In this three part series of posts I’m reflecting on the lessons to be learned from the sharply contrasting results last month at the International Criminal Court with a conviction entered in Ntaganda and reasons finally being released for the dismissal of the Prosecution case in Gbagbo and Blé Goudé. The former involved a rebel commander accused of being a co-perpetrator of attacks against civilians, including sexual offences and sexual enslavement. Yesterday, I noted that in Ntaganda the ICC Office of the Prosecutor (OTP) benefitted from its long engagement with, and consequent deep knowledge of, the relevant conflict. It also ran a well-prepared case targeting a rebel leader (as both a direct and indirect perpetrator) and had framed charges based in common facts and a limited number of key incidents. Gbagbo and Blé Goudé involved allegations that the former president of Côte D’Ivoire organised attacks upon civilian supporters of his principal political rival in post-election violence. The key question, of course, is what accounts for the difference in outcomes?
Today I will examine Gbagbo and Blé Goudé in some detail, and tomorrow I will ask – looking at the OTP’s new strategy document – whether the right lessons have been learned.
What went wrong in Gbagbo and Blé Goudé
The majority in the Gbagbo and Blé Goudé Trial Chamber for the no case to answer motion were Judges Henderson and Tarfusser, Judge Herrera-Carbuccia dissenting. For reasons of space, I will focus on the Henderson and Tarfusser separate opinions (although technically, Judge Tarfusser concurs in Judge Henderson’s reasons for dismissing the case which makes his opinion the Chamber’s “reasons”). In sum, though, their account of what went wrong for the Prosecutor was: a poorly conducted investigation was conducted which then had to underpin an inflexible and overly simplistic case theory, which was in turn poorly executed in the courtroom. “In a nutshell, the majority acquitted Mr Gbagbo and Mr Blé Goudé because the way in which the Prosecutor depicted their actions and omissions from a legal point of view could not be sustained by the evidence” (Judge Henderson, Preliminary remarks, para 2). These opinions do not make for comfortable reading.
The underlying investigation
Judge Henderson is blunt about the quality of the investigation and evidence, noting hundreds of documents originating in the Côte d’Ivoire Presidential Palace were not inspected until 10 months after Mr Gbagbo’s arrest and no evidence was led as to how (or if) they were secured from tampering (paras 34-35) or indeed “bombing, pillaging or raiding” (Tarfusser, para 92). Judge Tarfusser further recounts that he:
“saw with disbelief that, instead of formally seizing original items, the Prosecutor and her investigators had simply photocopied them; Witness P-0045 confirmed that OTP investigators had made copies of notes they had received from the witness and selected as relevant. As a result, the record contains copies of documents … which no longer exist in original form, due to … their disappearance, misplacement or destruction. Also, sections and sentences ended up being cut away in the process of photocopying; the quality of these copies is sometimes bad and makes them virtually illegible …” (para 91).
Judge Tarfusser also notes that “critical interviews” were entrusted to “staff with limited mastery of French” (para 93). He also appears to consider it a weakness that “much of the evidence was essentially provided by the current [Ivorian] government, which is headed by political opponents of the accused” (para 92) – a point which, while concerning, if stringently applied would, more or less by definition, pose serious problems in any prosecution of a former head of state.
Judge Henderson is also not flattering as to the quality of oral testimony: “An extraordinary amount of evidence in this case rests upon hearsay, which the Prosecutor submitted on a prodigious scale. The Prosecutor’s relaxed approach to its use raises serious questions about her methodology” (para 42). The point being not that such evidence is inadmissible, but rather that it is of little or no value without some corroboration, which was seldom provided.
It also appears that some early version of the Prosecutor’s case theory (below) may have distorted evidence gathering: “Unfortunately, the Prosecutor has, on occasion, been selective in the evidence she collected” (Henderson, para 81). Thus, while evidence of payments to youth group leaders could be relevant to a case that they acted as part of a plan, such evidence could only be properly interpreted against a relatively complete set of records of such disbursements to other persons and groups by the Presidential administration. Unfortunately, only the “youth leader” payments were excerpted from those records.
