Vafa Behnam

When does trade protection boost growth and when does it harm growth? I study this question in the context of US industrialization in the 19th century. I develop a novel empirical design leveraging differential county exposure to nationwide tariffs based on distance to international ports, using newly digitized tariff, import price and railway freight data. Exploiting a nationwide fiscal shock that suddenly raised tariffs for reasons unrelated to local industry conditions, I find protection reduces output growth for products close to the technological frontier but increases growth for products far behind it. I also find protection shifts patent activity towards far-from-frontier products, suggesting a reshaping of comparative advantage. A model of trade and innovation with variable markups generates these results: protection enables a far-from-frontier producer to capture the rents from innovating, while a close-to-frontier producer's high market share and low residual demand elasticity mean protection reduces the gain from innovating. This model provides an imperfect competition rationale for infant industry protection distinct from the usual learning-by-doing justification. I conclude by studying the welfare effects of trade, which can be negative unlike in workhorse trade models.

What were the impacts of the 2018 Trump tariff hike on manufacturing growth? Motivated by gravity models, I develop an empirical strategy that exploits cross-state variation in pre-treatment sectoral import volumes, generating differential exposure to nationwide tariff increases. I find a $1 tariff-induced reduction in imports raises output by $0.63 at the state-sector level. Accounting for a number of spatial spillover and general equilibrium channels, I estimate an implied nationwide manufacturing output gain from import protection of $69.5 to $127.5 billion. However, I find a $1 retaliation-induced reduction in exports results in a $3.4 loss in output, roughly five times the per-dollar output gain from import protection. A back-of-the-envelope calculation implies an aggregate loss of $177.8 billion, suggesting export losses are larger than the gains from import protection.

I study India's 2012 Modified Special Incentive Package Scheme (MSIPS), which offered a 25% capital expenditure subsidy to electronics and semiconductor manufacturers at a time when nearly all chips and half of consumer electronics were imported. Using plant-level data from the Annual Survey of Industries, I employ an event study to estimate the effects of this policy. Treated plants experience a large and persistent increase in total factor productivity in quantities (TFPQ), with no evidence of local spillovers. Markups rise modestly in the long run but not in the short run, suggesting the subsidy is mildly anticompetitive over time. Using an aggregation framework (Levinsohn and Petrin 2012), I find this policy led to a 1.4% increase in aggregate productivity in the formal manufacturing sector in India, driven almost entirely by the direct productivity effect rather than reallocative effects across plants.