The decision issued by the majority of the ICC Appeals Chamber in the Bemba case on Friday 8 June – and the multiple separate and dissenting opinions published simultaneously and subsequently – have been the subject of intense discussion among academics and practitioners. It is disappointing that some of the debate, at least online, has been portrayed as a division along partisan lines; with those who have expressed concern about the outcome of the case portrayed as reactionary or in thrall to the emotional impact on victims or the self-interest of the OTP, while those who have been less critical of the content of the opinions have been lauded for their objective analysis of substantive legal issues. It is possible to view the Majority’s decision and reasoning as flawed without necessarily subscribing to the viewpoint that any acquittal represents a failure to deliver justice or a disaster for the Court. Likewise, it is possible to believe that the decision to overturn Bemba’s conviction was the correct one in the circumstances without endorsing all of the procedural and substantive conclusions reached by the judges. This post will highlight three issues which, in the view of the authors, represent a troubling precedent and have the potential to paralyse the work of the ICC and regress the development of international criminal law more generally – not because Jean-Pierre Bemba was acquitted, but because of the standard of reasoning and the legal and procedural analysis applied by the Majority.
The divergence of views among the Appeals Chamber on key points
Firstly, the multiplicity of opinions and issues discussed within them actually confuse rather than clarify the reasoning of the judges. The Bemba appeal judgement consists of four separate opinions – the Majority Decision, issued on behalf of Judges Morrison, van den Wyngaert and Eboe-Osuji, the Dissenting Opinion of Judges Monageng and Hofmanski, and then two separate opinions from different factions within the Majority, one Joint Separate Opinion from Judge Morrison and Judge van den Wyngaert and one Separate Opinion from Judge Eboe-Osuji (which was not made public until Friday 15 June, one week after the rest of the decisions). The Majority Decision is remarkably short for a substantive appeal judgement at only 80 pages; the Dissenting Opinion is 269 pages long and includes a detailed analysis of the evidence; the Joint Separate Opinion from Judge Morrison and Judge van den Wyngaert is a mere 34 pages long but addresses a wide range of issues not discussed in the Majority Decision; while Judge Eboe-Osuji’s Separate Opinion is 117 pages long and likewise discusses many questions of law other than the grounds of appeal which were held to be determinative in the Majority Decision. It appears that Judges Morrison, van den Wyngaert and Eboe-Osuji agreed only on the specific grounds of appeal upheld in the Majority Decision, but held divergent opinions on other legal issues which necessitated the issuance of two separate opinions to expound in greater detail.
Although a 3-2 split between Appeals Chamber judges and the issuance of separate or dissenting opinions is not unusual or unprecedented, it is particularly unfortunate in this case – only the third substantive appeal judgement in the ICC’s history – because it appears to indicate a significant lack of agreement between the five judges on important issues of law, and therefore makes it more difficult for parties in future cases and even other judges at the pre-trial and trial phase to assess how best to comply with its findings. For example, the judges were split 3-2 on the appropriate standard of appellate review, 2-2-1 on the appropriate disposition (Judge Eboe-Osuji had originally favoured a retrial but relented in order to avoid “an inconclusive judgement” – Separate Opinion, para 22), 2-2-1 on whether the contextual elements for crimes against humanity had been established (treating the assessment of the contextual elements in paras 57-72 of the Joint Separate Opinion and Judge Eboe-Osuji’s more abstract discussion of the organisational policy for crimes against humanity at paras 283-289 of his Separate Opinion as distinct findings on Ground IV of the appeal), 3-2 on whether only allegations explicitly included in the Document Containing the Charges and the Confirmation Decision can form the basis of a conviction (although Judge Eboe-Osuji expressed a separate opinion as to whether the Trial Chamber can amend the indictment after the start of the trial, section VI) and 2-2-1 on the correct approach to causation as an element of command responsibility (this issue was also subject to different interpretations in the Dissenting Opinion, Joint Separate Opinion and Separate Opinion).
These are important legal issues which will have significant implications for future cases, and although all of the decisions expressed regret at the lack of unanimity, none addressed in any detail how their disparate conclusions and approaches should be interpreted or relied on by the parties or judges in future cases. It is interesting to note that this was the last decision issued by Judge van den Wyngaert (for the Majority) and Judge Monageng (for the Minority) before their term of office as judges expired, which means that the Appeals Chamber now consists of two judges from the Majority, one from the Minority and two newly-appointed judges (Judge Ibáñez Carranza and Judge Bossa) who have not previously ruled on these issues at the ICC. How should the OTP or the Trial Chamber currently hearing the Ongwen case, which includes allegations based on command responsibility, deal with the issue of causation, for example, in which there was no clear majority among any of the judges? How, as is discussed further below, should the OTP and Trial Chambers in the three current trials treat allegations which were not explicitly included in the Confirmation of Charges decision? How should those Trial Chambers formulate their eventual trial judgements to provide sufficiently exhaustive reasoning and conclusions on specific pieces of evidence to pre-empt the Appeals Chambers’ newly-announced standard of appellate review? In that context, the reference in Prosecutor Bensouda’s statement to the Majority’s departure from previously accepted standards and jurisprudence and the impossibility of further appeals should be seen as expressing concern about a lack of clarity for how the OTP should approach future cases and not, as one commentator put it, “an emotional rejection of the Appeals Chamber’s majority view, instead of… a self-critical analysis of its work”.
