as poorly when it prosecutes nonviolent crimes against humanity as it has in the past, even under a new prosecutor. The argument implicit in this objection is that the ICC is and will perform poorly enough to outweigh the benefits of holding to account individuals who commit nonviolent crimes against humanity. I do not think these assumptions and argument are sustainable. But I attempt to put forward the most convincing components of this objection in order to fairly evaluate them.

One possible way to make this objection is to argue that the prosecutor has made mistakes. For example, Adam Branch suggests the ICC’s arrest warrants against the Lord’s Resistance Army’s (LRA) leadership may have been detrimental to the Ugandan peace process, and possibly even contrary to international law because it is not at all clear that Uganda’s government was either unable or unwilling to try the LRA leadership (Branch Reference Branch2007: 186–7). Furthermore, the ICC tacitly exonerated the Ugandan government, which may have committed abuses punishable under the Rome Statute (Branch Reference Branch2007: 187–8). Another way that the prosecutor exercised his powers poorly was by issuing an open instead of a sealed arrest warrant for Sudan’s president even though Bashir had threatened publicly to expel aid organizations if he were to be indicted. Bashir’s expulsion of aid organizations very likely cost lives. A third charge is neocolonialism. As Sriram puts it, this claim is ‘difficult to source, because while it is a concern that has been aired in conferences and in diplomatic circles, it is seldom articulated fully or in print’ (Sriram 2010: 320). The idea is that because the court is located in the Netherlands and has so far indicted and prosecuted only Africans, the ICC is a neocolonialist institution.

These criticisms are generally reasons why the prosecutor should be more politically astute, and I admit that Ocampo made mistakes. Ocampo should have investigated and possibly indicted Ugandan government officials and issued a sealed warrant for Bashir. But these objections do not undermine my argument for the expansion of crimes against humanity to nonviolent offences for the following reasons. First, trying some alleged international criminals is better than trying none, and trying more is generally better than trying fewer. Adopting my argument would allow the ICC to bring to trial more individuals suspected of unjustly killing and harming multitudes. Second, it would allow for a wider range of individuals to be brought to trial, including those from outside Africa, and also for the prosecutor to avoid the charge of neocolonialism. The charge of neocolonialism is somewhat confusing for another reason. One of the main problems of colonialism was that colonial authorities were unaccountable and hence were able to get away with harming and killing masses of people to further their own ends. The ICC in general and my argument in particular aim to close a gap in accountability. What is more, the ICC can exercise its jurisdiction only if a state is unable or unwilling to prosecute, preserving a large degree of state sovereignty. It is difficult to sustain the claim that an institution is neocolonialist if the vast majority of those victimized by an alleged international criminal indicted by the ICC wants the accused prosecuted at the ICC. That is exactly what occurred in one case where individuals were polled. Some 98 per cent of Darfuri refugees in Chad wanted Bashir tried at the ICC (Hamilton Reference Hamilton2011: 160). Third, and perhaps most importantly, none of these arguments even attempt to trump the idea that holding to account individuals who commit international crimes is important. Another way to put this is to imagine the alternative: not holding individuals to account who commit nonviolent crimes against humanity is unjustifiable. Despite these criticisms – which are healthy and which will accompany any institution, especially a new one that challenges inveterate ideas and practices about law and state sovereignty – the ICC is working to a degree that expanding its mandate is morally justifiable. Alleged international criminals are behind bars, trials are proceeding, and the court is not in imminent danger of collapse.

My argument dilutes crimes against humanity

Sixth, because the concept of a crime against humanity is such a serious claim, some may believe that expanding its legal or moral scope must necessarily and problematically dilute it. I do not think this is the case for two reasons. First, theoretically violent harms are no worse than nonviolent ones. Indeed nonviolent harms may result in more severe suffering than some violent ones. Imagine someone who has led a relatively healthy life, with adequate opportunity, a fine education, comfortable housing, adequate health care, and so on until they were murdered or expelled from their home, say in ethnic cleansing in the Balkans. I see no reason why that person suffered more than someone in Chad who had poor health most of her life, who was malnourished, and who finally succumbed to a treatable disease because a kleptocratic president stole the money that was earmarked to fund a food security programme and local health clinic. I do not want or need to claim one scenario is worse than the other. I want to suggest only that the nonviolent death is sufficiently morally abhorrent, and that the ICC’s prosecutor has the authority, to investigate high-ranking Chadian officials on charges of crimes against humanity.

