Combining Computational and Archival Methods to Study International Organizations: Refugees and the International Labour Organization, 1919-2015

Abstract

Researchers studying international organizations have access to growing and varied archives due to digitization efforts. While developments in computational methods confer efficiency gains for examining these materials at scale, they raise concerns about their validity when applied to interpretive tasks in historical settings. In response, we present a general and flexible workflow that uses simple computational techniques from linguistics to enhance archival researchers' interpretive skills and sensitivity to historical contexts. These techniques also identify patterns that can serve as evidence of causal mechanisms when embedded within strong research designs and theoretical expectations. Then, we demonstrate our mixed-method approach by applying it to a dataset of International Labour Organization's (ILO) annual reports spanning ninety-three years. Examining the ILO's engagement with refugees as described in these documents, we identify key moments during which refugees have been particularly salient for this organization, and the emergence of new issues on its high-level policy agenda.