It is ironic that the Majority Decision and both Separate Opinions from the three Majority judges spoke about the importance of exhaustiveness and clarity of judicial reasoning. Although it is clear which two grounds of appeal the Majority felt were determinative (the scope of which criminal acts were included in the conviction and the Trial Chamber’s assessment of whether Bemba had taken all necessary and reasonable measures to prevent or punish crimes committed by MLC troops under his effective control), they do not provide an extensive discussion of these issues, the evidence assessed by the Trial Chamber relating to necessary or reasonable measures, or prior caselaw. Some of the more alarming pronouncements – such as the suggestion in paragraph 170 of the Majority Decision that “[c]ommanders are allowed to make a cost/benefit analysis when deciding which measures to take”, a statement which seems to infer a quasi-‘military necessity’ defence to a commander’s omissions – are stated as a declarative sentence and not supported by reference to any jurisprudence or academic commentary. Judge Eboe-Osuji states in paragraph 31 of his Separate Opinion that “a key motivation” for his discussion of legal issues outside the scope of the appeal “is to assist in the understanding of those aspects of the law… which have vexed the mind of jurists over the years”. However, the fact that his opinions on those issues are restricted to his Separate Opinion makes it clear that they were not shared by his fellow judges in the Majority or they would have been included in the Majority Decision.
Judge Morrison and van den Wyngaert say that the Chamber’s failure to achieve unanimity “is not just a matter of difference of opinion, but appears to be a fundamental difference in the way we look at our mandates as international judges” [Joint Separate Opinion, para 4]. It is interesting to note that ten years ago, the Appeals Chamber of the ICTY in the Orić case had an opportunity to dramatically alter the parameters of command responsibility by extending it to the actions and omissions of commanders in relation to crimes committed by their forces before they had assumed command – essentially, a failure to punish crimes committed at a time when the commander did not have effective control over the direct perpetrators. In the Hadžihasanović appeals judgement only a few months earlier, Judge Shahabuddeen had entered a separate dissenting opinion arguing in favour of the extension of potential liability to subsequently appointed commanders. In the Orić appeal, however, he declined to join with dissenting Judges Schomburg and Liu to adopt that standard by bare majority, stating:
“a decision to reverse turns upon more than theoretical correctness; it turns upon larger principles concerning the maintenance of the jurisprudence, judicial security and predictability…. Since I was one of the two dissenting judges in the earlier case… a reversal should await such time when a more solid majority shares th[ose] views”. [paras 14-15]It is unfortunate that the Majority in Bemba did not share this degree of judicial restraint or consideration for the perception of legal authority which attaches to bare majority appellate decisions.
Implications for the confirmation of charges process
The Majority held that specific criminal acts of rape, murder and pillage which were added to the case after the issuance of the Confirmation of Charges decision could not form part of Bemba’s conviction because they “did not form part of the ‘facts and circumstances described in the charges’ – to the extent that the document containing the charges was not amended to reflect them” [Majority Decision, paras 74-115]. The framing of the charges in the Document Containing the Charges (DCC) and the operative part of the Confirmation Decision was formulated to include the temporal and geographic scope of ‘categories of crimes’ – murder, rape and pillage – followed by a non-exhaustive list of specific criminal acts, prefaced by the phrase “including, but not limited to”. The Majority held that this formulation was “too broad to amount to a meaningful description of the charges against Mr Bemba” [Majority Decision, para 110] and that his conviction would therefore be limited to only those specific individual criminal acts which had been mentioned in the Confirmation Decision and Amended DCC. The Majority therefore excluded eighteen acts of murder, rape and pillage which had been proven beyond a reasonable doubt but had not been explicitly included in the DCC and Confirmation Decision [Majority Decision, paras 116-117].
