Research

1. Agglomeration and Local Spillover Effects of US Exporters (Job Market Paper) [SLIDES]

   • • Solo-authored

This paper leverages a novel detailed dataset of US plant level exporting activity to analyze the effects of information frictions on establishment export decisions. I establish the existence and magnitude of local exporter spillover effects in the US for the first time. I exploit county level variation in export activity to show that reductions in uncertainty about export demand can explain up to a quarter of the variation in establishment export destination and product choice. This effect drives both entry and exit decisions as well as dynamic product-country portfolio composition decisions.

2. Selling Abroad: A Comprehensive Portrait of US Exporters

   • • (with Stefania Garetto, Lindsay Oldenski, Nitya Pandalai-Nayar, and Natalia Ramondo)

We document a comprehensive set of facts about US-owned exporters. We compare the behavior of US exporters located within the US with the behavior of US-owned exporters located abroad--- i.e., foreign affiliates of US multinational firms. We bring together firm-level data from the US Census Bureau and from the Bureau of Economic Analysis (BEA) that span many years and industries. While the Census data has rich information on the export activities of US-based firms, the BEA data contain detailed information on the activity of foreign affiliates of US multinational firms, in particular, their export activities. This permits a novel comparison in the literature between ``domestic'' and ``affiliate'' exporters, both for the cross section and time series, and in particular regarding firms' life-cycle dynamics. These facts are aimed at better informing models of exporters and multinational firms behavior.

3. The Geography of Exports

   • • (with Christoph Boehm, Aaron Flaaen, and Nitya Pandalai-Nayar) (updated 4/5/2021)

Many important economic questions require export data at the establishment level or by region. However, assigning exports to establishments or regions in the U.S. is typically not possible or imprecise. Customs data measures trade at the firm level and firms with multiple establishments can typically not be assigned to a unique region. Further, the largest firms, responsible for the vast majority of trade, are multi-establishment firms, exacerbating this measurement issue. This paper develops a novel method to measure exports at the establishment level, using confidential U.S. Customs and Census data. We then construct an accurate county-level trade dataset for the U.S. and apply this new dataset to study the regional implications of the 2007-2009 collapse in trade.