Congress is set to make a down-payment on innovation in America

SENATOR HARLEY KILGORE, a West Virginia oil prospector’s son who carried around a horse chestnut for good luck, had a vision for American science. It was too dominated, he thought, by big business and by the university system: the country’s practical needs were an afterthought. In 1942 Kilgore proposed creating a federal bureaucracy, responsive to the public, that would guide scientific research for the good of the country and distribute its benefits geographically.

Kilgore was opposed by Vannevar Bush (pictured above), who led American R&D during the second world war. Bush felt that scientific research should be directed by the scientists themselves. In a report for the president called “Science: The Endless Frontier”, Bush summarised his ideas. Government, he said, should fund research. But rather than direct this research towards meeting social needs, it should instead seek to advance science for its own sake: basic, not applied, science was to be the primary objective. Bush won the day. The National Science Foundation (NSF), born in 1950, has largely followed the principles he laid out.

Kilgore is about to get his revenge. The Senate will probably soon pass the US Innovation and Competition Act, known until recently as the Endless Frontier Act. Though the bill is named after Bush’s report, it will take American science policy in a more Kilgorian direction. It lays out ten “key technology focus areas”, such as artificial intelligence, biotechnology and advanced materials science, to which new research funding will be directed. It allocates funding for regional tech hubs spread across the country. And its objectives are clear: the goal, in true Kilgorian fashion, is to “enhance the competitive advantage and leadership of the United States in the global economy”.

When it was first introduced in May 2020, the Endless Frontier Act planned to set aside $100bn for a new Directorate for Technology and Innovation within the NSF. This would have borrowed characteristics from the Defence Advanced Research Projects Agency (DARPA), the military-research office responsible for spearheading research that led to the internet, the computer mouse and mRNA vaccines.

The act’s ambitions have since narrowed. Instead of the full $100bn, the NSF’s new tech directorate will get $4bn. Some of the money has gone to pork. A significant portion has been diverted to the Department of Energy’s national labs. Though more than $50bn of funding will go to NSF, much of it either replaces existing funding or is earmarked for causes other than R&D, such as STEM education.

Innovation experts advise against looking a gift horse in the mouth, however. Federal spending on research has fallen from more than 1.2% of GDP in 1976 to less than 0.8% today. As a portion of the federal budget, it has dropped from 12% at its 1960s peak to 3%. The money set aside for R&D in the new bill will not reverse this slide. But the law will still deliver substantial additional funding to the NSF: Its budget for 2022 will be 27% higher than in 2021 and will double over the next five years. Jonathan Gruber of the Massachusetts Institute of Technology, whose work helped to spur the legislation, views the act as a “down payment” towards future innovation.

Others view it as a missed opportunity. Samuel Hammond of the Niskanen Centre, a think-tank, acknowledges that the influx of cash is valuable, but pines for what might have been. Science, he says, needs not just new funding but new institutions as well. Some researchers spend more than 40% of their time on administrative tasks such as grant-writing. Studies have found grant evaluations are inconsistent and subjective. Since the number of grant applications has increased faster than available funding, high-quality work may languish unfunded. And though scientists tend to do their best work in their younger years, the recipients of research grants have been getting steadily older.

Mr Hammond believes that government agencies that fund research have become sclerotic: he sees a “compliance culture” resulting from a risk-averse leadership wary of heavy-handed congressional oversight. That is a problem, says Benjamin Reinhardt, an independent researcher who has studied DARPA, because big wins come from taking risks. “All the value,” he says, “is in the long tail.”

One reason to create new research-funding institutions is to turn the scientific process on itself. Some economists have suggested prizes for big breakthroughs. New Zealand has experimented with lotteries for grant funding. Two researchers, Adam Marblestone and Sam Rodriques, have proposed Focused Research Organisations, stand-alone research efforts concentrated on solving single, well-defined science or technology problems.

Mr Gruber agrees that existing funding agencies are too conservative, and wishes the bill were bigger. But he believes it is a good start. The promotion of regional tech hubs, he says, could result in a virtuous cycle: once science and technology are no longer concentrated on the coasts, Americans may become more receptive to increases in R&D funding in the future. And, in the selection of ten key technology areas to focus on, he sees the beginnings of a less tentative approach to innovation. “You can call it picking winners,” he avers. “I call it taking risks.” ■

For more coverage of Joe Biden’s presidency, visit our dedicated hub