Essays on intellectual property and innovation

Chapter 2 uses the universe of trademark registrations in the United States to better capture gains from trade. Specifically, trademarks help overcome one major shortcoming: the inability to disentangle production from the design origin of new goods when relying on trade data. I show that distinguishing between the country of design and production is crucial, since traditional trade data overstates variety contributions from manufacturing hubs like China while understating the role of high-GDP countries as designers. Moreover, I quantify the variety gains from trade, obtaining 2-3 times larger variety gains for the years 1995-2014 compared to using sectoral trade flows and country-of-origin pairs. Chapter 3 disentangles two channels through which import competition from China affects the introduction of new American varieties. I use trademarks and detailed Chinese customs data to analyze the origins of US imports from China. I find that only 40% of them originate from Chinese-owned firms located in China, while the rest originate from multinational enterprises operating in China. I show that imports from Chinese-owned firms located in China reduce the introduction of new American varieties. In contrast, imports from multinational enterprises located in China increase the introduction of new American varieties for US buyers. Chapter 4, a project co-authored with Pinelopi Goldberg, Rèka Juhàsz, Nathan Lane, and Jeff Thurk, examines industrial policies in the semiconductor industry, a critical driver of economic growth and innovation. Using historical analysis, natural language processing, and a model-based approach, we document global industrial policies, explore their rationale, and evaluate their cross-border impacts. Our findings show that subsidies have been vital to the industry's growth, with technology transfers across countries playing a crucial role. Preliminary results reveal modest learning-by-doing effects, estimated at 3%, which is about 15% of the magnitude estimated by Irwin and Klenow (1994). However, we find significant cross-border effects, meaning that firms benefit from chip production in other countries, for instance, through technology transfers.