Towards an ‘International Criminal Evidence’?

Does the International Criminal Court (ICC) have an ‘evidence problem’? Recent practice would seem to suggest so. Numerous cases have been dismissed at the confirmation of charges stage, with the Pre-Trial Chamber determining that the ‘substantial grounds to believe’ standard of proof had not been met. Last year, the Al Hassan Trial Chamber disappointingly admitted alleged ‘torture-tainted’ evidence, limiting its analysis to whether the statements were obtained through ill-treatment, and not the broader system of detention in which the accused was held at the time he gave those statements. High-profile acquittals such as those of Laurent Gbagbo and Charles Blé Goudé at the no case to answer stage of proceedings, and that of Jean-Pierre Bemba Gombo on appeal, included sharp criticisms of the Office of the Prosecutor (OTP) for failing to bring sufficient evidence to support its case.

The Bemba Appeal Judgment, its accompanying dissenting and separate opinions, and the paroxysm of commentary that surrounded it, provided a fascinating insight into some of the key tensions, disputes and uncertainties that have plagued the ICC since its inception. It showed that fundamental questions, such as the extent to which Trial Chambers should underscore which evidence supports a factual finding and how that evidence led the Chamber to its conclusions, are answered differently by different judges. It also demonstrated a sharp dispute in interpretation on the standard of appeal for findings of fact, and the role of the confirmation of the charges stage of proceedings in shaping the later trial. The Bemba Appeal Judgment highlighted that some of the cracks that had begun to show in earlier ICC case law had now become a gaping chasm.

Judgments must tell a story that is accountable to the evidence. But fact-finding has traditionally been perceived as the poorer sibling to the analysis of legal principles. This dichotomy between facts and law is a false one, to the extent that each legal rule is, in the words of Jerome Frank, a ‘conditional statement referring to facts’. That is to say, the consequences of a legal rule (in the case of international criminal law, the consequence of punishment) can only follow where a particular set of facts has been proven. And substantive law develops through a dialogue with the particular factual features of each case.

At the conference to mark the ICC’s 20th anniversary held on 1 July this year, several commentators remarked that the Court is now leaving its difficult adolescent years and entering full maturity, but many questions remain unanswered. Now is the time to reflect upon and consider the best epistemic conditions for international criminal fact-finding. In my forthcoming book, ‘Proving International Crimes’, I argue that, because of the flexibility built in to the legal and procedural frameworks of international criminal courts and tribunals, practice on many key issues of proof is still evolving, and as such is unpredictable and uncertain. At the same time, investigators, lawyers, scholars, human rights advocates and other fact-finders look at the law and practice of the ICC to establish what standards need to be met in the rigorous collection, preservation, presentation, and analysis of certain types of evidence, only to find the Court’s jurisprudence gazing blankly back at them.

To develop international criminal evidence as a coherent body of law, ICC judges need to do three key things:

1. Abandon the ‘submission’ model for evidence, whereby decisions on the admissibility and relevance of the majority of tendered evidence are deferred until the end of trial. This relatively new practice, now very popular amongst Trial Chambers, gives little guidance to either the parties or the wider public as to what evidence was deemed reliable and why, unless it is explicitly discussed in the final judgment. For example, in Bemba et al., the parties provided competing arguments as to whether material found from an individual’s Facebook account was necessarily reliable and attributable to that person, given that anyone could, in principle, set up an account under someone else’s name. The Chamber did not explicitly address the admissibility of these photographs found on Facebook in its judgment, which means that parties in future cases seeking to admit such evidence have no guidance as to what is needed to show authenticity and reliability. More broadly, as Freeman and Vazquez Llorente have argued, the submission model appears incompatible with the huge volumes of digital evidence that are likely to play a central role in future trials.
2. Decide on a coherent epistemic framework. The drafters of the ICC Statute, as part of the unique compromise between legal traditions reached at Rome, built a great deal of flexibility into the Court’s legal framework. At trial, each presiding judge can give directions for the conduct of proceedings (or indeed, decide not to give any such directions but leave it to the parties to decide). This flexibility comes at the cost of consistency and coherence. On coherence, to paraphrase Mirjan Damaška, creating a successful fact-finding system combining elements of different legal traditions is ‘not like shopping in a boutique’ where one can pick up parts of one system without considering its fit with the broader systemic framework as a whole. On consistency, as Kate Gibson has argued, we can see two Trial Chambers under the same roof, running trials at the same time before the same Court under what look like very different systems, on key practices like the admission/submission of evidence, victim participation, and witness preparation. The judges, to their credit, recently adopted a model ‘Direction for the Conduct of Proceedings’ as part of the latest iteration of the Chambers Practice Manual, which may go some way towards consistency in practice. However, aside from the obvious argument about whether the Manual constitutes judicial procedural law-making through the back door, both the deliberations surrounding this development and the model direction are shrouded in secrecy, with the former having taken place at a 2021 judges’ retreat and the latter being included as a confidential annex to the Manual. Because of this, we are left in the dark about both the outcome of judges’ deliberations (although this can be guessed from reading more recent Conduct of Proceedings directions), and the reasons for their adoption. My book argues for the development of consistent practice for proving international crimes driven by two guiding principles: rectitude of decision and the highest standards of fairness.
3. Within that epistemic framework, consider what approach needs to be taken to the evaluation of evidence and the rigorous linking of fact-finding to just legal outcomes at each stage of proceedings, in line with its standard of proof. As Simon de Smet has convincingly argued, deeper reflection on the different epistemic models and related methodologies is warranted. ICC judges have, on occasion, fallen down the rabbit hole of the ‘atomism v. holism’ debate on the evaluation of evidence. This may prove more of a distraction than a rigorous self-evaluation, given that there are no ‘pure’ examples of either approach. At the heart of this debate, however, is the important distinction of whether judges’ approach is more driven by intuition (i.e. they feel that the evidence as a whole supports their conclusion, and are not compelled to set out each stage in the process of inferential reasoning that led to that conclusion) or rationality (a recognition that one cannot hold all the information in one’s head, and that therefore a rigorous framework for evaluating one’s intuitions is required). An approach with rectitude of decision and fairness at its heart would consider a more methodical approach towards fact-finding, which could benefit from technology to support the construction of arguments linked to the evidence and the drawing of inferences from the evidence presented.

The ICC’s evidence problem is much bigger than the well-documented downsides of investigating from a distance, often after the fact and with limited resources (although that seems not to be the case with the current Ukraine investigation). The development of a coherent model of international criminal evidence needs a firm commitment to reflecting upon how the ‘best epistemic fit’ for the admission, exclusion, and evaluation of evidence, and the marshalling of that evidence to support factual propositions, might be found.