Working Papers

Trademarks and Gains from Variety: The Role of Multinational Enterprises

Winner of the 2024 Best Job Market Paper Award (European Economic Association and UniCredit Foundation).

Abstract | Paper

Novel differentiated products, or varieties, are a key source of welfare gains from globalization. Yet standard measures based on trade data cannot separate production from the design origin of goods. However, imports are not always "foreign" varieties as we might imagine them, since firm boundaries often span national borders. I address this limitation using U.S. trademark registrations, which provide a consumer-oriented measure of varieties tied to design rather than production. Two findings emerge. First, variety gains from trade between 1995 and 2014 are two to three times larger when measured with trademarks than with sectoral trade flows. Second, by combining trademark data with Chinese customs records, I re-evaluate the effect of imports on product entry. Imports from Chinese-owned firms reduce the introduction of new U.S. and foreign products, while imports from multinationals located in China increase it. These results show that multinational activity biases standard estimates of variety gains and reshape our understanding of how import competition affects innovation.

Industrial Policy in the Global Semiconductor Sector (with Pinelopi Goldberg , Réka Juhász , Nathan Lane , and Jeff Thurk )

Coverage: NBER Digest .

Abstract | Paper | NBER WP | CESifo WP | CEPR DP

The resurgence of subsidies and industrial policy has raised concerns about their efficiency and compatibility with multilateral principles. Critics argue that such policies may divert resources toward less efficient firms and trigger retaliation by other countries, potentially generating a wasteful "subsidy race." At the same time, subsidies may generate global benefits in sectors characterized by dynamic economies of scale and cross-border effects. Assessing these competing arguments requires accurate measurement of the policies governments actually pursue. This paper undertakes that task for the semiconductor industry -- a key driver of economic growth and innovation, with potentially important dynamic economies of scale and strategic significance arising from its dual-use applications. We document and quantify recent industrial policies in the global semiconductor sector and examine the economic rationale underlying them. Combining historical analysis, natural language processing, and a model-based approach, we measure the extent and composition of government support across countries and over time. We find that government support has played an important role in the industry's development, with subsidies constituting its predominant form. Although China provides substantial subsidies to its semiconductor sector, the scale of its support is not exceptional relative to that of other countries once differences in market size are taken into account.

Trade Deflection: The Spillover Effects of U.S. Tariffs on Import Prices in Bystander Countries (with Sofie Grouws, Sharon Jeon, and Justin Pierce )

Abstract | Paper

Even as U.S. imports from China have fallen sharply since tariffs starting in 2018, Chinese exports have continued to expand globally. But despite fears in many countries about a potential surge of lowpriced Chinese imports redirected from the U.S. market, there is little systematic evidence on how trade flows and import prices have responded in these bystander countries. This paper estimates the extent to which the U.S. tariffs deflected Chinese exports to bystander countries. We show that products targeted by U.S. tariffs experienced significant reallocation of Chinese exports toward other destinations, especially emerging markets, countries in Asia, established Chinese markets, and low-price markets. In addition, these higher trade flows arrive at lower prices, a finding with implications for both local producers and monetary policymakers.

How Knowledge Crosses Borders: Evidence from Semiconductor Manufacturing (with Pinelopi Goldberg , Réka Juhász , and Jeff Thurk )

Abstract | Paper | NBER WP | CESifo WP | CEPR DP

We show that cross-border learning is an important feature of modern semiconductor manufacturing and that several important channels – such as knowledge sharing along global supply chains, the movement of engineers within multinational firms, and foreign direct investment – operate through market mechanisms rather than anonymous technological spillovers. These findings have significant policy implications. When cross-border learning is mediated by purposeful economic relationships, access to frontier knowledge is neither automatic nor guaranteed. Firms and governments can facilitate - or restrict - the international diffusion of technology, implying that policies to foster domestic technological upgrading are likely to be most effective when coupled with continued access to global sources of frontier knowledge.

Works in progress

Human Capital Spillovers Across Cities (with Pablo Valenzuela-Casasempere )

Abstract | Slides

This paper studies human capital spillovers among Spanish workers. Leveraging administrative matched employer-employee data for Spain between 2006 and 2020, we study whether the experience gained by workers in larger cities transfers to workers in smaller cities when they relocate. We find that, for workers in medium-sized cities, their three-year wage growth is 1.2 log points larger for each year of experience their coworkers accumulated in a large city. This positive knowledge spillover may mitigate the negative welfare effects of restricting the growth of large cities – an avenue we plan to explore in future research.

Assortative Matching in Patent Production

Abstract | Paper

This paper studies the collaboration process of inventors using the universe of patents filed at the European Patent Office between 1978 and 2017. I develop a one-factor static Roy model predicting that inventors sort into leaders and supporters based on an endogenous cut-off rule, and then match together to produce patents in teams of different size. Consistent with the model predictions, empirically I find that more productive leaders and supporters sort into bigger teams. Most importantly, leaders' productivity exhibits a strong positive correlation with the productivity of supporters working with them, but this correlation decreases in magnitude as team size increases. The extent of the assortative matching is significant: in the most conservative case, a one standard deviation increase in leaders' productivity is associated with doubling the average supporter productivity. The paper concludes with empirical evidence on the life-cycle of inventors, who move to bigger teams as they age and become leaders after gaining experience and learning from working on previous patents.