The textual dynamics of international policymaking: A new corpus of UN resolutions, 1946–2018

I introduce a new dataset of all United Nations Security Council and United Nations General Assembly resolutions passed from 1946–2018, as well as machine-learning-based measures of their references to other resolutions, textual alignment, and topics. I suggest applications of this data for a variety of questions in international relations from the development of international law to the influence of state power in international organizations. I illustrate the utility of this dataset by investigating why policymakers employ references in the drafting of legal documents, and how the inclusion of these references affects political outcomes. I draw on theories of international lawmaking to argue that for states deciding whether to vote in favor of a resolution, these references, by signaling ideological consistency with a state’s foreign policy goals and existing consensus amongst negotiators, serve as a strategy to obtain support for resolutions. I found that the inclusion of references did increase political support for resolutions, using my measure of textual alignment to hold resolution text constant while isolating variation in the inclusion of references. I found that even accounting for foreign aid flows as a canonical alternative explanation of vote choice, reference dynamics were an important predictor of state support for resolutions.