The Prosecutor’s case theory
Henderson notes that that the factual case presented was “exceedingly complex” and that this was likely a strategic mistake:
“The Prosecutor, in a Herculean effort, has attempted to bring within the scope of this single litigation several years of Ivorian history … However, in framing the factual scope of the charges so broadly, the Prosecutor may have bitten off more than she could possibly chew with the resources that were available to her” (Preliminary remarks, para 5 and see para 10).
In particular he notes the requirement to prosecute complex international crimes that one must “develop a thorough understanding of the political, social, cultural, and military background” giving rise to those crimes (Preliminary remarks, para 10; and para 77).
The Prosecutor’s essential case theory was that, by the time of the 2010 election, Mr Gbagbo and his “inner circle” of supporters had formed a plan that he should remain in office no matter the election outcome. To this end “adopted a policy to engage in an attack against the part of Côte d’Ivoire’s civilian population that supported his main political rival, Mr Ouattara.” (Henderson, para. 54). This was thus to be a “a straightforward story about a leader who could not accept that he had come to the end of the line and was unwilling to hand over power to his natural successor” (Henderson, para. 77).
Of this case theory Judge Henderson observes:
“The Prosecutor’s narrative is largely internally coherent and prima facie plausible. Nevertheless, there are a number of points of serious concern. The main concern is that the Prosecutor seems to have presented a rather one-sided version of the situation in Côte d’Ivoire. There is a reason why we ask witnesses to undertake to tell the ‘whole truth’. This is because withholding part of the relevant information may be highly misleading” (para 66).
He notes Mr Gbagbo “was never a ‘normal’ president in a ‘normal’ situation” but rather “there appears from the evidence to have been an ongoing … [and] drawn-out civil war [which] must have been at the forefront of [the mind of] everyone involved in the post-electoral crisis” (paras 67 and 72). In this context, presenting a situation as involving deliberate attacks on civilian neighbourhoods risks overlooking that the regular army was subject to “frequent attack” and ambush in “urban guerrilla warfare” against a Commando invisible (para 68; and see Tarfusser paras 105-107 on the role of these attacks “in destroying the Prosecutor’s narrative”). Further, local concerns about possible intervention by the former colonial power, France, whether justified or not “may have gone a long way to explaining a lot of the accused’s choices, rhetoric, and conduct” (para 70).
In particular, the Prosecutor seems to have presented a (none of the judges use this term) ‘Rwandan-ized’ narrative in which the country had serious ethnic or religious cleavages which mapped onto “population was entirely divided into supporters of Mr Gbagbo and supporters of Mr Ouattara” (Henderson, para 76; see Tarfusser, para 12). While this narrative is simple enough to understand, it was not borne out by the evidence of Prosecution witnesses on the stand. Judge Tarfusser quotes at length from testimony in which witnesses note they lived in cosmopolitan neighbourhoods, not ones split along ethnic or political lines, and in which it was often not possible to tell the origins of one’s neighbours (Tarfusser, para 13).
As the Prosecutor lacked “almost any direct evidence for her version of events, [and instead] she has advanced an elaborate and multi-faceted evidentiary argument that is built almost entirely upon circumstantial evidence.” (Henderson, para 78). Thus,
“The Prosecutor relies on a wide range of factual allegations from different areas, ranging from events that predate the post electoral crisis by a decade to minute details about meetings and small -scale financial and other transactions. According to the Prosecutor, this amalgam of hundreds of individual facts together form a system of evidence that can only be fairly assessed as a whole. However, simply declaring that everything is connected and that it is only possible to understand the true significance of the individual constituent elements by viewing all aspects of the case in their totality, is one thing. Articulating what these myriad connections are and demonstrating that they actually exist and form a coherent whole, is quite something else.” (ibid).
In Judge Tarfusser’s terms, the Prosecutor’s evidence largely consisted of “nothing more than a combination between neutral, institutional conducts, on the one hand, and readings of such conducts so as to make them consistent with the ‘case theory’, on the other hand”, an approach which can only succeed if no other readings are plausible and no contradictory evidence emerges (Tarfusser, para 15). Thus, ordering a curfew might be evidence of a military take-over; but when conducted in accordance with long-established national law it might only be a measure to restore public order. Thus, the weakness of circumstantial evidence is that wrong inferences may be drawn from perfectly true circumstances (ibid). Further, the Prosecutor’s witnesses gave evidence that the curfew had been recommended to the President by the military, not the other way around (Tarfusser, para 55).
