Date of Award

Summer 7-1-2022

Degree Type


Degree Name

Doctor of Philosophy (PhD)




Mitra, Devashish

Second Advisor

Lovely, Mary

Subject Categories

Economics | Social and Behavioral Sciences


This dissertation studies issues at the intersection of trade policy, labor market outcomes, and firms' productivity. It consists of three chapters.

Chapter 1 investigates firms' importing and exporting decision-making and how these decisions impact firms' future productivity. First, I document some facts about firms' importing and exporting behavior. Firms display remarkable persistence in exporting and importing status. However, firms' export and import status do not overlap perfectly, as some exporters do not import while some importers do not export. Motivated by these facts, I then develop a dynamic discrete choice model of a firm's decision to import and export. A firm needs to pay either sunk costs to initiate an activity or fixed costs to maintain that activity, anticipating that the decision would impact its future productivity and expected revenue. Estimating the model for a couple of Chinese industries, I find that both activities benefit a firm's future productivity. Then, I investigate the impact of trade liberalization and government subsidy policies on firms' importing and exporting behavior using the estimated model. A simulation of market expansion suggests that export market expansion raises both exporting and importing participation rates and generates a modest increase in mean productivity. Subsidizing either the sunk or fixed costs of one activity raises the activity's participation rate while discouraging the other activity. In addition, subsidizing the fixed costs of exporting generates a larger increase in total foreign market revenue than subsidizing the sunk costs of exporting.

Chapter 2 examines how trade policy uncertainty may impact firms' decisions and labor market outcomes. Specifically, this chapter links the grant of permanent normal trade relations (PNTR) status to China by the U.S. Congress in 2000 to the evolution of manufacturing employment in China from 1998 to 2007. Generalized difference-in-differences estimates show that Chinese manufacturing industries with high-NTR gaps experience greater employment increases than industries with low-NTR gaps after the grant of PNTR. Possible mechanisms include that more firms enter the ordinary exporting market, foreign firms invest more, and exports grow more to the U.S. than to the rest of the world in high-NTR gap industries than low-NTR gap industries after the grant of PNTR.

Chapter 3 studies how the recent US-China trade war may have disrupted the global supply chain by looking at the hetergeneous responses of goods to ``China tariffs''. Through a series of levies beginning in July 2018, the United States increased its average tariff on imports from China by 17.2 percent points. One goal of the tariffs was to reduce US sourcing from Chinese companies. This goal is at odds with a defining feature of US trade – about 60% of American imports are sourced from foreign multinationals operating in China rather than from domestic manufacturers. This paper investigates the extent to which imports sourced from foreign multinationals and thus linked to global supply chains were more responsive to the "China tariffs" than goods sourced from domestic firms. We apply the difference-in-differences and instrumental variable strategies Fajgelbaum et al. (2020) employed to estimate import demand and export supply elasticities for highly dis-aggregated US imports. We measure multinational presence at the same level of dis-aggregation by the share of US imports originating in foreign-invested enterprises, calculated from Chinese Customs Records. Using tariff and import data through December 2019, our results reinforce those Fajgelbaum et al. (2020) found for a wider set of tariffs and a short time period: US tariffs caused large short-tun declines in average import quantities but had no effect on import prices. We do, however, find significant heterogeneity in import demand associated with a multinational activity. In reduced form estimates, import quantities fell farthest for those products with very low or very high multinational presence compared to those at the median. Two-stage least squares estimates indicate that the import elasticities of products with FIE shares of 20% or 80% is more than two times the import elasticity of products with a median FIE share.


Open Access

Available for download on Monday, April 08, 2024

Included in

Economics Commons