Search

This article studies how federal funding affects the innovation outputs of university researchers. We link person-level research grants from 22 universities to patents, publications, and career outcomes from the U.S. Census Bureau. We focus on the effects of large, idiosyncratic, and temporary cuts to federal funding in a researcher’s preexisting narrow field of study. Using an event study design, we document that these negative federal funding shocks reduce high-tech entrepreneurship and publications but increase patenting. The lost publications tend to be higher quality and more basic, whereas the additional patents tend to be lower quality, less general, and more often privately assigned. These federal funding cuts lead to an increase in private funding, which partially compensates for the decline in federal funding. Together with evidence from industry-university contracts, the results suggest that federal funding cuts shift university research funding from federal to private sources and lead to innovation outputs that are less openly accessible and more often appropriated by corporate funders.

We examine innovation following the Great Depression using data on a century’s worth of U.S. patents and a difference-in-differences design that exploits regional variation in the crisis severity. Harder-hit areas experienced large and persistent declines in independent patenting, mostly reflecting the disruption in access to finance during the crisis. This decline was larger for young and inexperienced inventors and lower-quality patents. In contrast, innovation by large firms increased, especially among young and inexperienced inventors. Overall, the Great Depression contributed to the decline in technological entrepreneurship and accelerated the shift of innovation into larger firms.

Heterogeneity in the taxation of asset returns can create ownership clienteles. Using a simple model, we demonstrate that an important consequence of tax-induced ownership segmentation is to limit risk-sharing, creating regions of the aggregate demand curve for the asset that are "downward-sloping.'' As a result, the constraints of the ownership clientele impact the asset price response to variations in asset supply, and make the asset's price more sensitive to movements in idiosyncratic risk. We test these predictions on U.S. municipal bonds, where cross-state variation in state tax privilege results in different levels of in-state ownership. In states with high tax-induced ownership segmentation, we find greater susceptibility of municipal bond yields to supply variation and heightened sensitivity of muni yields to local political uncertainty.

U.S. universities, which are important producers of new knowledge, have experienced a shift in research funding away from federal and towards private industry sources. This paper compares the effects of federal and private university research funding, using data from 22 universities that include individual-level payments for everyone employed on all grants for each university year and that are linked to patent and Census data, including IRS W-2 records. We instrument for an individual’s source of funding with government-wide R&D expenditure shocks within a narrow field of study. We find that a higher share of federal funding causes fewer but more general patents, more high-tech entrepreneurship, a higher likelihood of remaining employed in academia, and a lower likelihood of joining an incumbent firm. Increasing the private share of funding has opposite effects for most outcomes. It appears that private funding leads to greater appropriation of intellectual property by incumbent firms.

U.S. universities have experienced a shift in research funding away from federal and towards private industry sources. This paper evaluates whether the source of funding – federal or private industry – is relevant for commercialization of research outputs. We link person-level grant data from 22 universities to patent and career outcomes (including IRS W-2 records). To identify a causal effect, we exploit individual-level variation in exposure to narrow federal R&D programs stemming from pre-existing field specialization. We instrument for the researcher’s funding sources with aggregate supply shocks to federal funding within these narrow fields. The results show that a higher share of federal funding reduces patenting and the chances of joining an incumbent firm, while increasing the chances of high- tech entrepreneurship and of remaining employed in academia. A decline in the federal share of funding is offset by an increase in the private share of funding, which has opposite effects. We conclude that the incentives of private funders to appropriate research outputs have important implications for the trajectory of university researcher careers and intellectual property. Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

We study the use and economic impact of artificial intelligence (AI) technologies among U.S. firms. We propose a new measure of firm-level AI investments, using a unique combination of detailed worker resume and job postings datasets. Our measure reveals a stark increase in AI investments across sectors in the last decade. AI-investing firms see higher growth in sales, employment, and market valuations. We use a novel identification strategy, instrumenting firm-level AI investments with firms' ex-ante exposure, based on alumni networks, to the supply of AI-skilled labor from universities historically strong in AI research. The positive growth effect of AI comes primarily through increased product innovation, reflected in trademarks, product patents, and updates to product portfolios. AI-powered growth concentrates among the ex-ante largest firms, leading to higher industry concentration and reinforcing winner-take-most dynamics. Our results highlight that new technologies can contribute to growth through product innovation.

Why do young firms pay less? Using confidential microdata from the US Census Bureau, we find lower earnings among workers at young firms. However, we argue that such measurement is likely subject to worker and firm selection. Exploiting the two-sided panel nature of the data to control for relevant dimensions of worker and firm heterogeneity, we uncover a positive and significant young-firm pay premium. Furthermore, we show that worker selection at firm birth is related to future firm dynamics, including survival and growth. We tie our empirical findings to a simple model of pay, employment, and dynamics of young firms.

This paper studies how corporate research and development (R&D) investment affects labor mobility. We use employer-employee matched data in ordinary least squares and instrumental variables analyses to assess four hypotheses. R&D has no effect on worker retention, exit from employment, or mobility to incumbent firms. Instead, it increases employee departures to entrepreneurship, leading employees to join the founding teams of startups that are venture capital-backed, high tech, high wage, and in different sectors than the parent firm. These high-growth, high-risk startups emerging from R&D benefit from a focused, standalone incentive structure and have poor complementarities to the parent firm’s assets.

Heterogeneity in the taxation of asset returns can create ownership clienteles. Using a simple model, we demonstrate that an important consequence of tax-policy-induced ownership segmentation is to limit risk-sharing, creating regions of the aggregate demand curve for the asset that are "downward-sloping." As a result, the constraints of the ownership clientele impact the asset price response to variations in asset supply and demand, and make the asset's price more sensitive to movements in idiosyncratic risk. We test these predictions on U.S. municipal bonds, where cross-state variation in state tax privilege policies results in different levels of home-state-biased ownership of local municipal bonds. In states with high tax-induced ownership segmentation, we find greater susceptibility of municipal bond yields to demand and supply variation, heightened sensitivity of muni yields to local political uncertainty, and greater difficulties in raising capital for public projects.

We test whether network connections to other firms through executives and directors increase value by exploiting differences in survival rates in response to a common negative shock. We find that firms that had more connections on the eve of the 1929 financial market crash have higher 10-year survival rates during the Great Depression. Consistent with a financing channel, we find that the results are particularly strong for small firms, private firms and firms with small cash holdings relative to the sample median prior to the shock. Moreover, connections to cash-rich firms are stronger predictors of survival, overall and among financially constrained firms. Because of the greater segmentation of markets in the 1920s and 1930s than in modern data samples, we can mitigate the potential endogeneity of network connections at the time of the shock by exploiting variation in the local demand for directors’ services. We also find evidence that the information that flows through network links increases the odds that a firm will be acquired.

Why do employees become entrepreneurs? This paper shows that firm financial distress drives the exit of workers to pursue entrepreneurship. In a difference-in-difference setting, I find that, following unexpected industry shocks, employees at relatively more financially levered public firms are more likely to exit to found new firms. These new firms are created because financially distressed firms are less able to retain productive workers who exit to found start-ups. Entrepreneurs exiting financially distressed employers earn higher wages prior to leaving paid employment and after founding start-ups, compared to entrepreneurs exiting non-distressed firms. Consistent with financial distress driving productive workers into entrepreneurship, start-ups created following financial distress have high future employment growth. Distressed firms are less able to retain entrepreneurial workers in part because ex ante contracts restricting employee mobility are not enforceable. In support of this argument, I find that the effect is concentrated in states with weaker enforcement of non-compete agreements. The results suggest that the social costs of distress might be lower than the private costs to financially distressed firms.