Part II of this comment will give an overview of how the causation element of command responsibility has been applied in prior international criminal tribunals, including the Bemba case. Part III will offer prescriptions for how the ICC and future tribunals should deal with the causation problem moving forward.

This comment seeks to answer two distinct questions: Is causation a requirement of command responsibility pursuant to Article 28? If so, due to the lack of consensus on the issue, what degree of causation should be required in order to prove this element?

To the first question, I argue that due to the failure of the Appeals Chamber to reach a consensus on the causation element of Article 28, the issue remains very much unresolved. While nine of the eleven judges who reviewed the case concluded that causation is an distinct element under Article 28, the two judges who do not happen to be in the majority of the Appeals Chamber.10 Even so, the third judge in the majority, Judge Eboe-Osuji, argued unambiguously that causation is an element.11 Thus, it appears likely that future Judges in future command responsibility cases before the ICC will follow this conclusion.

To the second question, I argue that the standard adopted by the Pre-Trial and Trial Chambers is manifestly unfair to the accused as it places a severely low burden of proof on the Prosecution. The Court must require the Prosecution to prove beyond a reasonable doubt that the criminal acts of the subordinates were in some way influenced by the failure of the superior to effectively control their troops or their failure to submit prior offenses to a competent authority. Inherent in such a standard would be limitations on sort of remoteness considerations, both geographically and temporally. For example, it would be difficult to show that a commander’s omissions five years prior influenced a subordinate’s actions in the present. Likewise, it will be equally difficult to demonstrate that a commander’s failure to punish a subordinate in Kenya influenced the actions of a soldier in the Congo, without more.

II. Causation in International Law Jurisprudence

A. Causation in Command Responsibility Before Bemba

The theory of that a commander could be held legally responsible for the crimes of their subordinates has been around since the mid 19th century, though there was no mechanism of international enforcement of this concept until much later as that was generally the primary responsibility of domestic legal systems.12 Indeed, the first codified international treaty agreement recognizing the existence of a causal connection between a commander’s failure to exercise her duty to control her subordinates and the commission of war crimes by those subordinates was in 1977 adoption of Article 86(2) of Additional Protocol I to the Geneva Conventions.13 This stated in relevant part:

The fact that a breach of the Conventions or of this Protocol was committed by a subordinate does not absolve his superiors from penal or disciplinary responsibility, as the case may be, if they knew, or had information which should have enabled them to conclude in the circumstances at the time, that he was committing or was going to commit such a breach and if they did not take all feasible measures within their power to prevent or repress the breach.14

Article 86(2), while unmistakably establishing that the omissions of a superior could result in the commission of crimes by subordinates, does not go so far as to require causation as an entirely separate element of proof in the prosecution of such crimes. Further, as will be discussed below with regard to the opinion of Judge Eboe-Osuji, it does not state, as Article 28 of the Rome State does, that a superior will be held responsible for the crimes of their subordinates.15 Rather, it implies that failure to prevent or repress a breach of the Protocol by a subordinate is an entirely separate offense.

The language of Additional Protocol I was adopted largely verbatim by the statutes of the International Criminal Tribunal for the Former Yugoslavia (ICTY)16 and the International Criminal Tribunal for Rwanda (ICTR).17 For example, Article 7(3) of the ICTY Statute reads:

The fact that any of the acts referred to in articles 2 to 5 of the present Statute was committed by a subordinate does not relieve his superior of criminal responsibility if he knew or had reason to know that the subordinate was about to commit such acts or had done so and the superior failed to take the necessary and reasonable measures to prevent such acts or to punish the perpetrators thereof.18

This language again accepts that a commander’s failure to take necessary and reasonable preventative measures can have a causal effect on a subordinate’s ability or willingness to commit a crime. Yet, unlike the wording of Additional Protocol I, the ICTY statute distinctly separates the commander’s failure to punish and the commander’s failure to prevent. Article 86(2) speaks only of a commander’s knowledge as it related to crimes being committed or about to be committed. Article 7(3), on the other hand, relates to crimes the subordinate “was about to commit… or had done so.”19 This is an important distinction and lends credence to the idea that command responsibility, at least as it relates to failure to punish, under the ICTY and ICTR statutes was intended to criminalize superiors under a separate offense and not for the actual criminal acts of the subordinates.

The leading ICTY case on command responsibility is Prosecutor v. Delalić (Čelebići ).20 The Trial Chamber in Čelebići, while recognizing “the central place assumed by the principle of causation in criminal law,” argued that it found:

[N]o support for the existence of a requirement of proof of causation as a separate element of superior responsibility, either in the existing body of case law, the formulation of the principle in existing treaty law, or, with one exception, in the abundant literature on this subject.21

The Court further argued that because a causal link cannot possibly exist between an offense committed by a subordinate and the subsequent failure of a superior to punish the perpetrator of that same offense, no requirement of causality could ever exist as part of the doctrine of command responsibility.22 Essentially, the Čelebići Court was stating that if causation cannot exist in the failure to punish branch of command responsibility, then it cannot also exist in the failure to prevent branch because Article 7(3) does not distinguish between the two with regard to elements of proof.23

Logically speaking, the Court is correct in that a commander’s failure to punish a subordinate for a crime that has already been committed could not be said to have caused that crime. Accordingly, it would be both unnatural and unjust to punish the commander as a principal or accomplice for a crime that they neither participated in nor could have prevented. To be sure, the argument has been advanced that punishment as principal under such circumstances would be permissible under the theory that the commander’s failure to punish the crime is evidence of her tacit approval of such crimes and, therefore, she shared the mens rea of the underlying actor.24 This approach has some limited acceptance in international criminal jurisprudence. In Prosecutor v. Sawada before the Tokyo Tribunal, Lieutenant-General Sawada was held to have “ratified” crimes that were committed by his subordinates while he was away and that he failed to punish upon his return.25 Additionally, the Trial Chamber in the Halilović case reasoned that the tacit acceptance argument as “not without merit,” because a failure to punish is “so grave that international law imposes upon him responsibility for those crimes.”26 Yet, this is a rather strained reading of the statute as it seeks to infer and attach a mental state to an individual without any additional proof. Regardless, it is not the position adopted by the ICTY.