Second, if someone wants to term the crime something different than ‘a crime against humanity’, and objects to the language of the Rome Statute, I might agree: the category of crimes against humanity is theoretically suspect. I do not have space to develop this argument, but I will briefly note a key point Altman (Reference Altman2006) makes in an article titled ‘The Persistent Fiction of Harm to Humanity’. Altman writes, ‘The vital interests of most humans are simply not at stake’, when crimes against humanity are committed (Altman Reference Altman2006: 371). As some define it, the concept of harm has to be stretched beyond useful meaning to support the theoretical, as opposed to legal, concept of crime against humanity (see May 2005, 2006a, 2006b; Luban Reference Luban2004, Reference Luban2006; Mayerfeld Reference Mayerfeld2006; Vernon Reference Vernon2002). My argument does not weaken the moral foundations of a crime against humanity even while it allows room to reformulating the concept of a crime against humanity.

Pogge’s explanatory nationalist critique

Seventh, Pogge might object that holding to account only national and nonstate actors overlooks those international leaders who sustain global economic and political orders that are probably morally and legally liable according to my argument because the global economic order sustains severe poverty from which 18 million people needlessly die every year (Davidsson Reference Davidsson2005; Øverland 2011; Pogge Reference Pogge2008: 2 and passim). Pogge might suggest that my analysis is overly nationalistic and does not sufficiently consider international actors who perpetuate the global property rights regimes that he argues harms the poor. I accept that the global institutional economic order, especially concerning property and borrowing rights, is problematic and needs modification (Pogge Reference Pogge2008: 118–22). Pogge might suggest that international actors could be tried for other harms as well. For instance, UN-imposed and US-backed sanctions against Iraq in the 1990s caused the deaths of half a million innocents (Gordon Reference Gordon2006: 87; Gordon Reference Gordon2010). Pogge’s objection is not an argument against my thesis that state and nonstate actors should be held to account for nonviolent crimes against humanity. Pogge’s objection is an argument to further expand the reach of my argument.

My argument problematically undermines state sovereignty

Eighth, some might object to my attempt to expand global legal accountability by arguing that states are sovereign and therefore state leaders should not be accountable to anyone other than their citizens. Due to limitations of space, I cannot adequately defend my views here on why states should have conditional and limited sovereignty. States leaders themselves and scholars have argued and come to accept that state sovereignty should not be absolute. For instance, a broad range of scholars have advanced a diverse set of views grounded in collective self-determination and human rights for why an absolutist interpretation of Westphalian state sovereignty is morally unsupportable (Pogge Reference Pogge1992; Walzer Reference Walzer1977; Walzer Reference Walzer1980; Beitz Reference Beitz1980; Beitz Reference Beitz1999; Beitz Reference Beitz2009a; Luban Reference Luban1980a; Luban Reference Luban1980b; Doyle Reference Doyle2009). State leaders themselves have agreed to many of these limitations to their sovereignty through international laws, universal human rights, and international institutions. It is important to note that the ICC exists only because states ratified the Rome Statute. Now, more than half of all states have ratified it. In 2005, state leaders themselves even accepted the ‘responsibility to protect’ which allows foreign military intervention as a last resort if states are unable or unwilling to protect their citizens from mass atrocities (ICISS 2001; United Nations 2005; Evans Reference Evans2008; Bellamy Reference Bellamy2009). In one sense, this objection is lifeless because states have already agreed to the Rome Statute and my argument merely argues that the ICC should try a category of crime that it has avoided prosecuting so far so state sovereignty is no weaker now than 1 July, 2002, when the Rome Statute entered into force. A stronger way to rebut this objection is to again weigh the relative moral values of state sovereignty and holding to account individuals that slaughter nonviolently large numbers of people. I have attempted to show that such harms are on par with widely accepted violent international crimes and concomitant infringements of state sovereignty. Crimes against humanity are explicitly actions that states have agreed limit their sovereignty in the nonbinding 2005 World Summit Outcome Document (United Nations 2005: paras 138–139). Even if my argument does undermine somewhat current conceptions of state sovereignty, it is legally permissible and morally essential.