It is hard to describe such trenchant criticism neutrally. At best the suggestion is that the Prosecutor, having formed a case theory, engaged in motivated reasoning, presenting a bricolage of de-contextualised evidence and inviting the Court to join the dots. As Judge Henderson put it:
“the Prosecutor seems to have started from the premise that her case theory is correct and that this theory provides the necessary coherence to link the disparate evidentiary elements she relies upon. However, this is putting the cart in front of the horse. In order to prove her case, the Prosecutor must first demonstrate the aforesaid connections and coherence. This has not been done.” (para 79).
Worse, Judge Tarfusser notes that there was no effort on the part of the Prosecutor “to adjust and progressively amend her narrative” in light of the evidence as it unfolded, rather her case theory remained the same (para 104). “Throughout the trial and until her closing statements, the Prosecutor’s failure to meaningfully address facts and circumstances coming on the record through her own witnesses which were not consistent with her own ‘case-theory’ was striking” (ibid).
As Judge Henderson puts it later: “If it is a combination of evidence that allegedly proves a fact, then the Prosecutor must clearly identify all the pieces of the puzzle and, crucially, explain how they fit together” (para 87); but “[u]nfortunately, because of the Prosecutor’s ‘everything-proves-everything’ approach, it has proved impossible to conduct a linear analysis of the evidence” (para 91). Judge Tarfusser described the Prosecutor’s approach as “a vortex of circularity, self-reference and repetition that has not made the Chamber’s task any easier” (para 51).
The conduct of the case
Judge Tarfusser is not complimentary about the OTP’s filings or courtroom technique. As for the “documents … filed by the Prosecutor throughout the trial” and “in particular, the pre-trial brief, the Trial Brief and the response to the defence’s motions” are strik[ing] for the degree to which they are lacking in structure, organisation and clarity, a lack compounded by their overall repetitiousness, circularity and redundancy” (para 40). He continues: “none of these flaws is papered over by the staggering size of the submissions or of the number of footnotes; quite the contrary” (ibid).
Judge Tarfusser is also critical of the court-room management of the case: “More than ten trial lawyers … took the floor and questioned witnesses in front of the Chamber” which he took to betoken an unacceptable degree of fragmentation and compartmentalisation of case management (para 96). From the bench it seemed each witness examination had been prepared in isolation “according to a mechanical script” leading prosecution lawyers to be “taken by surprise by developments in their own line of questioning”; there also appeared to him to be no plan “to have, where appropriate, witnesses appearing at a later stage to comment statements previously made in the courtroom by other witnesses” (paras 97-98). “More specifically, one is unable to guess what might have led the Prosecutor … not to assign the questioning of witnesses expected to testify on the same or related topic to … the same lawyer” (para 101).
Judge Tarfusser also considers many days of court time were wasted on witnesses with nothing relevant or probative to contribute, including four days questioning two documentary film makers one of whom spoke no French while the other had only such a grasp of it as to permit “remarkably” or “outrageously” inaccurate translations (para 24).
The mode of liability
This was also a much more complex case for the Prosecutor, in that there was no evidence of Mr Gbagbo acting as a direct perpetrator. Further, the majority was obviously unpersuaded by the circumstantial case that Mr Gbagbo could be considered a co-perpetrator through participation in a common plan. Insofar as command responsibility was alleged, not enough was done by the Prosecutor to show how Mr Gbagbo was expected to exercise control “in a context as difficult and chaotic such as the post-electoral crisis” and regarding events alleged to have occurred between 16 December 2010 and up to and including “the hours and days following his arrest” on 11 April 2011 (Tarfusser , para 17). Quite apart from the difficulty of a President punishing conduct which occurred after he was deposed, there was only a four-month window in which Mr Gbagbo could have conducted investigations and taken disciplinary measures. Judge Tarfusser pointedly notes that the Prosecutor’s own investigations into such matters took several years.