The flaw in this reasoning is evident and, interestingly, the Čelebići Court even seems to acknowledge this. Indeed, the Court reasons that a causal connection between a commander’s failure to punish past crimes and the commission of future crimes is “not only possible but likely.” But the Court seems to view this causal connection not as a failure to punish but as a failure to prevent. The commander’s lack of punishment, in this sense, should be viewed as a preventative measure that could have been taken to avoid the commission of future crimes. This conflation of the two branches of command responsibility misses the point. As Professor Robinson notes:

‘[F]ailure to prevent’ and ‘failure to punish’ are not redundant; they offer two distinct ways to prove the failure of the commander. A prosecutor may prove either inadequate preventative measures or inadequate efforts to investigate and prosecute crimes. Either provides the dereliction that, if accompanied by a blameworthy state of mind and a contribution to crimes, can ground accomplice liability for resulting crimes. Thus, respecting the contribution requirement does not render the ‘failure to punish’ branch superfluous.27

The Čelebići Court could have easily deduced a causality requirement in Article 7(3) while still recognizing that it cannot possibly exist with regard to past crimes, but only to future crimes. Nevertheless, the approach adopted in Čelebići became the dominant position for future ICTY and ICTR Courts.28

B. Competing Interpretations of the Bemba Court

1. Pre-Trial and Trial Chambers

i. Majority Opinions

The Pre-Trial and Trial Chamber opinions take a strong anti-accused position by essentially adopting the approach pushed for by the Prosecution and the amicus curiae brief of Amnesty International.29 The opinion states that:

[T]o find a military commander or person acting as a military commander responsible for the crimes committed by his forces, the Prosecutor must demonstrate that his failure to exercise his duty to prevent crimes increased the risk that the forces would commit these crimes.30

The opinion does not elaborate further on how much the risk would need to be increased; however, the decision argues that because:

[P]unishment is an inherent part of the prevention of future crimes, a commander’s past failure to punish crimes is likely to increase the risk that further crimes will be committed in the future.31

Thus, the standard adopted by the Pre-trial chamber would allow the Prosecutor to obtain a conviction under Article 28 simply by demonstrating that:

1. a subordinate committed a crime;

2. the commander did not punish that crime; and

3. a future crime was committed by either the same or another subordinate.

Although the opinions of the Pre-Trial and Trial Chambers explicitly contradict ICTY case law on command responsibility by affirming that causality must exist between the superior’s failure to control her troops and the subordinate’s underlying crimes, that is about as far as they are willing to go. Indeed, the “risk aggravation” theory, according to the Pre-Trial Chamber, states that “there is no direct causal link that needs to be established”32 between the failure to punish or prevent and the underlying criminal acts. On its face, the internal contradiction in the reasoning that causality is a necessary element of command responsibility but no direct causal link between the omission and the crime needs to be established appears hard to reconcile. Yet, what the Pre-Trial and Trial Chambers appear to be saying is that causality can be inferred from the increased risk because that increased risk must have been at least one of the causes of the underlying crime.

ii. Separate Opinion of Judge Steiner

Clearly unpersuaded by the extraordinarily weak standard of causation proposed by the Pre-trial Chamber and the majority opinion above, Judge Steiner attempts to find a middle ground by arguing that:

[T]he causality requirement would be satisfied where, at least, there is a high probability that, had the commander discharged his duties, the crime would have been prevented or it would have not been committed by the forces in the manner it was committed.33

Recognizing that the position adopted by the majority would essentially turn command responsibility into a strict liability crime, Judge Steiner seeks to raise the threshold to force the Prosecution to produce evidence of an actual connection between the commander’s failure to control and the underlying crime rather than simply showing the former.

This is a notable step in favor of the accused, but it still does not provide a clear, enforceable standard. Primarily, Judge Steiner does nothing to clarify what the difference between the “increased risk” and “high probability” standards are. The opinion states that the “high probability” standard is something below the position argued for by the defense: that that crimes be “(i) the ‘certain’ consequence of the omission [of the commander]; or (ii) ‘a direct and reasonably foreseeable’ consequence of the omission.”;34 however, Judge Steiner does not define what, exactly, that difference is.

2. Appeals Chamber

The majority opinion of the Appeals Chamber does not discuss nor make any conclusive determination on the issue of causation. The opinion only states that Article 28 is not a form of strict liability,35 implying that the majority believes there is a causation requirement but fails to clarify further. Nonetheless, because the majority opinion does not elaborate on the causality problem, this section will focus only on the varying interpretations of the two separate opinions and the dissenting opinion.

i. Separate Opinion of Judges Van den Wyngaert and Morrison

This opinion is by far the most deferential to the accused. Judges Van den Wyngaert and Morrison suggest that proximity of command is paramount in determining whether the acts of a superior could have a causal effect on the crimes committed by subordinates.36 They reason the because superior commanders—those at the highest level of military command—cannot possibly be expected to monitor the day to day activities of the troops on the ground, they cannot be said to have “effective control” of those troops.37 Thus, since this essential element of Article 28 cannot be established, it will be nearly impossible for a prosecutor to prove that a superior commander is liable for the crimes of those at the bottom or the military hierarchy.

