Bemba Acquittal Rests on Erroneous Application of Appellate Review Standard

[Jennifer Trahan is a Clinical Professor at the NYU Center for Global Affairs.]

While there are many subsidiary issues and it will take substantial time to fully analyze each of them, the crux of the decision in the case of Jean-Pierre Bemba Gombo, sentenced to 18 years by the Trial Chamber and acquitted of all charges upon appeal, rests on the margin of deference to accord Trial Chamber determinations of an issue of fact (or mixed fact and law). The key issue (aside from one related to charging) came down to whether Mr. Bemba, whose subordinates were found to have committed murder as a war crimes and crime against humanity, rape as a war crimes and crime against humanity, and pillage as a war crime, in the Central African Republic between 2002 and 2003, took the “necessary and reasonable measures” to prevent or repress the commission of the crimes or to submit them to the competent authorities for investigation and prosecution. Such measures are required under the command responsibility standard of Article 28(a)(ii) of the International Criminal Court’s Rome Statute.

Based on a lengthy trial in which 77 witnesses testified, and 733 items of evidence were admitted, the Trial Chamber found that he did not do so, concluding that the measures taken “patently fell short of ‘all necessary and reasonable measures’ to prevent and repress the commission of crimes within [Mr. Bemba’s] material ability.” (Trial Chamber judgment, para. 731.) Under established jurisprudence, this factual determination was entitled to deference when the issue went up on appeal. Unfortunately, in a “significant and unexplained departure” from prior jurisprudence (and similar jurisprudence of the ad hoc tribunals), the Majority applied an unworkable and arguably erroneous standard of appellate review in examining this key issue on which the outcome largely hinged.

The standard to be applied is whether a reasonable trier of fact could have reached the finding in question, based on the evidence that was before the Trial Chamber. (Dissenting Opinion, para. 47.) In the Dissenting Opinion of Judges Monageng and Hofmanski, they persuasively point out that this was not the standard applied by the Majority who voted to acquit (Judges Van den Wyngaert and Morrison, with President Eboe-Osuji concurring based on a separate opinion). Instead, the Majority took selected pieces of evidence, and based on a review of only those pieces of evidence, essentially revisited the Trial Chamber’s factual findings, finding “some doubt,” and substituted their judgment as to a factual determination made by the Trial Chamber. But the reason that a margin of deference is given to Trial Chamber judges as to determinations of fact is that they have reviewed the totality of the evidence presented to them—especially important regarding such lengthy and complex trials as are conducted by the ICC. (If the “some doubt” or “serious doubt” standard were correct, as Judges Monageng and Hofmanski point out, it would mean that in any case with a dissent at the Trial level, there should be acquittal on appeal, as the Dissenter obviously had doubt. (Dissenting Opinion, para. 13.) The Majority’s new appellate review standard is thus not only unprecedented, it would make ICC convictions nearly impossible. While the ICC may not be strictly bound to apply its own precedent, the Majority makes no case for departing from the established standard of appellate review.

Moreover, whether a commander has taken “necessary and reasonable measures” to prevent or repress crimes committed by his troops/subordinates is necessarily fact and context-specific. It is not “necessary and reasonable measures” in the abstract, but in the specific circumstances in which the crimes were committed, at the time of the crimes, and given the actual ability of the particular commander to exercise measures related to those troops.

Reading both the Majority and Dissenting opinions, it is difficult for this author to conclude whether or not Mr. Bemba undertook such “necessary and reasonable measures” to prevent or repress the crimes committed by members of the Mouvement de Liberation du Congo, because I have not been following the Bemba trial; I have not been sitting in court every day of the trial to hear the evidence and assess the credibility of witnesses; and I have not thereby gained an appreciation of the particular factual context in which the crimes occurred, nor the conditions in the CAR, nor what measures reasonable could have been undertaken. That is why I will not opine on whether Mr. Bemba took the “necessary and reasonable measures” to prevent or repress the crimes, and that is exactly why the job of appellate review is also not to make such determinations afresh based on a limited review of selected pieces of evidence.  This is why the question on appeal is simply: whether a reasonable trier of fact could have reached the finding in question based on the evidence that was before the Trial Chamber—something the Majority never examined.

Judges Van den Wyngaert and Morrison also, in their separate opinion, also set off into uncharted territory by suggesting that it is really not the job of a high-level commander to prevent or repress crimes, but the job of mid-level commanders “to keep their troops in check. . . .” (Separate Opinion, para. 33)—thereby seemingly suggesting the evisceration of command responsibility for high-level commanders. They also are inexplicably sympathetic to Mr. Bemba as a “remote” commander (Separate Opinion, para. 33), as if we don’t have modern forms of communication, whereby a commander who is physically remote could nonetheless be fully aware of the actions of his or her subordinates and/or impose measures to prevent or repress crimes. (Similarly flawed logic also crept into the now widely denounced Perisic acquittal before the International Criminal Tribunal for the former Yugoslavia, who was also deemed “remote” from the Republika Srpska forces who directly perpetrated the crimes.) (This is a very convenient finding for modern militaries, whose high-level commanders often will be remote from battlefields.)

It is exceedingly distressing for such a significant case to turn on the misapplication of the appellate review standard, and one can only empathize with the massive number of victims in the CAR who must be confounded by this inexplicable result. Prosecutor Bensouda is understandably dismayed, as revealed by her statement of June 13.

Maybe the worst concluding logic is offered by the Majority, when they consider whether they should instead remand the case (as Judge Eboe-Osuji would have done), but Judges Van den Wyngaert and Morrison basically conclude that would take too long to do so. (Separate Opinion, para. 73.) Yes, cases should not take decades, but, if remand were the proper path (and this author is not convinced it was), the time delay of their own court is not an excuse not to remand. Remand, however, was not necessary; it was never shown that the Trial Chamber could not have reasonably reached the decision that it did on the command responsibility charges as to whether Mr. Bemba undertook necessary and reasonable measures to prevent or repress the crimes. Absent such a showing of error, the Trial Judgment finding of command responsibility should have been affirmed.