Research

Job Market Paper

“Stellar Skills: Superstar Firms and the Specialization of Human Capital” (PDF)

Technological change is the key driver of sustained economic growth, but new ideas can only transform production if the supply of technology-specific inputs supports their adoption. This paper studies how imperfect competition in markets for technology-specific capital shapes the direction of firms' technology adoption and workers' specific human capital accumulation. I argue that when production requires combining technology-specific capital and labor, providers of specialized capital have a direct stake in the formation of complementary skills. I develop a general equilibrium model in which dominant capital providers set rental rates and subsidize training in the skills required to operate their systems, internalizing their effect on the market for technology-specific labor. The framework features a policy-relevant trade-off: market power in markets for specialized capital raises misallocation yet can stimulate investment in high-quality training. I quantify how the balance between these forces shapes technological specialization and welfare, showing that stronger competition in input markets is not always welfare-improving, particularly when workers’ preferences make skills less substitutable. I test the theory in the digital infrastructure industry. Using large-scale job posting microdata and developer surveys, I study the open source release of platform-specific software by a major supplier of cloud computing power. I observe patterns consistent with my theory of subsidies to skill acquisition costs: the release of specific software is associated with increased adoption of the complementary platform, but also a lowered wage premium for the related specialized skills.

Other Working Papers

“Market Entry and Plant Location in Multiproduct Firms” (with Juanma Castro-Vincenzi, Eduardo Morales and Alejandro Sabal) (Slides) (Draft available upon request)

We develop a quantitative model in which firms decide where to produce and sell each of their products. Cannibalization forces imply that our model exhibits substitutabilities in the firm's decision to sell different products in the same destination market; transport costs increasing in distance and increasing returns at the plant-product level imply our model displays complementarities in the firm decision to produce and sell the same model in geographically close markets. We provide a novel solution algorithm for multiple discrete choice single-agent problems where the sign of the interdependence between any pair of choices is known. We estimate our model using our solution method and data on the global car industry with information by firm, country, and car model on production, sales, and prices. We evaluate the effect of recently proposed production subsidies, consumption subsidies, and tariffs on the global structure of automobile production and prices and access to car varieties across countries.

“Risk-based Organizational Choices” (Draft coming soon)

This paper studies how production uncertainty shapes firms’ organizational boundaries and, more broadly, the allocation of control within global supply chains. I develop a theoretical model of vertical integration under uncertainty about supplier performance and show that the gains from integration increase with the variance in supplier reliability. The model highlights a novel channel through which supply chain risk affects firm structure: integration can serve as a mechanism to insure against input disruptions by allowing the buyer to respond faster to productivity shocks that hit the supplier. Using establishment-level data, I find empirical support for this mechanism. Multinational firms operating in environments with greater supplier performance volatility tend to own a larger number of subsidiaries. The results contribute to the debate on supply chain resiliency by linking risk exposure to the organization of production within global firms.

Work in Progress

“Product and Labor Market Power: Evidence from Firm-to-Firm Networks” (with Chiara Motta)(Extended Abstract)

Resource misallocation is a central driver of productivity differences across countries, yet distortions in product and labor markets are typically analyzed in isolation. We develop a general-equilibrium model of heterogeneous firms that compete imperfectly in both input and output markets, extending the standard misallocation framework to allow for correlated markups and markdowns. We show that even modest distortions can generate large aggregate productivity losses when product and labor market power coincide, suggesting that policies targeting product market competition or labor market regulation in isolation underestimate potential efficiency gains. We quantify these effects using newly linked Italian administrative data that combine firm-to-firm transactions, corporate balance sheets, and matched employer-employee records. These integrated sources allow us to infer firm-level markups and markdowns, as well as market-level concentration on both sides of the economy. Armed with these data, we quantify the degree of correlation between product and labor market dominance.

“Firms’ AI Skills and the Direction of Product Innovation”(with David Argente and Sara Moreira)