UC San Diego

The field of International Political Economy (IPE) has long theorized and studied phenomena which have inherent geographic elements. Either for the sake of parsimony or due to methodological constraints, few studies theorize or test arguments which explicitly account for the role of geography. In this dissertation, I introduce a number of key methodological and theoretical concerns with the assumptions based on geographic dispersion that are inherent in many of the standard IPE models. I argue that by accounting both for the geography of interests and the geography of institutions, I can explain a number of results which are puzzling for prior scholarship. To test these arguments, I introduce a framework to think for how the geography of interest and institutions interact. I present an original dataset on electoral boundaries which allows me to test the arguments. I illustrate the effect of geography primarily in the area of agricultural policy. Specifically, I focus on explaining Congressional support for reforming the American Food Aid program, broad policy support for agriculture as a whole, and crop-specific policies. This research leads to three key results. First, by aggregating our analysis of a geographically specific phenomenon to the level above the explicit political interaction, we can obscure important mechanisms and relationships. Second, by accounting for the geographic nature of this interaction, we can make sense of a number of empirical observations which are puzzling to current theories. Third, both the results for food aid and on crop specific supports suggest that advances in remote sensing of data and spatial analysis allow us to measure interests at a very fine-grained level, in this case down to the specific crops or commodity. All three of these findings suggest that the dataset of electoral boundaries produced here will help us to further refine and test theories looking at the role of special interests in foreign- focused economic policy