To compete with China, America needs the Endless Frontier Act

Tuesday, September 8, 2020

Issues in Science and Technology

Is the United States’ science and technology enterprise currently configured and funded to meet the challenge posed by China?

The future of the US economy and security hangs on that question. Yet the answer is no. Changing that should be a fundamental goal of US science policy.

Sadly, the question has received too little attention from policy-makers, who have tended to focus more on how to weaken China than on how to strengthen the United States. Meanwhile, China has made becoming a world leader in areas such as artificial intelligence an explicit national goal. It is lavishing the effort with money and setting up a system to enable its companies to capitalize quickly on technological advances. Although China’s plans do not guarantee achievement, one need look no further than 5G communications—where the United States has little to offer, and the entire West is in a weak position—to grasp the real possibility and the significance of falling behind. We should focus less on criticizing China’s ambitions and more on fostering our own.

Meeting these challenges requires a visible, focused, and sustained research effort in key technologies—a concentrated push on the research questions most likely to generate long-term technological progress. In effect, we need to fill the gap in the nation’s research system that opened with the disappearance of institutions such as Bell Labs, which did pathbreaking research that resulted in revolutionary technologies such as the transistor. And we also need to accelerate the movement of technological advances into the marketplace.

This requires a meaningful increase in federal funding for research; total federal research and development is now below 2003 levels in real dollars. But additional money is not enough. Just doing more of the same will not get us where we need to be.

One important step toward filling the critical gaps in our research system would be for Congress to pass the Endless Frontier Act. This bipartisan bill would authorize $100 billion over five years for a new Technology Directorate at the National Science Foundation to fund use-inspired basic research, primarily at universities, focused on advancing key technology areas.

Under the bill, NSF would also offer additional scholarships, fellowships, traineeships, and postdoctoral awards in fields related to the key technology areas, and would provide grants to enable universities to experiment with new ways to speed the transition from lab to market. It would also fund test beds for new technology and authorize $10 billion for the Department of Commerce to fund regional technology centers, in which state and local governments, industry, and universities would partner to promote economic development based on new technologies.

No single measure is going to address all aspects of the nation’s competitiveness. Complementary bills might, for example, help free up financing for new technologies or provide incentives for industry to engage in more long-term research. Other legislation might encourage researchers to pursue the incremental technical improvements that contribute to a healthy economy or on ways to advance manufacturing.

But critically, the Endless Frontier Act would boost the kind of research needed in key fields where we face stark competition. The United States cannot afford to observe from the sidelines while others invent the next transistor, or the next internet. We must be at the forefront of the kind of transformative advances that grow out of targeted research, but are years in the making. A common misreading of the bill is that it is designed to finance applied research or large engineering development projects; in fact, it is geared to support long-term, use-inspired basic research that will lay the groundwork for new technologies.

Some critics worry, understandably, that a new directorate would detract from NSF’s traditional mission of funding discovery research projects that bubble up from the scientific community. The bill would, after all, assign an additional mission to NSF, and the new directorate would need to approach research in different, more goal-oriented ways. But the bill makes clear that NSF’s traditional work is to continue, and the bill rightly creates a wall between the new and existing directorates. The funding of the existing directorates is protected, and they would also get 15% of the new funding.

The Endless Frontier Act plants a new technology mission at NSF precisely because it has been so successful. Moreover, NSF knows universities and how to work with them (and vice versa), which will be central to the technology directorate’s effectiveness.

A new directorate is not a mere organizational reshuffling. It will give NSF an expanded mission; a different way of approaching problems; to some extent, a different way of selecting and managing projects. That means it will need to be structured so that it can draw on, and contribute to, the mission of the existing NSF directorates while being separate enough to not change their valuable fundamentals.

Some current NSF programs do nod in the direction of technology development, and those efforts could remain in the existing directorates. They are no substitute for a new directorate that would have the sole purpose of fostering basic research explicitly to advance specific technological solutions, whether those solutions relate to artificial intelligence, climate change   or cyber security.

