The University Dividend: Capital, Research, and Regional Renewal
Can Policy Effectively Bridge Innovation Supply with Investment Needs?
A fundamental shift in how university inventions were handled after 1980 acted as a powerful magnet, drawing venture capital investment to specific regions and industries aligned with university research strengths, according to new research ("Innovation and capital" by Daniel C. Fehder, Naomi Hausman, and Yael V. Hochberg). This suggests focusing policy on boosting innovation supply might be more effective than just trying to attract capital directly.
Why it matters
Venture capital is highly concentrated (e.g., top 3 cities got 60% of US VC in 2019), and policymakers often struggle to cultivate innovation hubs in other regions. This study explores whether boosting the supply of local, commercializable ideas is a key missing piece.
The 1980 Bayh-Dole Act dramatically changed the landscape for university research commercialization.
Before: Ownership of inventions from federally-funded research (the majority of university research) typically rested with the government. Universities had weak incentives and faced hurdles (like negotiating individual agreements) to patent or license discoveries. Only about 5% of government-owned patents were licensed.
After: Bayh-Dole gave universities the rights to own, patent, and keep royalties from inventions derived from federal funding. It required them to promote commercialization and share royalties with inventors, spurring the creation of Technology Transfer Offices (TTOs) and a surge in university patenting (from <300 in 1976 to >2000 by 1990).
The Setup: The study uses a difference-in-differences approach, leveraging a clever identification strategy. Since universities have diverse research specializations (e.g., UT Austin strong in engineering, Johns Hopkins in biosciences pre-1980), Bayh-Dole created varied shocks across regions and, critically, within regions across different industries based on the local university's pre-existing strengths. By comparing industries within the same county that were more vs. less related to the local university's expertise (using an "innovation index" based on pre-1980 university patent fields and industry relevance), researchers could isolate the impact of the innovation supply shock, controlling for broader geographic and industry trends using extensive fixed effects (County x Industry, County x Year, Industry x Year).
Data & Insights
The analysis tracks US patent data (NBER, USPTO, augmented with Gross & Sampat data), NSF federal funding records, and venture capital deals (from VentureXpert) at the county-industry-year level from 1970-1990.
VC Follows University Innovation: Post-1980, VC funding, the number of deals, and the number of active investors systematically increased in geographic areas and specific industries most affected by the newly accessible university innovation. Event studies confirm these trends weren't apparent before 1980.
A one standard deviation increase in a county-industry's pre-1980 university "innovation index" led to $109,000 more annual VC funding after 1980. This also corresponded to statistically significant increases in the number of local VC deals (+0.038) and active investors (+0.102).
Looking at raw patent influence: each citation-weighted pre-1980 university patent corresponded to an increase of $45,000 in annual VC funding in that specific county-industry post-1980, plus related increases in deals (+0.016) and investors (+0.042).
Elasticities: A 1% increase in the innovation index yielded a 2.7% to 3.9% increase in VC dollars, deals, and investors.
Quantifying the Impact: The Bayh-Dole innovation shock appears to account for over 16% of the total increase in VC funding flowing into these university counties between the 1970s and 1990. (This rises to over 1/3 if the early 90s VC boom is included).
Federal Funding as a Mechanism: The positive effect on VC was significantly stronger in areas where universities received more federal research funding before 1980 (triple-interaction model), supporting the idea that Bayh-Dole's impact was tied directly to unlocking this specific pool of federally funded research.
Geographic Concentration: Effects were strongest in the university's home county, decaying significantly in adjacent counties and further out (tested up to 75 miles), consistent with the local nature of knowledge spillovers and known VC investment preferences (monitoring/advising ease).
Local Conditions Matter: The positive impact of university innovation on VC attraction was amplified in counties that already had higher median incomes, more college graduates, larger populations, higher wages, greater urbanization, and larger average establishment sizes pre-1980, suggesting absorptive capacity matters.
"Horse Race" Results: When comparing predictors of post-1980 VC flows, the university innovation index strongly and significantly predicted VC attraction. Indices measuring pre-existing VC activity levels (scaled by R&D, à la Kortum & Lerner) and pre-existing corporate patenting levels showed much weaker or insignificant predictive power in the same models. This suggests the new supply of university innovation was a primary driver, not just path dependence.
Broader Innovation Spillovers: Beyond attracting VC, pre-1980 university innovation strength also significantly predicted where corporate patenting would increase after Bayh-Dole, even controlling for prior corporate patenting activity. This aligns with findings that government R&D can seed persistent clusters.
High-Growth Entrepreneurship Link: Using Startup Cartography Project data (1988-1995), the study found university counties (especially those with high pre-1980 federal funding) had significantly more high-growth entrepreneurial activity (firms with eventual IPOs/acquisitions). Controlling for contemporary VC investment significantly reduced the direct statistical link between university innovation and entrepreneurship outcomes, suggesting VC acts as a key intermediary, identifying and funding promising ventures emerging from the university ecosystem.
Between the lines: The Bayh-Dole Act coincided with regulatory changes (like ERISA's "Prudent Man" rule clarification and Small Business Act changes favoring LPs) that significantly increased the total capital available to VC funds. This research shows that the supply of commercializable innovation from universities played a key role in directing where this influx of capital was deployed, towards regions and industries benefiting from the newly accessible frontier research, rather than solely reinforcing existing corporate strongholds or prior VC hubs.
My Notes
This research highlights an important sequence: Public funding generates research, and then policy choices, like the Bayh-Dole Act, fundamentally reshaped how these ideas became investment opportunities. Financial firms, particularly Venture Capital (VC), subsequently played a central role in directing investment towards these innovations.
Universities are key players here, they are underused tools to help boost local economies or specific industries. However, using them effectively for that purpose isn't easy; it takes more than just letting them patent inventions, as local conditions and planning matters.
The heavy reliance on VC funding also raises questions. VCs aim for big financial wins, which might not always line up with the public good goals behind the original research (like broad benefits or helping struggling regions). This makes it important to think about other ways to fund these new ideas, perhaps through different kinds of banks or investment funds focused more directly on public benefits.