With regard to punishing subordinates for crimes they have committed or referring their criminal behavior to competent authorities, Judges Van den Wyngaert and Morrison state that failure to do this “can only be said to have a causal effect on subsequent crimes in very exceptional circumstances.”38 In order for the Prosecutor to convict on these grounds, she would have to show beyond a reasonable doubt:

[T]hat the perpetrators of the subsequent crimes knew about the initial crimes, knew that the commander was aware of this crime and that he or she deliberately failed to take any measures to have the perpetrator held to account, and that this played a significant role in their decision to commit crimes themselves.39

This is a nearly impossible standard to meet.

The opinion of Judges Van den Wyngaert and Morrison, if applied in future cases, would create a situation whereby the highest level of military commanders could essentially never be held liable for the crimes committed by ground level troops and, even in situations where a military commander has direct control, the prosecutor would have to gather evidence about the actual knowledge and mental state of both the superior and the subordinate. This is clearly unworkable and would essentially mean the end of the doctrine of command responsibility in international criminal law.

ii. Separate Opinion of Eboe-Osuji

Judge Eboe-Osuji argues that the “result of” element of Article 28 needs to be analyzed through careful observation of the purpose, language and scope of the Rome Statute itself.40 Judge Eboe-Osuji notes that causation is indeed and essential element of command responsibility under Article 28.41 The bulk of the opinion, however, deals with the causation threshold and the burden of proof on the prosecution to establish this element. Judge Eboe-Osuji states:

What is required is merely that the defendant’s ‘contribution must be more than negligible or not to be so minute that it will be ignored under the ‘de minimis’ principle.42

Yet, the opinion goes on to argue that because of the unique nature of the military commander during times of war:

[T]he manner in which the opportunity [for the subordinate to commit the crime] was afforded goes well beyond merely ‘permitting’ or ‘not preventing’ the forbidden harm—it indeed involves ‘setting in motion’ a chain of tangible events (notably training, arming and deploying the subordinates), which culminated in the resulting harm.43

Thus, a commander who “enables the prospects of harm by creating the danger of their occurrence”44 can and should be held liable for the underlying offenses committed by her subordinates.

Judge Eboe-Osuji reaches this conclusion by comparing the two main competing theories of command responsibility: dereliction of duty and complicity. While dereliction of duty is a sui generis crime, complicity is a theory of liability holding a commander responsible as a principal for the underlying crimes committed by the subordinate. The prevailing jurisprudence of the ad hoc tribunals held that dereliction of duty was the proper lens through which to assess the criminal responsibility of commanders.45 These ad hoc tribunals traced their reasoning to Article 86(2) of Additional Protocol I to the Geneva Conventions.46

Although Article 86(2) does not clearly differentiate between dereliction of duty and complicity as the proper standard for command responsibility, the difference in language is clear. Whereas Article 28 of the Rome Statute states that a commander will be held responsible “for” the crimes committed by their subordinates, Article 86(2) merely states that superiors will not be absolved “from penal or disciplinary responsibility, as the case may be.”47 As Judge Eboe-Osuji notes, the tendency of the judges in the ad hoc tribunals to learn towards the more lenient dereliction of duty standard makes sense: there is obvious discomfort in punishing an individual for crimes committed by others when that individual neither participated in the actus reus nor shared the mens rea.48 Nonetheless, the language of Article 28 is relatively clear as the insertion of the word “for” indicates that the drafters intended to hold commanders responsible as accomplices to their subordinates and not for a separate crime. Moreover, Article 28 states that the crimes for which the commander is prosecuted must be “within the jurisdiction of the court.” Dereliction of duty—or some analogous crime—is not one of the crimes enumerated in Article 5. To that point, there is a serious problem of pleading as Bemba was tried and convicted not for dereliction of duty, but for war crimes and crimes against humanity.49 Thus, as Judge Eboe-Osuji reasons, accomplice liability is the proper lens through which to view command responsibility.

Because Article 28 purports to hold a commander responsible as an accomplice for crimes they did not participate in, Judge Eboe-Osuji argues that proving the causal connection places:

[An] exacting burden upon the Prosecution to prove the indicated links of complicity between the subordinate’s crime and the commander’s prior connivance or condonation…Failure to discharge the burden of proof on the criminal standard will result in an acquittal, as is the case with the present appeal.50

Indeed, Judge Eboe-Osuji very clearly states that the causality requirement would be met when:

(1) Where it is shown that the commander really had effective control over the subordinates, but wilfully declined or abstained from exercising such effective control, resulting in failure to prevent or repress the crime; or that he wilfully failed to punish or prompt the punishment of rogue subordinates in a manner that emboldened them subsequently to commit the same or kindred crimes that the commander voluntarily failed to punish on a previous occasion. As part of this condition, it must be shown that the failure to exercise proper control (which is the commander’s indicted failing) is attributable to the commander’s wilful abstention from exercising effective control that was within his ability at all material times: it may not be enough that the failure to exercise effective control is merely symptomatic of his real lack of effective control in the first place (except possibly in circumstances where the doctrine of endangerment may prevent him from escaping criminal responsibility, if he created the danger in the first place, but failed to put commensurate measures in place ahead of time to control the danger); (2) Where it is shown that the crimes committed after the failings indicated above are the demonstrable ‘result of’ those failures. Indeed, this is in the sense that those failings had encouraged the crimes that had been committed;51

The opinion goes on to clarify that part (2) of this test must be shown “beyond a reasonable doubt.”52

This test is interesting because it does, as indicated, provide for an exacting burden on the Prosecution, but it does not provide clarity on the issue of causation specifically. Rather, it sets an extraordinarily high bar for the prosecution to prove with regard to the mens rea of the commander, but not to the subsequent causation. Once the former is established, proof of the latter appears to be much easier. It will be very hard for the prosecutor to demonstrate beyond a reasonable doubt that the commander willfully abstained from her duties. Assuming a commander is willfully failing to control her troops, it follows that the causal connection between such failure and the underlying offenses of subordinates will be easier to establish because there is likely a widespread culture of impunity. In the end, while Judge Eboe-Osuji does provide a thorough analysis of the causation element, his ultimate holding does little to clarify what type of proof is needed for the Prosecutor to demonstrate causation beyond a reasonable doubt.

iii. Dissenting Opinion

The dissenting opinion follows the exact line given by Judge Steiner in her separate opinion at the Trial Chamber. The opinion begins by analyzing whether the “result of” element of Article 28 requires a showing that the superior’s failure to exercise control properly caused the commission of crimes by her subordinates.53 In finding that it does, the dissent relies on three primary reasons.

