139 Despite acknowledging the variance in data quality, it is surprising that in the LRA Crisis Tracker coding system, ‘Human Rights Watch (HRW) reports are given a verification rating of 5 [the highest ranking] and considered confirmed reports due to their rigorous and thorough research methods’: LRA Crisis Tracker (n 138). Several research studies, however, have systematically detailed severe methodological flaws in HRW's reporting practices, reinforcing the idea that these publications should not be relied on without independent verification: Steinberg, Herzberg and Berman (n 51). For instance, a study conducted by Professors Gerald Steinberg and Abraham Bell of Bar Ilan University, evaluating the reporting of HRW and Amnesty International during the 2006 war between Israel and Hezbollah, found that in ‘nearly every case, [HRW's] initial estimation of Lebanese casualties was exaggerated’ and ‘the lack of reliable sources of information [was] prominent’ (ibid). The study concludes that ‘[HRW's] reports were closer to unverified claims than researched conclusions’. In parallel, William Arkin, an independent researcher and former senior military analyst for HRW, conducted an independent analysis of the Lebanon War while serving as the military adviser to the UN fact-finding mission (ibid). The study contradicted many of HRW's claims, while applying more rigorous methodological standards.
Similar conclusions were reached by scholars from the University of London and the Conflict Analysis Resource Center (CERAC), a Bogota-based conflict think tank. In their 2007 study (Andres Ballesteros and others, ‘The Work of Amnesty International and Human Rights Watch: Evidence from Colombia’, 1 February 2007, http://www.scribd.com/doc/30748914/The-Work-of-Amnesty-International-and-Human-Rights-Watch-Evidencefrom-Colombia), they found that HRW followed a ‘non-systematic approach that includes opaque sourcing and frequent changes in the objects they measure’. Moreover, their research showed ‘a failure to specify sources, unclear definitions, an erratic reporting template and a distorted portrayal of conflict dynamics’ among the methodological problems with HRW's publications (ibid). Furthermore, the report accuses HRW of ‘bias against the government relative to the guerrillas’.
The reliability of HRW reporting has also been called into question at the ICC. In the case of Prosecutor v Callixte Mbarushimana, defence counsel objected to an attempt by the Prosecutor to admit HRW reports into evidence (ICC, Prosecutor v Callixte Mbarushimana, Hearing, ICC-01/04-01/10-T-7-RED, Pre-Trial Chamber I, 19 September 2011, 73). Counsel expressed concern that these reports did not ‘base their assertions on interviews with victims. They merely make a reference to second–hand hearsay emanating from unidentified and unidentifiable villages’ (ibid). Counsel also raised the issue of significantly inconsistent casualty figures:
Why did Human Rights Watch say in February that dozens of civilians were killed … only to change this figure to seven in the December 2009 report? Did people come back from the dead? What does this tell us about Human Rights Watch's working methodology? What does this tell us about Human Rights Watch's statistics? (ibid 80)
Based on these arguments, the Court ruled:
78. The evidentiary weight to be attached to the information contained in documents emanating from Human Rights Watch will be assessed on a case-by-case basis. As a general principle, the Chamber finds that information based on anonymous hearsay must be given a low probative value in view of the inherent difficulties in ascertaining the truthfulness and authenticity of such information.
(ICC, Mbarushimana, Decision on the Confirmation of the Charges, ICC-01/04-01/10, Pre-Trial Chamber I, 16 December 2011).
In another case, Prosecutor v Germain Katanga & Mathieu Ngudjolo Chui, the Court excluded HRW reports, as well as other NGO publications (publications by Amnesty, FIDH and the International Crisis Group were also excluded) specifically over issues relating to reliability (ICC, Prosecutor v Germain Katanga & Matheiu Ngudjolo Chui, Decision on the Prosecutor's Bar Table Motions, ICC-01/04-01/07, Trial Chamber II, 17 December 2010).