The Majority’s views on the level of forensic detail which is necessary at the confirmation phase has serious implications for future cases at the Court. The confirmation of charges process is provided for in Article 61 of the Rome Statute, but has been criticized as unwieldy and “failing to reach its intended goals” in practice. Prosecutors at the ad hoc Tribunals routinely used the ‘including but not limited to’ formulation in their Indictments, although their pre-trial procedural system did not include an equivalent to the confirmation of charges process. The Majority condemned the formulation of the charges as potentially including “all such crimes committed by MLC soldiers in a territory of more than 600,000 square kilometers… over a period of more than four and a half months” [Majority Decision, para 103] but did not show what specific prejudice had been suffered by Mr Bemba’s defence by the inclusion of evidence of individual criminal acts of rape, murder and pillage at trial which had not been explicitly mentioned in the Confirmation Decision. The purpose of the confirmation of charges process is, as the Minority put it, “determining whether there is a case to be tried… and not… confirming or crystallising the totality of the factual allegations underpinning these charges for the purposes of the trial” [Dissenting Opinion, para 21].
It is hard not to see the Majority’s findings on this issue as having substituted their own preferred procedural and evidential approach in place of the process which has been followed in all cases at the ICC to date and – as Alex Whiting has pointed out – which had previously been endorsed by the Appeals Chamber in the Lubanga case. The OTP has rightly been criticized by judges in the past for the practice of ‘phased investigations’; obtaining just enough evidence to meet the ‘reasonable grounds to believe’ standard for an arrest warrant, then a little more to establish ‘substantial grounds to believe’ for the confirmation hearing, then hoping to find enough to meet the ‘beyond reasonable doubt’ threshold prior to the trial [Kenyatta Confirmation Decision, Dissenting Opinion of Judge Kaul, para 52]. However, there is a very distinct difference between condemning deliberately incomplete investigations and rejecting all evidence of individual criminal acts obtained or introduced post-confirmation. Judge van den Wyngaert has made it clear in the past that she does not approve of the use of post-confirmation documentary evidence or interviews, although the Appeals Chamber has previously recognized that the Prosecution is not obliged by the Statute to complete all investigations prior to the confirmation hearing, although any additional evidence must still be disclosed to the defence in a timely manner [paras 52-55].
The Majority’s approach has the potential to massively disrupt both current and future investigations and cases. It means, in essence, that the OTP must mention every fact and allegation which may become relevant to its case by the time the Document Containing the Charges is filed – not just each individual criminal act, but also other allegations relating to the mode of liability. The Majority’s absolutist approach to the confirmation process led to the condemnation of one of the Trial Chamber’s key findings relating to whether Mr Bemba had in fact taken all necessary and reasonable measures to prevent or punish crimes, one of which was his failure to exercise his power to withdraw or redeploy MLC troops. However, since the redeployment of troops was not explicitly mentioned in the Amended DCC, the Majority held that Mr Bemba had not been given sufficient notice of this issue and that the Trial Chamber should not have relied on it [Majority Decision, paras 185-188]. Leaving aside the ridiculousness of the suggestion that Mr Bemba may not have realised that his ability to redeploy or withdraw troops from the army he created could be relevant to his prosecution for command responsibility, the implication and consequences of the Majority’s proscription are severe.
Given the time limits on submitting the DCC, holding a confirmation hearing and issuing the Confirmation Decision once an accused has made their first appearance, and the fact that the OTP often has no way of knowing whether an accused person will come into the custody of the Court one week or ten years after the issuance of an arrest warrant, the OTP will now be taking a big risk by pulling the trigger on any case unless they are effectively trial-ready. In terms of allocation of resources, the Investigation Division will become even more important since every scrap of evidence necessary for all parts of the case theory will now need to be verified, analysed and incorporated into the charging documents years before the trial begins. The confirmation of charges process, already criticized for being a ‘mini trial’, will now need to operate as exactly that since it will permanently dictate not just the general factual parameters of the charges but the reliability of all specific items of evidence which may become relevant at trial. In addition, evidence introduced by the Legal Representative of Victims during the trial cannot, on the basis of the Majority’s decision, now be included within any subsequent conviction unless the specific factual allegation was mentioned in the Confirmation Decision, which would not have been possible in the Bemba case since the Legal Representative of Victims was not authorized to submit evidence at the pre-trial phase [Majority Decision, para 96].
