Measuring the Effects of Human Rights Treaties

Do human rights treaties improve human rights conditions on the ground? In the end, this critical question is empirical in character. The effectiveness of any regulatory strategy turns on whether its rules and institutions actually mitigate the problems they are designed to address. Although empirical questions require empirical study, bad data is worse than no data. In a recent study, Professor Oona Hathaway purports to quantify the effect of human rights treaty ratification on human rights violations. Her findings are striking. She contends that ratification is associated with worse human rights practices (when other important variables are held constant). Of course, it is unsurprising that some states continue to commit substantial human rights abuses even after ratifying human rights treaties. It is, however, startling to suggest that treaty membership - including the labeling, monitoring, and reporting of abuses - actually increases violations. In our view, any study advancing such wildly counterintuitive claims carries a heavy burden. While we support the empirical study of these phenomena (and indeed we rely on many such studies in formulating our critique), we identify several problems with Hathaway's project. We suggest that these problems demonstrate serious deficiencies in her empirical findings, theoretical model, and policy prescriptions.