Published online by Cambridge University Press: 24 October 2016
Human rights monitoring reports play important roles both in the international human rights regime and in productions of human rights data. However, human rights reports are produced by organizations subject to formal and informal pressures that may influence the topics considered salient for attention and scrutiny. We study this potential using structural topic models (STMs), a method used for identifying the latent topical dimensions of texts and assessing the effects of covariates on these dimensions. We apply STMs to a corpus of 6298 State Department Country Reports on Human Rights Practices (1977–2012), identifying a plausible set of topics including killings and disappearances, freedoms of expression and movement, and labor rights, among others. We find that these topics vary markedly both over time and space. We also find that while US domestic politics play no systematic role in shaping topic prevalence, US allies tend to receive more attention to violations of physical integrity rights. These results challenge extant research, and illustrate the usefulness of STM methods for future study of foreign policy documents. Our findings also highlight the importance of topical attention shifts in documents that monitor and evaluate countries.
Benjamin E. Bagozzi is an Assistant Professor in the Department of Political Science & International Relations, University of Delaware, Smith Hall, 18 Amstel Ave, Newark, DE 19716 (firstname.lastname@example.org). Daniel Berliner is an Assistant Professor in the School of Politics & Global Studies, Arizona State University, Coor Hall - Sixth Floor, 975 S. Myrtle Ave, Tempe, AZ 85287 (Daniel.Berliner@asu.edu). Earlier drafts of this paper were presented at the 2016 ISA Annual Convention and at the ASU Center on the Future of War conference “How Do We Know What We Know? Charting the Future for Human Rights Documentation and Analysis.” The authors would like to thank Ken Benoit, three anonymous PSRM reviewers, as well as Chad Clay, David Davis, Christopher Fariss, Jillienne Haglund, LaDawn Haglund, Milli Lake, Will Moore, Kathryn Sikkink, Ryan Welch, Reed Wood, and Thorin Wright for their helpful comments and suggestions. To view supplementary material for this article, please visit http://dx.doi.org/10.1017/psrm.2016.44