Estimating State Preferences from United Nations Security Council Speeches

• In Event: Text-as-Data and Historical Foreign Policy Decision Making

Abstract

The United Nations Security Council (UNSC) has been a venue for debates and decisions over the world’s most important security issues for over 75 years. However, unlike the UN General Assembly, the UNSC has few contested votes, thus complicating efforts to estimate change and continuity in state preferences over vital security matters. We develop and test a measurement model to estimate dynamic state preferences from UNSC speeches. The model combines features of the Fighting Words method and WordFish in a new Bayesian measurement model estimated in RStan. The estimates have considerable face validity and yield interesting insights about which countries have and have not adopted rhetoric favored by China over that favored by the United States. Overall, we find that China remains isolated with less than a handful of non-permanent members that have adopted positions close to it. The paper discusses several extensions and applications of the measurement model, most notably to estimate evolving state preferences over specific topics.

Authors

• Erik Voeten, Georgetown University

• Michael A. Bailey, Georgetown University

• Cathy Sona Lee, Georgetown University