First, the fact that, unlike other provisions of Article 28 of the Rome Statute, explicitly calls for a causation requirement by stating that the crimes of the subordinates be committed “as a result of” the commander’s failure to exercise control properly, strongly indicates the intention of the drafters to make causality a distinct element.54 Second, holding a superior “’criminally responsible for the crimes within the jurisdiction of the Court’ committed by his or her subordinates… without causation would be incompatible with the culpability principle.”55 (Dissenting opinion at ¶ 333). The ICTY Appeals Chamber in Tadić appropriately described that principle as follows:

The basic assumption must be that in international law as much as in national systems, the foundation of criminal responsibility is the principle of personal culpability: nobody may be held criminally responsible for acts or transactions in which he has not personally engaged or in some other way participated (nulla poena sine culpa).56

Thus, the dissenting opinion argues, it would be irreconcilable with basic tenets of criminal law to not require a personal nexus to the underlying offenses. Finally, the addition of the “result of” element in Article 28, when it was not present in prior international criminal statutes, strongly favors the interpretation that the drafters of the Rome Statute believed the ICTY and ICTR statutes did not go far enough.

The opinion then moves on to determine what the causality threshold should be. Similar to the Pre-Trial and Trial Chambers, the dissent rejects the applicability of a strict “but-for” test. The opinion argues that the nature and complexity of the cases before the ICC are such that it would almost always be possible to construct a hypothetical scenario in which, despite the best efforts of the commander, the crimes still would have been committed because of some intervening event. Thus, it follows that, if such a test were applied, no commander would ever be convicted pursuant to Article 28. In the end, the dissent adopts the view of Judge Steiner’s “high probability” test because it will ensure that “responsibility only arises when there is a demonstrably close link between the commander’s omission and the crimes committed by his or her troops.” Likewise, it will avoid acquittal on the grounds of possible, yet highly unlikely hypothetical scenarios that could have also caused the crimes to be committed.

III. Where Does the Issue Stand Now and Where Should It Go From Here?

As is evident from the above discussion, the legal standard for causation in the context of command responsibility is far from settled. Opinions on the issue range from no contribution requirement at all to a highly stringent “but-for” test. Due to the extraordinary lack of consensus, the logical question becomes: what standard should the ICC employ moving forward? The answer to this question depends on how much precedential weight should be given to each of the competing separate majority opinions versus that of the dissention opinions, which is more consistent with both the Pre-Trial and Trial Chambers. A more careful analysis on the precedential value of split decisions on the ICC is the topic of an entirely separate paper.

For the purposes of this comment I will first try to distill a common thread from each of the opinions discussed above, giving the greatest weight to those in the majority of the Appeal Chamber. From there I will discuss potential flaws in the positions taken by the Appeals Chamber and suggest more desirable alternatives. I argue that the position taken by Judge Eboe-Osuji provides the clearest and most reasonable solution to the causation problem but that it still leaves unanswered major questions such as the factors to be taken into account by the court and the burden of proof required by the Prosecutor to satisfy each.

A. Where the Court is Now

The lack of consensus at the Appeals Chamber on the issue of causation is concerning as it leaves an important area of law wholly unresolved. Complicating this is the fact that, even among the majority, opinions differed as to the element of causation. Even so, taken together, the overwhelming majority of judges at the ICC support the notion that causation is a distinct element of command responsibility that needs to be proven. Both the Pre-Trial and Trial Chambers unanimously agreed that causation is required under Article 28. Additionally, three of the five judges in the Appeals Chamber agreed on the same issue. The problem, of course, is that two of those three judges wrote the dissenting opinion. Further muddying the waters is that the lone opinion in the majority supporting the notion of causality fundamentally differs on the standard that should be applied from the Dissenting Opinion and both the Pre-Trial and Trial Chambers, each of which are basically in total agreement.

The result is a very narrow holding that will leave a lot of work for future Courts to unravel.57 Ultimately, it is highly likely that causation will be found to be an essential element of command responsibility under Article 28. Nine of the eleven judges that reviewed the case found in favor of causation, and nearly every scholar that has written on the issue has concluded the same. Indeed, the Separate Opinion of Judges Van den Wyngaert and Morrison represents a startling departure from nearly unanimous legal consensus on this issue. Moreover, for an opinion that departs so drastically from the positions set out by the Pre-Trial Chamber, Trial Chamber and even Judge Eboe-Osuji, it cites to almost no authority, either judicial or scholarly, to support its conclusion. Thus, judges presiding over future cases before the ICC will more likely than not conclude that causation is a required element. The question of what the causation threshold should be, however, remains uncertain.

B. Where Should the Court Go From Here?

To answer the question of what the causation standard should be moving forward, it is necessary to apply the principles of interpretation set out in Articles 31 et seq. of the Vienna Convention on the Law of Treaties.58 As such, the language of the statute must be read in the ordinary meaning of the terms of the treaty “in their context and in the light of its object and purpose.”59 This section will first analyze whether Article 28 does, in fact, require proof of causation as a distinct element of command responsibility. Next, this section will attempt to determine what the appropriate causation threshold should be by evaluating the merits of each of the various tests proscribed by the Judges above. It should be noted that because I ultimately conclude that causation will be found to be a distinct element of command responsibility under Article 28, I will not consider the Separate Opinion of Judges Van den Wyngaert and Morrison in this analysis as they believe it is not.

