<p>  </p> <p>This dissertation is composed of three independent chapters in the field of the economic geography and international trade. However, there is one uniting theme between all three chapters: geographic spillovers. In each chapter, a source of geographic spillovers that is relevant to policymakers is investigated. Further, each chapter addresses a theoretical or data-driven challenge to identifying these spillovers and implements an improved methodology for estimation. In particular, the first chapter studies agglomeration and congestion spillovers, the impact of employment density on the productivity of workers and the amenities associated with living in a location, respectively, by using variation from a natural experiment. The second chapter studies fiscal multipliers, the effect of government spending on economic activity, by using variation from the same natural experiment. Finally, the third chapter studies import spillovers, the impact of neighboring firms' experience sourcing from foreign markets on the likelihood that firms in the same location and industry enter into those same markets, by using detailed data on Serbian firms.</p> <p><br></p> <p>In the first chapter, <em>Agglomeration and Congestion Spillovers: Evidence from Base Realignment and Closure</em>, I quantify agglomeration and congestion spillovers using variation from a natural experiment by instrumenting for changes in local employment with proposed changes to civilian employment at military installations through the Base Realignment and Closure (BRAC) process. I find an agglomeration spillover elasticity consistent with the existing literature. However, my estimate of the congestion spillover elasticity is smaller in magnitude than common parameterizations of quantitative economic geography models. All else equal, with a weaker congestion spillover elasticity, more of the distribution of economic activity across space is due to natural advantages and disadvantages. This result implies smaller gains from implementing the optimal spatial policy. </p> <p><br></p> <p>In the second chapter, <em>Local Fiscal Multipliers and Defense Spending</em>, I estimate county-level fiscal multipliers using shocks to military employment to instrument for local defense spending. Aggregate shocks to military employment are subject to the Base Realignment and Closure process, which is designed to isolate the recommendations of the Department of Defense from political influences. By exploiting variation in military employment, I address the endogeneity of government spending when using changes in defense spending to estimate fiscal multipliers. In addition, this method addresses the attenuation bias due to geographic measurement error that results from using data on military contracts alone with small geographic units. This extends the common method for estimating state- and national-level fiscal multipliers using variation in defense spending to more local geographic units. My estimates imply a local income multiplier between 0.5 and 0.8, which is smaller than existing estimates that use non-defense-based sources of variation, but larger than the existing estimates based on variation in defense contract spending. </p> <p><br></p> <p>In the third chapter, <em>International Sourcing and Firm Learning: Evidence from Serbian Firms</em>, we find that compared to non-importers, importers are more productive and pay higher wages as they source better quality and cheaper production inputs. However, little is known about how these firms learn about their sourcing markets. We quantify the impact of neighboring firms' importing experience on the decision to start sourcing inputs from new markets using merged customs and administrative data from Serbian firms. We find that firms are more likely to start importing from a new market if firms in the same industry and location have imported from that market and if those firms increased their imports over time. Further, our results support a distinction between imports and exports for the decision to enter foreign markets; unlike exports, import sourcing choices are not independent across countries, but are substitutes. We also investigate origin-country and firm heterogeneity. Our results indicate that the impact of neighboring firms' importing experience is greater for source countries in the European Union market and for firms that are high productivity, foreign owned, and previous importers. Together, these findings suggest that a firm's spatial connections are an important factor in its access to global markets as sources for inputs.</p>