Research

Click here to download my latest CV.

Government R&D Contracts as Platforms for Experimentation (Job Market Paper) (Click here to download the latest PDF)

Startups experiment with technologies and markets to reduce uncertainty, but such experimentation is costly. While venture capital supports a narrow range of technologies, high-tech startups outside this scope often rely on government R&D contracts. This study examines the effect of experimenting with technologies and markets on startup performance. Using data on $42 billion in U.S. government R&D contracts awarded to high-tech startups between 1989 and 2019, I estimate effects on employment, sales, and exit outcomes. Identification leverages variation in eligibility for small business programs and fiscal year-end spending surges. Results show that experimenting with markets, rather than technologies, drives startup performance. Market experimentation improves private and public sales, diversifies product offerings, and increases exit likelihood—particularly for low-growth startups. In contrast, technology experimentation has limited or negative effects. These findings suggest that government R&D contracts can serve as platforms for market testing, shaping which startups grow and how innovation unfolds.

Guaranteed Demand and Corporate R&D (with Sharon Belenzon), NBER Working Paper 28644 (Click here to download the latest PDF)

The U.S. government incentivizes firms to develop innovative technologies by awarding R&D contracts that often carry an implicit promise of “guaranteed demand.” Firms that demonstrate strong technological capabilities are rewarded with noncompetitive production contracts for the resulting products and services. Using newly assembled data on $4.2 trillion in government procurement contracts from all federal agencies, matched to U.S. publicly traded firms, we document a “crowding-in” effect, where government R&D contracts lead to increased investment in corporate scientific research. This effect is concentrated in large, vertically integrated firms and limited to upstream R&D. We argue that these patterns are best explained by a guaranteed demand mechanism: firms co-invest in upstream research when success offers a credible path to future noncompetitive production contracts. We develop a theoretical framework to explain when it is optimal for the government to bundle R&D and production contracts. Our analysis shows that guaranteed demand can produce higher quality at a lower total cost for upstream R&D projects when the R&D firms have production capabilities. Our empirical results support these predictions. Additionally, we find that the crowding-in effect has weakened over time as the government has increasingly decoupled R&D contracts from production contracts. We discuss the potential implications of this decoupling.

The Effect of Public Science on Corporate R&D (with Ashish Arora, Sharon Belenzon, Lia Sheer, and Hansen Zhang), NBER Working Paper 31899 (Click here to download the latest PDF)

We study the relationships between corporate R&D and three components of public science: knowledge, human capital, and invention. We identify the relationships through firm-specific exposure to changes in federal agency R&D budgets driven by the political composition of congressional appropriations subcommittees. Our results indicate that R&D by established firms, which account for more than three-quarters of business R&D, is affected by scientific knowledge produced by universities only when the latter is embodied in inventions or PhD scientists. Human capital trained by universities fosters innovation in firms. However, inventions from universities and public research institutes substitute for corporate inventions and reduce the demand for internal research by corporations, perhaps reflecting downstream competition from startups that commercialize university inventions. Moreover, abstract knowledge advances per se elicit little or no response. Our findings question the belief that public science represents a non-rival public good that feeds into corporate R&D through knowledge spillovers.

Media Coverage: The Economist, Marginal Revolution

The Private Value of Innovating for the Government (with Ashish Arora, Sharon Belenzon, and Elia Ferracuti), NBER Working Paper 33880 (Click here to download the latest PDF)

We quantify the private returns to government R&D contracts awarded to firms. We present new evidence that R&D contracts not only finance innovation but also embed an implicit government guarantee of noncompetitive future procurement for the winning R&D contractor. We measure its private value by analyzing stock market reactions to news about R&D contract awards. Using all federal R&D contracts awarded to U.S. publicly traded firms from 1984 through 2015, we find that the average private return on an R&D contract is 19 times its maximum potential revenue. However, returns are highly skewed, with only 7.5% of firms receiving at least one top-quartile contract. Private returns are linked to future production contracts, but only for noncompetitive awards to vertically integrated or large firms. These results suggest that a procurement regime bundling R&D and production contracts enhances value for firms with production capability. We develop a conceptual framework to clarify this innovation policy lever.

Back to the Future: Are Big Firms Regaining Their Scientific and Technological Dominance? Evidence from DISCERN 2.1 (with Ashish Arora, Sharon Belenzon, Lia Sheer, and Dror Shvadron)

In recent years, breakthrough innovations in artificial intelligence, quantum computing, and robotics have captured the attention of both the scientific community and the public. Do these advances signify a break in the long-standing trend of declining investment in scientific research within corporate R\&D? We answer this question using DISCERN 2.1, an extensive update to the original DISCERN dataset. By adding nine years of coverage, we extend the dataset to 2024, incorporate improved firm and subsidiary data, and integrate open-access scientific publications and patent data using the OpenAlex and PatentsView datasets. These improvements result in a more comprehensive and accurate coverage, evidenced by increases in the number of ultimate owner parent firms, subsidiaries, patents, and scientific publications. Our analysis reveals that the breakthrough innovations of the last decade represent unique opportunities in a limited number of fields, pursued by a select group of companies rather than a broad resurgence in corporate research investment.

Work in Progress

Bundling Upstream Research with Downstream Production (with Sharon Belenzon, Honggi Lee, and Jungkyu Suh)

The Role of Startups in Translating University Science (with Sharon Belenzon and Sharique Hasan)