Not all the blame necessarily rests with the Prosecutor. Judges have a role in imposing court room discipline, in conserving the use of expensive court time and translation facilities, and in not indulging counsel in irrelevancies or digressions. There were obvious tensions on the bench about the handling of evidence and other matters. Judge Henderson is sceptical of the value of admitting “over 4,610 items of documentary and other … evidence” without establishing their relevance or authenticity ab initio (para 32). In the result: “It is probably fair to say that a majority of documentary exhibits that were submitted by the Prosecutor in this case would not pass even the most rudimentary admissibility test in many domestic systems” (para 36; compare Tarfusser para 89). Tarfusser has some wide-ranging and trenchant criticisms of ICC administration and judgement writing in general which merit reading but which space prohibits covering here (paras 7-11, 18-20 and 97). Nonetheless, he openly refers to “the differences in approach within the bench, some of which so deep as to have repeatedly fractured the Chamber” and that “developments in the courtroom might also have led some to guess that I was not always supported in the choices I would have made [as presiding judge] as regards the conduct of the proceedings” (para 6). While better managed proceedings would not have changed the outcome they might have foreshortened the time in detention for an accused with no case to answer.
It is also regrettable that, once again, a visibly splintered Trial Chamber has failed in its obvious duty to deliver only one decision setting out the reasons of the majority and minority in a single document (Article 74(5)). Instead we have a cursory eight page record of the oral decision of 16 July 2019, 961 pages of “reasons” from Judge Henderson dissecting the paucity of evidence, a further 90 pages from Judge Tarfusser (who otherwise concurs with Henderson, while still criticising the length of his analysis), and 300 pages from Judge Herrera Carbuccia on why she would have let the case proceed. The best that can be said of this mess is that the length of the reasons given is almost certainly designed both to send a message to the OTP and to demonstrate to the Appeals Chamber that no stone was left unturned before the case was dismissed.
On the positive side of the ledger, Dov Jacobs has previously noted (as a general matter and not in relation to Gbagbo specifically) that it is not enough to reprimand a Prosecutor for poor management of a case and poor-quality investigations: “there is no pedagogic effect whatsoever if the case is ultimately allowed to proceed”. Here, the Prosecutor has faced sanction in the form of a no-case to answer motion being granted and the case not being allowed to proceed further.
At one level Gbagbo and Blé Goudé reinforces the lessons of Ntaganda. To succeed the OTP needs deep local knowledge, should not (in the first instance at least) target anyone too remote from the direct commission of crimes, and the charges should be simply framed and based on a clear chain of evidence. In Côte D’Ivoire one has the sense that the Prosecutor may have been tempted to work backwards from who was in custody: Mr Gbagbo. At first blush, this was a clear case of a president who refused to accede to the peaceful transfer of power after an electoral defeat. The temptation to look for evidence fitting a case theory that he had orchestrated violence to stay in power must have been strong. There was certainly not much time to investigate other theories: Mr Gbagbo was arrested on 11 April 2011, the Prosecutor officially requested Pre-Trial Chamber authorisation to open a proprio motu investigation on 23 June 2011 (granted on 3 October 2011), and Mr Gbagbo was surrendered to the ICC on 30 November 2011. By the end of December 2011, the OTP was effectively committed to prosecuting Mr Gbagbo even if confirmation of charges hearings did not open until February 2013.
Gbagbo appears to have been driven by an inflexible case theory, with the evidence selected to match it. Unfortunately, once witnesses – the Prosecution’s own – got into the witness box that narrative began to unravel. Rather than adjust, the Prosecutor dug in. Her case was further hampered by poor investigations, badly drafted filings, and disjointed courtroom management. None of this speaks well of the OTP’s ability to handle complex proceedings against former heads of State. While violence certainly occurred in Côte D’Ivoire and many victims suffered, the OTP failed to demonstrate through “linking evidence” that Mr Gbagbo ordered those crimes, planned them, or failed to do what was within his power (in a limited window of time) to repress or punish them. It is at least possible that no such linking evidence was found because none exists. It may simply be that Mr Gbagbo was the wrong defendant to choose, but having chosen him and set events in motion it was too difficult for the OTP to change tack.
In tomorrow’s post I will look at the OTP’s new strategic plan and ask if the right lessons have been learned from these experiences, as well as offering some general conclusions.