![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
ABSTRACT
We develop an industry-specific sub-national model of exports that takes into account international and domestic shipping costs. We generate econometric estimates of these shipping costs for the U.S. pharmaceutical and medicine manufacturing industry by fitting the model to 2017 U.S. export data at the district level. Then, we use the econometric model to estimate the value of exports originating from production in each state. Our estimates of states’ export intensities provide indicators of the exposure of workers in individual states to pharmaceuticals-related changes in patent protection, tariffs, and macroeconomic conditions in export markets.
Acknowledgments
The author is grateful to the editor and referees for helpful comments and suggestions. This article represents the opinions and research of the individual author. It does not represent in any way the view of the U.S. International Trade Commission or any of its individual Commissioners.
Disclosure statement
No potential conflict of interest was reported by the authors.
Supplementary material
Supplemental data for this article can be accessed online at https://doi.org/10.1080/08853908.2023.2279230
Notes
1 For example, Kim and Go (Citation2022) find positive effects of exports on employment in South Korean municipalities, and Flückiger and Ludwig (Citation2015) find positive effects on sub-regions of European countries.
2 Another approach is to estimate local exports based on origin-of-movement designations. This approach also has limitations that are discussed below.
3 Riker (Citation2020) develops a similar structural model that estimates the sub-national distribution of U.S. imports across states, rather than exports.
4 This is the value of exports as they depart from the port, before adding the costs of international trade.
5 The log-linear approximation in (4) is similar to the bonus vetus OLS approach in Baier and Bergstrand (Citation2009). Their model focuses on international trade, rather than sub-national trade, so their log-linear expansion is around an equilibrium with symmetric international trade costs, rather than symmetric domestic shipping costs.
6 The source for the export data is the USITC/DOC Trade DataWeb at https://dataweb.usitc.gov/.
7 The model does not analyze U.S. exports from geographically separate Puerto Rico.
8 The model does not include Alaska, Hawaii, or the U.S. territories, which are geographically disconnected from the rest of the U.S. market.
9 The value of shipments and employment in each industry are reported in the 2017 Economic Census database, at https://www.census.gov/programs-surveys/economic-census/data/tables.html.
10 There is some information on domestic shipments of exports from the U.S. Commodity Flow Survey. However, there are limitations on these data, which we discuss in the next section.
11 For example, Autor, Dorn, and Hanson (Citation2013), Acemoglu et al. (Citation2016), and Feenstra, Ma, and Xu (Citation2019) all use lagged employment shares in their econometric models of local labor market effects of trade.
12 To simplify the notation in (7), we have canceled several terms that would be included in both the numerator and the denominator of the ratio.
13 The data are publicly available at https://www.census.gov/data/datasets/2017/econ/cfs/historical-datasets.html.
14 Other limitations of the CFS data are that they are only available every five years when there is an Economic Census, and the Public Use File introduces additional noise in the individual CFS records to mask the identities of shipping parties.
15 Export shipments from distributors have a different pattern of concentration. 43.1% are from California, 27.6% from New York, 3.2% from Tennessee, 2.9% from Illinois, and 2.8% from New Jersey. These products were not necessarily manufactured in these states.
16 The data are publicly available at https://www.brookings.edu/research/export-monitor-2018.
17 The Brookings method is similar to earlier estimates in Testa, Klier, and Zelenev (Citation2003).
18 The share of a state’s employment in the industry that is attributable to exports is also equal to the export intensity ratio if the products that are exported and domestically shipped have similar labor requirements per dollar of output, even if the products have different attributes in the eyes of consumers.