NLP Approach to the Study Economic Sanctions

Abstract

Existing datasets of economic sanctions rely on secondary sources to record instances of economic coercion. This approach may miss economic sanctions and does not capture important details in how sanctions are threatened, imposed and removed, among other information. We present a computer-assisted analysis approach of text to retrieving sanctions-related government documents. We collect all sanctions events originating in the office of the U.S. President, and all Congressional sanctions for 1988-2016. Our approach has three advantages: (1) it is more complete, (2) it is disaggregated, (3) it includes the original language of the measures. These features directly shed light inter-branch delegation, (partisan) conflict, and policy priorities. Indirectly, they help us both measure and analyze the sources of effectiveness. The data and approach can advance the study of economic sanctions, and can facilitate progress in other areas of research in political science.