What should be done about evidence which, through no fault of the parties, does not become available prior to the issuance of the Confirmation Decision or the submission of the Amended DCC? It is not difficult to think of situations where a witness may not be confident about coming forward unless they are sure that the defendant will actually stand trial, particularly when the defendant is a senior political or military figure. The judges of the Majority resolutely refused to engage with the practical consequences of their findings on the scope of the charges, beyond the immediate implications for Mr Bemba’s appeal. The Majority Decision dodged the question entirely by stating that “this is not to say that adding specific criminal acts after confirmation would in all circumstances require an amendment to the charges – this is a question that may be left open for the purposes of disposing of the present ground of appeal” [para 115]. Judges Morrison and van den Wyngaert acknowledged but coyly sidestepped the “pragmatic concerns” about the realistic capacity of the OTP and the Pre-Trial Chamber to include every single underlying act in the DCC and Confirmation Decision, saying only that “to accept [these arguments] would be to affirm that the confirmation process is unfit for purpose” [Joint Separate Opinion, para 28]. Judge Eboe-Osuji, by contrast, departed from all four of his fellow judges and previous Appeals Chamber jurisprudence by pronouncing unilaterally that, in his view, it is possible for the Trial Chamber to amend the scope of the charges, if not the charges themselves, after the beginning of the trial [Separate Opinion, para 97-150]. Despite four separate opinions, the OTP and even their fellow judges in the Pre-Trial and Trial Chambers are left with no clarity or certainty on such a vital procedural issue.
‘Necessary and reasonable measures’ to prevent or punish crimes of sexual violence
One final troubling aspect of the Appeals Chamber decision is the lack of analysis of whether the measures taken by Mr Bemba were sufficient to investigate, prevent or punish acts of rape committed by MLC troops. As Susana SáCouto has also highlighted, the Trial Chamber’s assessment of the specific measures taken noted the fact that the Mondongo Inquiry file did not pursue any allegations of rape, and the mandate of the Zongo Commission was restricted to only the question of the transport of pillaged goods [Trial Judgement, paras 720, 722 and 726], which the Trial Chamber characterized as a “grossly inadequate response” [para 727]. The Trial Judgement did not differentiate between the sufficiency of measures taken to prevent or punish murder, rape and pillage, because it found that Mr Bemba had not taken ‘all necessary or reasonable measures’ to address any of the crimes. It did, however, find that the failure to provide MLC soldiers with adequate pay led them to “self-compensate through acts of pillaging and rape” [Trial Judgement, para 678].
Although the judges of the Appeals Chamber did discuss the issue of whether Mr Bemba had adequate notice of the crime of murder [Minority Decision, paras 307-318] and re-assessed the Trial Chamber’s conclusions on Mr Bemba’s motivations for pursuing the measures he did, none of the decisions analysed the question of whether the measures taken were sufficient to investigate, prevent or punish each of the specific crimes. It seems odd to treat the crimes committed by MLC troops as one homogenous block and not to consider the extent to which the defendant’s response met the ‘all necessary and reasonable measures’ standard for each category of crime, particularly when the crimes were considered separately for the purpose of the ‘actual knowledge’ standard and Mr Bemba himself argued that “the Trial Chamber erred by assessing the evidence relating to crimes of murder, rape and pillage together” for the purposes of establishing notice [Minority Decision, para 308].
This point is not raised merely to argue against Mr Bemba’s acquittal on the charge of rape; if it had been applied at the trial phase it is possible that another Trial Chamber could have found that Mr Bemba had in fact taken all necessary and reasonable measures to investigate and punish the crime of pillage, since many of the measures that were taken focused specifically on that category of crime. However, given the nature of the crime of rape and sexual violence and the degree of impunity traditionally enjoyed by direct perpetrators, it does not seem irrational to question whether a specific commander had taken all necessary and reasonable measures to address the crime of rape in particular, since it is so frequently overlooked in internal disciplinary mechanisms and domestic justice systems. It would be remarkable and deeply unfortunate if the Bemba appeal judgement became an authority for the proposition that all a commander needs to do to indemnify himself against acts of rape committed by troops under his effective control is to send a letter to an NGO, as Mr Bemba did with FIDH. For now, at least, the ICC’s conviction rate for sexual crimes returns to zero.
Certainty and predictability of the law is a fundamental right for all parties in a criminal justice process – not just for the defence, but also for the prosecution, victims and first instance judges. It is arguable that if the Appeals Chamber had upheld Mr Bemba’s conviction by a bare majority, the decision may not have been the subject of such controversy. However, the most troubling aspects of this judgement – particularly the abrupt reformulation of the standard of appellate review and ostensible reconceptualization of the confirmation of charges process 15 years into the work of the Court – have implications far beyond the disposition of the case against Mr Bemba. If the judges of the Appeals Chamber wanted to make such dramatic changes to the applicable law and procedure of the Court, with such profound implications for all ongoing investigations and trials, it would have been preferable to do so in a more unanimous and less fragmented way in order to provide the necessary clarity and certainty for not only the parties but also their fellow judges at the pre-trial and trial phase. The Bemba Appeal Judgement may come to be seen in retrospect as a watershed moment for the law and practice of the Court, or it may prove to be merely an idiosyncratic, opportunistic and ultimately unsuccessful attempt at judicial law-making.
The opinions expressed above are the personal views of the author and not representative of any organization or institution.