1. Is Causation a Distinct Element?

The Rome Statute, along with the statutes for the prior ad hoc international criminal tribunals, was designed to put an end to impunity for the perpetrators of unimaginable atrocities that deeply shock the conscience of humanity. (Eboe-Osuji at ¶ 244). Indeed, the preamble to the Rome Statute includes a cross-reference to the purposes of the UN Charter, the first of which is “to save succeeding generations from the scourge of war.” Thus, from a policy perspective, it is easy to see the purpose of the Rome Statute as aimed at protecting innocent civilians from the extreme dangers of warfare committed, or contributed to, by persons in positions of command. The significance of that policy must comprise questions about the criminal responsibility of those implicated in creating the danger of such ravages of armed conflicts upon innocent civilians.

Similarly, as noted by several of the opinions, command responsibility cases before the ICC are unique in that they are often:

[L]arge scale and complex by nature, often requiring an assessment of the culpability of individuals who are physically or geographically remote from the scene of the crimes, and frequently involving many intervening actors.60

The conflict between these competing realities is obvious. On the one hand, the international community wants to bring justice to the victims of the world’s worst crimes. On the other, warfare is inherently a:

[M]ortally dangerous transaction with limited social utility… [involving] civilian victims who are the veritable grass trampled upon by dueling elephants.61

Even an army guided by the most circumspect of commanders can cause significant civilian collateral damage. The law, while recognizing that the worst impulses of war must be punished and deterred, should not be too quick to impose liability on those who may truly have been acting in good faith. Accordingly, the ICC must abide by the core principle of criminal law:

[I]ndividual criminal responsibility should only attach to an accused where there is some form of personal nexus to the crime.62

The final language and drafting history of Article 28 also both strongly support the notion that it was intended to include a personal nexus requirement between the accused and the underlying crime. As noted above, the prior ad hoc tribunals rejected a causality requirement for command responsibility. Yet, those tribunals were acting under a different statutory framework. Neither the ICTY nor ICTR statutes included any language requiring proof of causality. The inclusion of the phrase “as a result of” in Article 28 takes on important meaning when viewed against the backdrop of prior command responsibility jurisprudence. The travaux préparatoires support this.63 It seems evident from the summary of negotiations that the State Parties intended to make the causality requirement explicit in Article 28, unlike the language of the prior international tribunal statutes. Moreover, unlike other provisions of the Rome Statute, Article 28 explicitly states the nexus requirement through the addition of the “result of” phrase. Thus, in accordance with the vast majority of the Judges in Bemba and scholarship on the subject,64 it seems evident that causation is a required element of Article 28.

2. What Is the Appropriate Causation Threshold?

i. Risk Aggravation and High Probability Tests

These two tests, adopted by the Pre-Trial and Trial Chambers, along with the separate opinion of Judge Steiner and the Dissenting opinion in Appeals Chamber, respectively, are the two most pro-Prosecution standards. I analyze them together because they are essentially the same test. The “high probability” standard adopted by Judge Steiner nominally appears to raise the causality threshold from the simply “risk aggravation” test, but neither of the two opinions supporting this test do anything to clarify what separates the two.

This interpretation of the causality element represents a natural progression in command responsibility jurisprudence given the obvious semantic differences between the statutes of the ad hoc tribunals and the Rome Statute. The insertion of the phrase “as a result of” in Article 28 clearly implies that the drafters intended to depart from ICTY jurisprudence holding that causation is not a distinct element of command responsibility. Nonetheless, it is likewise evident that the Pre-Trial and Trial Chambers were uncomfortable leaping too far in the opposite direction by adopting a strict “but for” causality test. To be sure, the “risk aggravation” test does comport with the plain language of Article 28, but the rationale provided for such a test in these opinions leaves much to be desired.

First, in dismissing the enforceability of a “but for” causality test, the opinions both overstate the predictability of positive acts and understate the predictability of omissions. The Pre-Trial Chambers argues:

[C]ontrary to the visible and material effect of a positive act, the effect of an omission cannot be empirically determined with certainty. In other words, it would not be practical to predict exactly what would have happened if a commander had fulfilled his obligation to prevent crimes.65

From a naturalistic perspective, it is true that an omission cannot possess “causal energy” in that the failure to commit an act will not set in motion a series of future events.66 For example, a car will not start without the positive act of sticking the key in the ignition and turning it. But such a perspective assumes a world at absolute zero, where no molecular motion exists until some intervening actor actively changes the status quo. The reality of the world is that things are constantly in motion and the outcomes of both acts and omissions are inherently impossible to predict with certainty. It is true that the act of shooting an individual at point-blank range will almost certainly cause a deterioration in health. Yet, it is also equally true that a soccer player who fails to shoot will certainly never score a goal. Thus, while a “but for” causality test may be extraordinarily difficult to apply to a wartime scenario with thousands of potential intervening actors, dismissing that test should not be grounded in the reasoning that the causal effect of omissions is too difficult to ascertain or prove beyond a reasonable doubt.

Second, the test adopted by the Pre-Trial and Trial Chambers purports to require the Prosecutor to prove that the superior’s omissions “increased the risk” of the subordinate committing the underlying offense, but the opinion’s dicta does all the Prosecutor’s work for her. By conceding that a failure to punish past crimes will likely increase the risk of future crimes, the Pre-Trial and Trial Chambers are essentially establishing strict liability for any commander that fails to punish crimes committed by their subordinates. According to this reasoning, it seems immaterial whether the subordinate was actually motivated by the failure to punish or even that the subordinate actually knew about the prior unpunished crime. Thus, such a standard, while nominally creating a causation requirement, does not really go further than that utilized by the ad hoc tribunals.