Some critics have argued that other agencies with specific governmental missions would be more appropriate to carry out this work. Certainly there should be more discussion of how agencies such as the Department of Energy, the National Aeronautics and Space Administration, and the National Institute of Standards and Technology could contribute to the goals of the legislation. But the aim here is to establish an entity with the clear responsibility to look broadly at the competitiveness needs of the nation—not to examine the government’s needs, as with most technology programs at other agencies. Because of the gap in our research enterprise, and the threat of global competition, “doing the same, just more of it” is an inadequate solution.

Another concern of the bill’s critics is that Congress would establish the list of initial key technology areas. But the bill deliberately identifies technology areas broadly enough to leave ample room for NSF, through its own consultative processes, to select the specific research questions and goals to pursue, altering them as a field evolves. In any case, there is nothing unusual or excessively political about Congress delineating broad research areas of national need; past legislation created research programs in high-performance computing, nanotechnology, and quantum information sciences. Those laws did not unduly constrain science or send scientists down unproductive or outmoded paths.

It should be encouraging that the Endless Frontier Act is rightly seeking focus at the new directorate. Spreading money thinly among many research areas, or constantly shifting pursuits, will not yield gains of competitive consequence. A technology directorate should be organized around solving a limited set of critical technological challenges in a cross-disciplinary manner with an effort sustained enough to produce results that industry or government can bring to scale.

With this interdisciplinary model, the new directorate could concentrate on projects that would give the United States a competitive edge. For example, within the broad field of artificial intelligence, NSF might define a goal of developing AI systems that require less data to be “trained,” thus eliminating China’s built-in data advantage from having a larger population and few privacy limitations. Such explorations might draw on research from existing research directorates at NSF—for example, inquiries about how babies learn and social science work on the impacts of AI.

Finally, some skeptical voices seem to assume that any aspect of the bill that hasn’t yet been fully fleshed out is simply being neglected. For example, it is true that universities will have to work with industry—or spin off new companies—to bring new ideas to market. The bill could say more about what that relationship might look like. But the bill does not neglect the subject; it explicitly allows industry to participate in university-led consortia, and it includes funding to identify better ways to enable university start-ups to mature their technology so that it is more likely to succeed in the market.

As a nation, we urgently need to concentrate our energies on a response to China that is equal to the task. Small-scale policy experiments and tinkering across multiple government agencies are unlikely to prevent the next 5G debacle or to yield the next industry-spawning invention. Some of the models cited as alternative approaches—such as creating additional Advanced Research Project Agencies—were greeted with just as much skepticism when they were first proposed. Tellingly, entities such as the energy-focused ARPA-E are now regularly seen as successes.

Putting together Science, the Endless Frontier 75 years ago, Vannevar Bush drew on his experience during World War II, but proposed something entirely novel—a federal agency with a unique structure that would fund university research to meet broad national needs. Like him, we need to be informed by history, not hamstrung by it. Of course, aspects of Bush’s vision were ultimately rejected, but if the whole concept had been shelved because it diverged from prior models, where would we be now? Those at universities who feared an influx of federal funding as too radical a shift now look timid rather than wise, even if some of their specific concerns had merit.

The Endless Frontier bill builds on what was best and most lasting in Bush’s vision while acknowledging where it may have fallen short. It is practically an axiom in science policy circles that some of Bush’s thinking was off the mark or no longer makes sense. Critics note that research and development is not a linear process; that the scientific community’s agenda can be too insular; that sometime to meet national needs, research needs to actually focus directly on those needs. Yet those critiques seem to vanish now that a proposal is being offered to expand the agency that grew out of Bush’s report so it can address those very weaknesses.

The United States is facing a kind of technological competition that it hasn’t encountered since becoming a world leader. The nation will not succeed if it tries to compete by changing as little as possible and reveling in past achievements.