Susana SáCouto is the Director of the War Crimes Research Office at the American University Washington College of Law. The views and opinions expressed here do not necessarily reflect the views and opinions of the Open Society Justice Initiative.
Two weeks ago, the Appeals Chamber of the International Criminal Court (ICC) reversed – in a divided opinion – the conviction of former military commander Jean-Pierre Bemba for the crimes against humanity of rape and murder and the war crimes of rape, murder, and pillaging committed by his troops in the Central African Republic (CAR) between October 2002 and March 2003. The decision is clearly a disappointment for the victims of the crimes committed by Bemba’s troops, who have been waiting for more than 15 years for a measure of justice. Among other things, the decision also means that nearly 16 years after the Rome Statute came into force – and despite increasing recognition of the prevalence of sexual violence in the situations under the jurisdiction of the court – the ICC has yet to issue a final conviction for sexual violence.
Others, including Leila Sadat, Alex Whiting, and Diane Amman, have already addressed some of the decision’s limitations. I write here to elaborate on the impact of the Appeals Chamber’s decision, particularly on cases involving charges of sexual and gender-based violence, that flow from: 1) its use and application of a modified standard for appellate review; 2) its approach to the role of the pre-trial chamber; and 3) its interpretation of “all necessary and reasonable measures” that a commander is required to take to avoid liability under Article 28 of the Rome Statute.
Standard of Review
As others have noted, the Appeals Chamber parted from the standard of review for factual errors used consistently not only by the ad hoc international criminal tribunals for the former Yugoslavia (ICTY) and Rwanda (ICTR), but also by the ICC’s own Appeals Chamber (see Lubanga case). Rather than determining “whether a reasonable Trial Chamber could have been satisfied beyond a reasonable doubt as to the finding in question,” which would mean giving the trial chamber a “margin of deference” with respect to its evaluation of the evidence, the Appeals Chamber adopted a new standard by which it may “interfere with the factual findings of the first-instance chamber whenever the failure to interfere may occasion a miscarriage of justice” and requiring it to do so when it is “able to identify findings that can reasonably be called into doubt.”
While application of this standard meant less deference to the trial chamber – which spent nearly four years hearing the case, including 77 witnesses and 773 pieces of evidence – a majority of the Appeals Chamber declined to “assess the evidence de novo,” meaning that even though it reviewed the record itself rather than relying on the trial chamber’s assessment, it did not assess anew all the evidence in the record. Indeed, relying on limited evidence, and without reference to arguably relevant evidence – including, as Whiting points out, a corrupt defense witness (D54), or, as the Dissenting Opinion [pdf] notes, the “evasive or contradictory” demeanor of another witness (P36) – the Appeals Chamber disagreed with the trial chamber’s assessment of Bemba’s liability as a commander, acquitting him of all charges.
This bodes ill, particularly for cases involving sexual violence crimes, which often require a comprehensive analysis of context to understand how such violence is actually perpetrated in times of conflict or mass violence. Sometimes rape is ordered by military or political leaders, but often it is not explicitly planned or orchestrated from the outset. Sexual violence may initially be committed because the atmosphere of war and the chaos that accompanies it permits it to occur. However, once it becomes clear that superiors do not disapprove of sexual violence, “opportunistic rapes typically then become more public, more frequent, and more violent,” becoming part of the overall violence against a targeted group. Recognizing this often requires the court to consider all the evidence holistically and make common sense inferences based on that evidence.
For instance, in In the Kvočka case, the ICTY found that rapes committed in a detention camp were foreseeable – despite the absence of evidence showing that the accused knew that women had been previously raped in that camp – by looking at all the evidence and drawing inferences from the surrounding circumstances. In particular, the trial chamber in the Kvočka case found the fact that “[a]pproximately 36 women were held in detention, guarded by men with weapons who were often drunk, violent and physically and mentally abusive and who were allowed to act with virtual impunity,” and concluded that “it would be unrealistic and contrary to all rational logic to expect that none of the women held in Omarska, placed in circumstances rendering them especially vulnerable, would be subjected to rape or other forms of sexual violence,” adding that this was “particularly true in light of the clear intent of the criminal enterprise to subject the target group to persecution through such means as violence and humiliation.”
Without a thorough review of all the evidence – including a contextual analysis of whether and how sexual violence that may at first appear unintended might actually be connected to the commission of other crimes and thus attributable to the individual(s) responsible for those crimes – sexual violence crimes may well go unpunished. The risk is not just theoretical. As Katanga (one of only two other cases to reach judgment on charges of sexual violence at the ICC) makes clear, even when sexual violence is committed by the same group of perpetrators against members of another group at the same time and place as other forms of violence perpetrated against that group, a limited assessment of the evidence can result in acquittal of the sexual violence charges. This risk is even greater if the standard of review permits an appellate chamber to interfere with the findings of a trial chamber when they can be called into doubt but its review is limited to select parts of the record.