Lastly, and perhaps most importantly, a standard that holds a commander liable for increasing the risk of crime even slightly is manifestly unfair to the accused. Certainly, victim protection is one of the primary purposes of the ICC, but it need not come at the expense of fundamental principles of criminal law. The increased risk standard allows for the conviction of a superior for the crimes of her subordinates without any showing of direct causal effect on that crime. The superior neither needs to possess the mens rea of the subordinate nor even encourage the actus reus to be held liable as an accomplice. Further, the Pre-Trial and Trial Chambers place no limit, either temporal or geographic, on the distance between the failure to punish and the underlying crime. Under such a standard, at least theoretically, a superior could be held liable for a crime committed by a subordinate that was not even under their control when the first crime went unpunished. Thus, even though victim protection is a laudable goal, superiors should not be held responsible for the underlying criminal acts of their subordinates without some showing of direct causal impact of their failure to control and those acts.

ii. Causation Beyond a Reasonable Doubt After Willful Failure to Control by a Commander

As noted above, the opinion of Judge Eboe-Osuji provides the most well-reasoned analysis of the causation issue of any of the Judges that presided over the Bemba case. Indeed, I believe Judge Eboe-Osuji was correct in his conclusion that Article 28 constitutes a form of accomplice liability for commanders and, as such, should require a close causal connection between the commander’s actions or omissions and the underlying offense. However, I think the test he provides is unreasonably strict in terms of mens rea requirements and unfortunately unspecific in terms of causality requirements.

With regard to mens rea, which Judge Eboe-Osuji says needs to be shown as a condition precedent to even getting to the causation element, the willfulness standard he provides sets an extraordinarily high bar on the prosecutor. Gathering evidence that shows a commander acted willfully in disregarding their duties pursuant to Article 28 will be very hard to do. Rarely will there be any documented record of a commander explicitly refusing to punish a subordinate, purposefully failing to train her troops or actively encouraging a subordinate to commit a war crime. If such evidence did exist, the case for the Prosecutor would be easy to satisfy regardless of the high mens rea standard.

More importantly, the proposed willfulness standard does not seem to fit with the language of Article 28. The only mens rea requirement mentioned is in Article 28(a)(i), which states that a commander’s failure to control only creates liability with respect to crimes they either “knew or, owing to the circumstances at the time, should have known”67 were being committed or about to be committed. Nowhere in the Statute is willful or deliberate ignorance of crimes mentioned as a requirement for liability. Thus, it seems that the standard proposed by Judge Eboe-Osuji is higher than that required by the Statute and would make prosecution under Article 28 exceedingly difficult.

Even if the Prosecutor could overcome the mens rea burden, proving causation under Judge Eboe-Osuji’s test would be another challenge. I think that Judge Eboe-Osuji is correct in that causation should be proven beyond a reasonable doubt, but he does not give much indication as to what factors will be weighed in order for the Prosecution to meet that burden. The opinion does mention remoteness and temporal proximity as potential factors that would influence causality, but he does not give either much weight.68 Indeed, he suggests that there is no statute of limitations on international crimes so temporal proximity between the failure to punish and the underlying offense plays almost no role in the possible causal connection with the two.69 Thus, there is no clear indication of how the Prosecutor would be able to convince Judge Eboe-Osuji that causation has been established beyond a reasonable doubt.

iii. Prescriptions

I argue that the standard adopted by the Pre-Trial and Trial Chambers is manifestly unfair to the accused as it places a severely low burden of proof on the Prosecution. A simple risk aggravation test would potentially open up liability to a wide range of commanders that may have allowed isolated infractions to go without punishment, but cannot be said to then tacitly condone all future war crimes that may be committed thereafter. While I recognize the sui generis nature of command responsibility vis-à-vis domestic criminal law, I find it hard to rationalize a step so far away from the foundational notion of nulla poena sine culpa.

Instead, the standard should be somewhat closer to that proposed by Judge Eboe-Osuji, requiring proof beyond a reasonable doubt that the underlying criminal acts were a result of the actions or omissions of the superior. Necessary factors in such a standard would be considerations of remoteness, both geographically and temporally, actual or constructive knowledge on the part of the subordinates regarding the failures of the commander, and whether the commander is still in a position of power at the time the crime is committed. Neither of these factors would be dispositive on their own, but must be weighed based on a totality of the circumstances. For example, it will be difficult to show that a commander’s omissions five years prior influenced a subordinate’s actions in the present, especially if that subordinate was not in the commander’s employ at the time of the original infraction. Likewise, it will be equally difficult to demonstrate that a commander’s failure to punish a subordinate in Kenya influenced the actions of a soldier in the Congo, without more.

Additionally, while it will be difficult or impossible to obtain direct testimonial evidence from the subordinate about their knowledge of the commander’s failure to punish or control her troops in the past, the Prosecutor may be able to establish proof of a culture of impunity by showing the widespread commission of minor unpunished infractions that may not constitute war crimes. Such a culture could lead a subordinate to believe that the commander would allow or permit more serious crimes to go unpunished. Ultimately, this standard is not perfect, but I believe it will allow for more consistent Prosecution and give the Judges a clearer lens through which to decide on the liability of commanders.

IV. Conclusion

Ultimately, the split decision in the Appeal Chamber acquitting Jean-Pierre Bemba Gombo has left the ICC with more problems than solutions. Of the many unanswered questions resulting from this decision, the question of causality in command responsibility is one that will need to be settled. As I have argued in this comment, the different causation tests proposed by each of the Judges above, if implemented, would lead to vastly different conviction outcomes. A clear standard would allow the Prosecutor to collect evidence to meet that standard instead of wasting time and resources. This would make the ICC more efficient and, consequently, help to build support among the international community.

