Abstract | In her article, ‘Measuring extraordinary rendition and international cooperation’, Rebecca Cordell seeks to subject the Rendition Flights Database to a model-based, statistical analysis. She argues that her analysis suggests that more countries were involved in the CIA’s rendition programme than our work has previously established, and that many more flights in the database are likely to be connected to rendition operations than we have identified. While we would not dispute the likelihood that both of these statements are correct, and we have always presented our findings with an acknowledgement that they provide a necessarily limited account of the rendition, detention and interrogation (RDI) programme, we suggest that Cordell’s work should be approached with some caution. This is so for two reasons. First, her findings – although derived through a different mode of analysis from our own – do not appear to move beyond those we have already published. Second, because Cordell has not triangulated her analysis with either data relating to transfers of specific prisoners, or with evidence relating to which countries hosted prisons for the CIA and the operational dates for these prisons, we question her claim that the flights she has identified from the Database are ‘likely rendition flights’. We aim to demonstrate the importance of having a clear understanding of the limitations of big data when researching violations of human rights, especially where such data is related to covert operations. We also seek to show why the Renditions Flights Database only has merit when it is triangulated with a wide range of supplementary sources, including first-hand accounts by prisoners themselves, declassified documents from the US Government, and the findings of parliamentary, journalist and legal investigations. It is this process of triangulation which gives the flight data meaning and which makes it of value for establishing the facts of human rights abuses. |
---|