Role of the Pre-Trial Chamber
Second, the Appeals Chamber in the Bemba decision took a different approach than that followed and previously supported by the Appeals Chamber with respect to the role of the pre-trial chamber in the confirmation of charges process. That process – unique to the ICC – requires the pre-trial chamber to find “substantial grounds to believe” that the accused committed the crimes alleged. As the court itself has repeatedly affirmed, this standard is low and intended to prevent “wrongful and wholly unfounded” charges from moving to trial. However, the Appeals Chamber in Bemba adopted what seems like a much higher standard, requiring the pre-trial chamber not only satisfy itself that the charges are not frivolous, but also confirm all individual acts underlying the charges. This is true even if, as was the case with Bemba, the accused receives adequate notice of any additional acts underlying the charges during the often lengthy period between confirmation and the first day of trial (as a report by the War Crimes Research Office (WCRO) shows, as of late 2015, this period ranged from 13 to 22 months).
Again, this bodes ill for cases involving sexual violence crimes. For a variety of reasons – including that sexual violence is sometimes mistakenly viewed as incidental, opportunistic, or isolated as opposed to part of a broader conflict or interconnected with other crimes – investigators often do not prioritize investigation of sexual violence crimes from the outset. Indeed, as the jurisprudence of the ad hocs makes clear, evidence of sexual violence often comes out late in the investigation of atrocity crimes (see, for example, Kabiligi & Ntabakuze [pdf] and Niyitegeka [pdf], where evidence of sexual violence gathered after the confirmation of the initial indictment led to amendment of the indictment) and sometimes even during the trial (see Akayesu [pdf], where evidence of rape came out during a witness’ testimony at trial, leading to the amendment of the indictment six months after trial started).
A rule barring the trial chamber from considering acts of which the accused receives adequate notice but are not specifically confirmed by the pre-trial chamber would not only lengthen an already prolonged confirmation process (as the WCRO report shows, as of late 2015, the amount of time between a suspect’s initial appearance and a final confirmation decision ranged from seven to 33 months), but would also likely have an adverse impact on sexual violence cases.
Finally, the Appeals Chamber approach to Article 28(a), combined with the court’s earlier approach to other modes of liability, increases the risk of impunity for crimes of sexual and gender-based violence. As a forthcoming article I co-authored with Leila Sadat and Patricia Sellers explains, the Appeals Chamber restrictive interpretation of Article 25(3)(a) is likely to serve as a particularly high bar for cases involving sexual violence charges, as it will likely be difficult to prove that individuals accused of such crimes “unquestionably… conceived the crime, oversaw its preparation at different hierarchical levels, and controlled its performance and execution,” as required by the court’s interpretation of co-perpetration (see Katanga), given that sexual violence – even when widespread – often occurs because it is tolerated and permitted rather than explicitly ordered or planned.
Second, while an accused could be held accountable for sexual violence crimes through other modes of liability, such as common purpose liability under Article 25(3)(d), this provision, as we argue in the article, has been applied in a discriminatory manner to sexual violence crimes, resulting in the acquittal of sexual violence charges (see Katanga). The Appeals Chamber’s interpretation of Article 28(a), if followed, would further narrow the modes of liability under which a perpetrator may be held accountable for sexual violence crimes.
In Bemba, the Appeals Chamber concluded that the trial chamber made a number of errors that “resulted in an unreasonable assessment of whether Bemba took all necessary and reasonable measures” which a commander is required to take to avoid liability under Article 28 of the Rome Statute. Among these errors was the trial chamber’s failure to assess the measures Bemba should have taken by reference to the specific crimes that were actually committed, which the majority suggests –without citing any specific authority – should have been limited to the crimes it found had been established beyond reasonable doubt, namely one murder, 20 acts of rape, and five acts of pillaging.
Although it emphasized that the scope of the duty to take such measures depended on the number of such crimes, which in this case was “comparatively low,” it said nothing about the sufficiency of such efforts with respect to the particular nature of those crimes. Having found that a vast majority of the crimes established beyond reasonable doubt were sexual violence crimes, it is somewhat surprising that the Appeals Chamber makes little mention of the fact that two of the chief mechanisms set up to investigate allegations of crimes committed by Bemba’s troops in CAR either did little to pursue reports of rape (the Mondonga Inquiry) or were limited to allegations of pillaging (Zongo Commission). Indeed, although there is evidence in the record that Bemba was copied on a case file which contained detailed information about acts of pillaging and rape allegedly committed by his troops in the CAR, the Appeals Chamber makes no effort to assess the competence or quality of that investigation with respect to the sexual violence allegations. The decision suggests that an assessment of “all reasonable and necessary measures” need not consider whether and how those measures adequately investigated the particular type of crimes committed. Again this bodes ill for cases involving sexual violence, as war crimes investigations, even when adequate as to other crimes, have often suffered from insufficiencies when it comes to sexual violence crimes.
The implications of the decision could be quite far-reaching not only for the ICC, but also for other courts with jurisdiction over similar crimes, including as the Special Criminal Court (SCC), the court established in CAR to address some of the very same type of violations that were the subject of the Bemba case. If the ICC could not hold Bemba responsible for the sexual violence crimes committed in CAR, how will the SCC hold other perpetrators accountable for sexual violence crimes?