Endnotes — (click the footnote reference number, or ↩ symbol, to return to location in text).

1. 1.

   The Prosecutor v. Jean-Pierre Bemba Gombo, ICC-01/05-01/08 A, Judgment on the appeal of Mr Jean-Pierre Bemba Gombo against Trial Chamber III’s “Judgment pursuant to Article 74 of the Statute” (AC, Jun. 8, 2018) [hereinafter Bemba Appeals Chamber Judgment], available online, archived. ↩

2. 2.

   Rome Statute of the International Criminal Court, Adopted by the United Nations Diplomatic Conference of Plenipotentiaries on the Establishment of an International Criminal Court, Jul. 17, 1998, UN Doc. A/CONF.183/9 [hereinafter Rome Statute], Art. 28, available online. ↩

3. 3.

   Andrew P. Simester, John R. Spencer, G. Robert Sullivan & Graham J. Virgo, Simester and Sullivan’s Criminal Law: Theory and Doctrine 84 (4th ed. 2010).

   (The requirement of causation is fundamental to our understanding of actus reus in criminal law). ↩

4. 4.

   See Phillip DeLacy, The Problem of Causation in Plato’s Philosophy, 32 Classical Philology 97 (Apr. 1939), doi paywall. ↩

5. 5.

   The Prosecutor v. Jean-Pierre Bemba Gombo, ICC-01/05-01/08, Decision Pursuant to Article 61(7)(a) and (b) of the Rome Statute on the Charges of the Prosecutor Against Jean-Pierre Bemba Gombo (PTC II, Jun. 15, 2009) [hereinafter Bemba Pre-Trial Decision], available online, archived. ↩

6. 6.

   The Prosecutor v. Jean-Pierre Bemba Gombo, ICC-01/05-01/08-3343, Judgment pursuant to Article 74 of the Statute (TC III, Mar. 21, 2016) [hereinafter Bemba Trial Chamber Judgment], available online, archived. ↩

7. 7.

   The Prosecutor v. Jean-Pierre Bemba Gombo, ICC-01/05-01/08-3636-Anx3, Concurring Separate Opinion of Judge Eboe-Osuji (AC, Jun. 14, 2018) [hereinafter Concurring Opinion of Eboe-Osuji], available online, archived. ↩

8. 8.

   The Prosecutor v. Jean-Pierre Bemba Gombo, ICC-01/05-01/08-3636-Anx1-Red, Dissenting Opinion of Judge Sanji Mmasenono Monageng and Judge Piotr Hofmański (AC, Jun. 8, 2018) [hereinafter Bemba Appeals Chamber Dissent], available online, archived. ↩

9. 9.

   Bemba Appeals Chamber Judgment, supra note 1. ↩

10. 10.

    The Prosecutor v. Jean-Pierre Bemba Gombo, ICC-01/05-01/08-3636-Anx2, Separate Opinion of Judge Christine Van den Wyngaert and Judge Howard Morrison, ¶ 56 (AC, Jun. 8, 2018) [hereinafter Separate Opinion of Wyngaert & Morrison], available online, archived. ↩

11. 11.

    Concurring Opinion of Eboe-Osuji, supra note 7, at ¶ 202. ↩

12. 12.

    Otto Triffterer, Causality, a Separate Element of the Doctrine of Superior Responsibility as Expressed in Article 28 Rome Statute?, 15 Leiden J. Int’l L. 179 (2002), paywall, doi. ↩

13. 13.

    Protocol Additional to the Geneva Conventions of 12 August 1949, and Relating to the Protection of Victims of International Armed Conflicts, Art. 86(2), Jun. 8, 1977, [hereinafter Additional Protocol I], available online. ↩

14. 14.

    Id. ↩

15. 15.

    Rome Statute, supra note 2, at Art. 28. ↩

16. 16.

    Statute for the International Tribunal for the Former Yugoslavia, S.C. Res. 827, Art. 7(3), U.N. Doc. S/RES/827 (May 25, 1993) [hereinafter ICTY Statute], available online. ↩

17. 17.

    Statute for the International Tribunal for Rwanda, S.C. Res. 955 Annex, Art. 8(2), U.N. Doc. S/RES/955 (Nov. 8, 1994) [hereinafter ICTR Statute], available online. ↩

18. 18.

    ICTY Statute, supra note 16, at Art. 7(3). ↩

19. 19.

    Id. ↩

20. 20.

    The Prosecutor v. Zejnil Delalić, Zdravko Mucić, Hazim Delić and Esad Landžo, IT-96-21-T, Judgement, ¶ 398 (ICTY TC, Nov. 16, 1998) [hereinafter Čelebići], available online. ↩

21. 21.

    Id.; M. Cherif Bassiouni & Peter Manikas, The Law of the International Criminal Tribunal for the Former Yugoslavia 350 (1996). ↩

22. 22.

    Čelebići, supra note 20, at ¶ 397. ↩

23. 23.

    Id. ↩

24. 24.

    Darryl Robinson, How Command Responsibility Got So Complicated: A Culpability Contradiction, Its Obfuscation, and a Simple Solution, 13 Melb. J. Int’l L. 1 (Oct. 28, 2011), available online, archived, doi. ↩

25. 25.

    Trial of Lieutenant-General Shigeru Sawada and Three Others, US Mil. Comm. in Shanghai, Case No. 25, at 15 (Apr. 15, 1946), available online. ↩

26. 26.

    The Prosecutor v. Sefer Halilović, IT-01-48-T, Judgement, ¶ 95 (ICTY TC, Nov. 16, 2005) [hereinafter Halilović], available online. ↩

27. 27.

