In the face of trade liberalization domestic firms are often forced out of the market, whereas others adapt and survive. In this paper we focus on a new channel of adaptation, namely the shift toward increased provision of services in lieu of goods production. We exploit variation in EU trade policy to show that lower manufacturing tariffs cause firms to shift into services provision and out of goods production. Additionally, we find that a successful transition is strongly associated with higher firm-level R&D stocks whereas higher physical capital stocks slow the shift into services provision.

Working Papers:

Feenstra (1994) developed, and Broda & Weinstein (2006) refined, a structural estimator to estimate import demand and supply elasticities. Working through the first principles of the methodology from Leamer (1981), this paper analyzes and improves the technique to provide a unified estimator of import supply and demand elasticities. The proposed LIML estimator corrects small sample biases and constrained search inefficiencies. Standard estimates are shown to overestimate the median elasticity of substitution by over 35%. Applied to US import data from 1993-2007, the biases of the standard estimates translate into an understatement of consumer gains from product variety by a factor of 6.

“Estimating Import Supply and Demand Elasticities: Analysis and Implications,”

Journal of International Economics, 96(1), May 2015: pp 1-17. (Lead article)

(Elasticity Estimates)

Product redesigns happen across virtually all types of products, yet there is little evidence on the market and welfare effects of redesigns. We develop a model of redesign decisions in a dynamic oligopoly model and use it to analyze redesign activity in the U.S. automobile market. We find automobile model redesigns are frequent despite an estimated average cost around $1 billion. Our estimates also suggest that redesigns lead to large increases in profits and welfare due to the strong preferences consumers have for redesigns. Yet, we show that welfare would be improved with modest declines in redesign activity during recession periods.

“Keeping it Fresh: Strategic Product Redesigns and Welfare,” (Joint with Christopher Knittel and Bruce Blonigen), International Journal of Industrial Organization, 53(C), 2017: pp 170-214,

We examine how multinational firms strategically source production to mitigate the consequences of wage bargaining with workers. When wage bargaining pressure differs across countries, firms allocate production of goods with high markups toward countries with relatively competitive labor markets, limiting the rents available to workers with strong bargaining power. We use product- level data from the universe of automotive production facilities in North America at a monthly frequency between 1988 and 2009 to structurally estimate variable price elasticities of demand for different vehicles. From the theory we derive an empirical strategy that allows us to distinguish the impact of wage bargaining pressure from other sourcing motives. We find robust evidence that multinational firms strategically source their products across countries in response to differences in wage bargaining pressure.

“Strategic Sourcing and Wage Bargaining,” (Joint with Nicholas Sly), Journal of Development Economics, 109, July 2014: pp 172-187.

“Market Size, Structure, and Access: Trade with Capacity Constraints,”  European Economic Review, 70, October 2014: pp 276-298.

This paper develops a model of international trade where firms are heterogeneous across capacity and productivity. A binding capacity constraint induces firms to raise prices in order to take advantage of access to new markets. This generates markets with a flexible competitive structure giving rise to instances where trade and trade liberalization negatively impact welfare. The key predictions of my model can be identified by observing the presence of small yet highly productive firms, and substitution by firms across markets as accessibility evolves. Using Thai firm-level data I establish the prevalence of these anomalous firms, and demonstrate they face capacity constraints.

Feenstra (1994) is widely implemented in international trade to estimate elasticities of substitution. Through a Monte Carlo experiment, simulated estimates suggest substantial biases due to weak instruments. However, the derivation of the elasticity of substitution drastically mitigates these biases.

“Investigating the Asymptotic Properties of Elasticity of Substitution Estimates,” Economics Letters, 190 (2), November 2010: pp 57-62.

“Measuring the Benefits of Product Variety with an Accurate Variety Set,”  (with Bruce Blonigen). Journal of International Economics, 82 (2), November 2010: pp 168-180.

Recent studies have used import data to assess the impact of foreign varieties on domestic prices and welfare. We employ a market-based data set on the U.S. automobile market that allows us to define goods varieties at a more precise level, as well as discern location of production and ownership of varieties. Our estimates of price and welfare changes from new varieties in the U.S. automobile sector are twice as large as standard estimates when using our detailed market-based data. We also show that new varieties introduced by foreign-owned affiliates provided an additional 70% welfare gain during our sample.


Empirical evaluations of international trade models with cross country heterogeneity have been limited by our estimates of trade elasticities. Specifically, identification of import demand and export supply elasticities has relied on strong assumptions of exporter homogeneity. This paper develops a tractable structural estimator that requires only readily available bilateral trade data in order to estimate heterogeneous export supply and import demand elasticities. Our elasticity estimates are shown to follow intuitive patterns of importer and exporter market power and produce believable distributions and magnitudes. The aim of the estimator is to be parsimonious enough to be applied to a wide range of studies assuming pairwise heterogeneity in trade. To highlight the flexibility of the estimator, we extend the cornerstone theories of optimal trade policy to a setting where exporters have heterogeneous supply elasticities. Applying our estimates, we show that heterogeneous export supply elasticities provide valuable insight into how countries apply tariffs over time in response to compositional changes in patterns of trade.

“Trade Elasticities, Heterogeneity and Optimal Tariffs,” Journal of International Economics, 114, September 2018: pp 44-62(Elasticity Estimates)

“From Selling Goods to Selling Services: Firm Responses to Trade Liberalization,” (Joint

with Holger Breinlich and Greg Wright), American Economic Journal: Policy, forthcoming.

Online Appendix

“Trade Restrictiveness Indexes and Welfare: A Structural Approach,” Working Paper,

Trade restrictiveness indexes (TRI) have become a staple for practitioners and policy makers to summarize international trade barriers.  TRIs theoretically found a measure of  trade barriers by calculating the uniform tariff that is welfare equivalent to the observed distribution of applied tariffs within a country.  This paper is the first to incorporate importer market power and exporter heterogeneity into calculations of TRIs and welfare globally.  To do so we structurally estimate a quantitative model of international trade.  The structure of the model allows tractable estimation of importer and exporter welfare and TRIs for every country in the world from 1990-2007.  Canonical estimates, which ignore exporter heterogeneity and importer market power, are shown to overstate efficiency losses from tariffs by a factor of 5 for the average importer.  Additionally, by not accounting for importer market power previous studies have failed to acknowledge  substantial welfare losses to exporters that are captured by importers through tariffs.  These channels are shown to significantly impact the measurement and interpretation of TRIs.

“Does the U.S. Export Global Warming? Coal Trade and the Shale Boom,” (Joint with Christopher Knittel, Kostantinos Metaxoglou and Andre Trindade).  Working Paper,

We examine the effect of the U.S. Shale Gas Boom on global trade and consumption of coal and CO2 emissions. We estimate a structural model that links the domestic to the international coal market and use it to simulate counterfactual scenarios. Our results show that the total quantity of coal traded around the world in the absence of the Boom is essentially the same as the actual. A compositional change towards dirtier coal could still have significant environmental effects; we show that this is not the case either. Hence, U.S. coal exports simply displaced other coal without affecting global emissions.