Portrait of Chiara Castrovillari

Chiara Castrovillari

PhD Student in Economics, Cornell University

I am a PhD student in Economics at Cornell University, working on labor and urban economics, with a focus on innovation, entrepreneurship, and place-based policies. Before starting my PhD, I was a Research Assistant in the IMF's Strategy, Policy, and Review Department and earned an MPA from Brown University's Watson Institute. You can find my CV here.

Selected Works In Progress

What are the Labor Market Effects of U.S. State-Level R&D Tax Credits?

Draft available upon request.

Previous work has established a link between innovation incentives and overall investment in research and development (R&D). However, there has been little attention to the R&D inputs spent on — particularly labor. This paper exploits the staggered adoption of state-level R&D tax credits in the U.S. using a differences-in-differences design to estimate the effect of these incentives on the R&D labor market. The credits raise R&D employment by over 14% and increase the number of R&D establishments by about 4% while having no effect on wages. This has important implications for innovation and labor market policies.

The Geography of Entrepreneurship: Evaluating the Role of Learning and Sorting across Locations, with Elio Nimier-David

This paper examines how geography and career histories shape entrepreneurial entry and performance. Using French administrative data that link workers and entrepreneurs across employment and self-employment spells from 2006 to 2021, we reconstruct detailed career trajectories — including location, firm, industry, and wage histories — for a large sample of the French working population. We ask whether entrepreneurial individuals sort into cities, or whether living and working in larger locations induces entrepreneurship. We use a movers design and an extended dynamic AKM framework to disentangle sorting effects, city-specific and firm-specific experience, that induce entrepreneurship. This speaks to the relative importance of learning versus sorting in explaining the concentration of entrepreneurship in large cities, with implications for place-based policy.

Unpacking Agglomeration: Industrial Diversity and Knowledge Spillovers, with Eleonora Patacchini and Davide M. Coluccia

This paper studies how exposure to industrial diversity in urban locations shapes the direction of innovation. Existing theories of agglomeration economies disagree on whether knowledge spillovers arise from specialization within an industry or from diversity across industries, but seldom address which specific technological relationships between inventors and their surrounding environment affect innovation. We develop new, technology-based measures of industrial relatedness by using large language models to extract the downstream technological systems and functional roles described in the text of 2.47 million U.S. patents issued between 1850 and 1950, classifying pairs of industries as technologically complementary or substitutable. We then exploit the staggered construction of bridges across the United States between 1920 and 1940 as a shock to the network of industries an inventor is exposed to, comparing inventors with different predetermined exposure to complementary versus substitutable industries around each bridge opening.

Pre-Doctoral Research

Macroeconomic Shocks and Conflict, with Tomohide Mineyama and Patcharaporn Leepipatpiboon

IMF Working Paper, WP/23/68.

This paper contributes to the research on the macroeconomic origins of conflict. Based on a sample of 133 low- and middle-income countries over a 30-year period, it analyzes to what extent changes in a country's commodity terms-of-trade can explain an increase in the incidence and intensity of conflict through their effect on aggregate income. The paper finds that a negative commodity terms-of-trade shock leads to an increase in the number of conflict events and fatalities, an effect that plays out over several years and is twice as large for low-income countries and fragile and conflict-affected states compared with the sample average.