    See Robinson, supra note 24, at 28. ↩

28. 28.

    See, e.g., The Prosecutor v. Tihomir Blaškić, IT-95-14-A, Judgment, ¶ 75–77 (ICTY AC, Jul. 29, 2004) [hereinafter Blaškić], available online

    (concluding that the existence of causality between a commander’s failure to prevent subordinates’ crimes and the occurrence of these crimes is not an element of command responsibility that requires proof by the Prosecution in all circumstances of a case);

    The Prosecutor v. Dario Kordić and Mario Čerkez, IT-95-14/2-T, Judgment, ¶ 445 (ICTY TC, Feb. 26, 2001) [hereinafter Kordić and Čerkez], available online; Halilović, supra note 26, at ¶ 78; The Prosecutor v. Enver Hadžihasanović and Amir Kubura, IT-01-47-A, Judgment, ¶ 40 (ICTY AC, Apr. 22, 2008) [hereinafter Hadžihasanović], available online; The Prosecutor v. Jean Mpambara, ICTR-01-65-T, Judgement, ¶ 26 (ICTR TC I, Sep. 11, 2006), available online. ↩

29. 29.

    The Prosecutor v. Jean-Pierre Bemba Gombo, ICC-01/05-01/08, Amicus Curiae Observations on Superior Responsibility Submitted Pursuant to Rule 103 of the Rules of Procedure and Evidence (PTC II, Apr. 20, 2009), available online. ↩

30. 30.

    Bemba Pre-Trial Decision, supra note 5, at ¶ 426. ↩

31. 31.

    Id. at ¶ 424. ↩

32. 32.

    Id. at ¶ 425. ↩

33. 33.

    The Prosecutor v. Jean-Pierre Bemba Gombo, ICC-01/05-01/08-3343-AnxI, Separate Opinion of Judge Sylvia Steiner, ¶ 24 (TC III, Mar. 21, 2016), available online, archived. ↩

34. 34.

    Id. at ¶ 20. ↩

35. 35.

    Bemba Appeals Chamber Judgment, supra note 1, at ¶ 67. ↩

36. 36.

    Separate Opinion of Wyngaert & Morrison, supra note 10, at ¶ 33. ↩

37. 37.

    Id. ↩

38. 38.

    Id. at ¶ 54. ↩

39. 39.

    Id. at ¶ 53. ↩

40. 40.

    Concurring Opinion of Eboe-Osuji, supra note 7, at ¶ 186. ↩

41. 41.

    Id. at ¶ 193. ↩

42. 42.

    Id. at ¶ 166. ↩

43. 43.

    Id. at ¶ 185. ↩

44. 44.

    Id. ↩

45. 45.

    Id. at ¶ 188. ↩

46. 46.

    Additional Protocol I, supra note 13, at Art. 86(2). ↩

47. 47.

    Id. ↩

48. 48.

    Concurring Opinion of Eboe-Osuji, supra note 7, at ¶ 191. ↩

49. 49.

    Id. at ¶ 194. ↩

50. 50.

    Id. at ¶ 202. ↩

51. 51.

    Id. at ¶ 199. ↩

52. 52.

    Id. at ¶ 216. ↩

53. 53.

    Bemba Appeals Chamber Dissent, supra note 8, at ¶ 326. ↩

54. 54.

    Id. at ¶ 329. ↩

55. 55.

    Id. at ¶ 333. ↩

56. 56.

    The Prosecutor v. Duško Tadić, IT-94-1-A, Judgement, ¶ 186 (ICTY AC, Jul. 15, 1999), available online. ↩

57. 57.

    See Leila Nadya Sadat, Fiddling While Rome Burns? The Appeals Chamber’s Curious Decision in Prosecutor v. Jean-Pierre Bemba Gombo, EJIL Talk (Jun. 12, 2018), available online, archived. ↩

58. 58.

    Vienna Convention on the Law of Treaties, Art. 31, May 23, 1969, 1155 U.N.T.S. 331, available online. ↩

59. 59.

    Id. ↩

60. 60.

    The Prosecutor v. Jean-Pierre Bemba Gombo, ICC-01/05-01/08-3343-AnxII, Separate Opinion of Judge Kuniko Ozaki, ¶ 21 (TC III, Mar. 21, 2016), available online, archived. ↩

61. 61.

    Concurring Opinion of Eboe-Osuji, supra note 7, at ¶ 239. ↩

62. 62.

    Bemba Trial Chamber Judgment, supra note 6, at ¶ 211. ↩

63. 63.

    Summary of the Proceedings of the Preparatory Committee During the Period 25 March-12 April 1996, U.N. Doc A/AC.249/1, Proposal submitted by the United Kingdom of Great Britain and Northern Ireland, at 85, May 7, 1996, available online.

    (“In addition to other (types of complicity) (modes of participation) in crimes under this Statute, a commander is also criminally responsible (as an aider and abettor) for such crimes committed by forces under his command as a result of his failure to exercise proper control […]” (parentheticals in original) ). ↩

64. 64.

    See, e.g., Triffterer, supra note 12; Robinson, supra note 24; Kai Ambos, Superior Responsibility, in The Rome Statute of the International Criminal Court: A Commentary 823, 860 (Antonio Cassese, Paola Gaeta, John R. W. D. Jones, eds., 2002), paywall, doi. ↩

65. 65.

    Bemba Pre-Trial Decision, supra note 5, at ¶ 425. ↩

66. 66.

    Ambos, supra note 63, at 860. ↩

67. 67.

    Rome Statute, Art. 28. ↩

68. 68.

    Concurring Opinion of Eboe-Osuji, supra note 7, at ¶ 258. ↩

69. 69.

    Id. at ¶ 213–14. ↩