Insider witnesses’ credibility and reliability: An empirical legal framework for international criminal justice

To date, we know surprisingly little about how legal practitioners decide whether to rely on a particular (insider) witness, i.e. how they assess whether an individual is providing truthful testimony. This is particularly unexplored in the context of international criminal justice, which faces particular evidentiary challenges. Assessing the practice is imperative to understand whether, and if so, how international criminal fact-finding could be improved. Hence, this study aimed to answer this question: What factors impact insider witness credibility and reliability and how are they assessed in the investigations and prosecutions of international criminal cases? For this purpose, an extensive analysis was conducted of all the ICTY, ICTR, and ICC trial judgements and other case-related documents from 1996 to 2019; along with 320 responses to a factorial survey. The jurisprudence was approached both qualitatively and quantitatively, as were the responses to the vignette study. The empirical data was supplemented with a review of academic literature and policy documents. What this study finds is that across international criminal courts and tribunals, a certain framework for witness assessment has developed over the years: by year 20212, a general witness assessment framework, sharing elements across institutions, was in place. This framework features multiple insider-specific factors, including a ‘caution’ directive for assessments of insider witnesses. The development of the framework is followed by the analysis of how this framework was operationalised regarding insider witnesses. The analysis of judicial practice was based on 1,359 insider witness assessments in the ICTY, ICTR, and ICC trial judgments of 1996-2019. The findings show that close to a half of the insiders at the ICC received a negative assessment outcome (47%), with the numbers considerably lower at the ICTR (28.20%) and the ICTY (17.45%). . Overall, negative rather than positive factors were mentioned more frequently and had more significant relationships with the assessment outcomes. Among quality considerations, those related to external validation (e.g., corroboration, contradiction) were more frequent than those related to internal quality (e.g., level of detail, coherence). Finally, an unexpected lack of attention to witness competence factors was found. In contrast to both previous research and judicial witness assessment framework, indicators of witness competence: e.g., memory, trauma, time-lapse, were mentioned with such low frequency that it was hardly possible to include them in the calculations. The final chapter provides another layer of understanding legal decision-making regarding insider witness credibility and reliability. By means of an experimental vignette study we explored factors considered by decision-makers when assessing insider witness evidence. Analysis of participants’ explanations found (i) a general lack of consistency in the inferences drawn by practitioners; (ii) a relationship between source and information qualities, and (iii) personal expectations of quality as an additional factor in witness evidence assessments. The relatively high degree of inconsistencies and noise in the decision-making observed in this study is problematic as it introduces considerable subjectivity, engendering the likelihood that similar witnesses or similar information could be assessed differently depending on the individual conducting the assessment. In criminal justice settings, where expectations of reliable decision-making are high, greater consistency is desirable.