The prospect that China will become dominant in R&D has made Congress eager to spend more money on science—but history shows that creates problems of its own.
At last, it seems that the American scientific community has found a new Sputnik. However, unlike that fateful day in October 1957 when the Soviet Union launched the world’s first artificial satellite, this new Sputnik is not a single object or event but instead a decades-long trend that has cumulated to what senators now call a “crisis,” one requiring a massive federal financial response. The new Sputnik is China. More precisely, it is the prospect that China will come to dominate the current century’s emerging high-technology industries and threaten America’s global standing.
The need to spend money on this “crisis” is one of the few things upon which the deeply divided Senate can agree, and so it recently passed the United States Innovation and Competition Act of 2021 (S. 1260) by a bipartisan 68–32 vote. Both the text of the bill itself and the comments of numerous senators frame their concerns as a zero-sum competition with China, one that can only be won through direct government support for these cutting-edge industries and that will determine which country will dominate the twenty-first century.
But 2021 is not 1957, and China’s technological ascendance is hardly a solitary Cold War satellite flung into the sky. So what are we to make of this? By looking carefully at the past 70 years of US science and technology policy, we can see that money alone will not deliver the transformations we desire. Rather than mimicking the panicked responses of the past, we should revise our strategies and institutions for the current moment and the future.
First, note that Chinese officials and commentators reacting to this “Sputnik moment” seem none too happy with the way the Senate has framed it. They object most overtly to sections of the bill that criticize China over various human rights and trade issues. But beyond those specifics, they also complain that the bill embodies a Cold War mentality, pushing up the level of competition between the countries to the point of hostility to China.
And yet, S. 1260 should be seen as a tremendous compliment to China and an acknowledgement of all that it has achieved in the last several decades. The legislation does not seek to counter Chinese advantages in cheap labor or lax environmental standards, but instead casts the threat from China as one of technological superiority. This is a situation the United States has not faced, or even imagined, since the Soviets launched Sputnik. One of the bromides of American politics is that it takes a serious threat, preferably an external one, to motivate policymakers to undertake a new, sweeping, and expensive program. China in 2021 serves that domestic US political function: it’s nothing personal.
The Sputnik framing does, however, pose potential problems for the United States. It’s understandable that scientific leaders might look with nostalgia on the decades immediately after World War II, seeing that period as a Golden Age for American science and technology. The policies and institutions that emerged during that period, the unprecedented expenditure of federal funds for research and development, and the immense obstacles other industrial countries faced as they rebuilt after the war made the United States a global scientific and technological hegemon.
Nonetheless, that postwar party could not go on forever. And over time the vaunted policies and institutions themselves created the conditions for many of the problems—such as declining success rates for grant proposals and a glut of postdocs—that the American scientific community has faced since funding flattened and declined. What is the point in being nostalgic for the 1950s if you only end up in the 1970s?
Which decade will S. 1260 land us in? It’s impossible to predict the bill’s ultimate effects. For starters, it will change as it goes through the House and conference committee and as appropriations committees decide how much money it actually gets. But we can look to the past to see how science policies of this ilk have played out. And over these decades, we’ve observed that not only does R&D fail to solve many economic and social problems, but also that more money does not even reliably solve all R&D problems—though working in a system with declining budgets presents more, and usually worse, pathologies than working in one with rising budgets.
Many science policy analysts and elected officials assume that technological innovation drives economic growth, protects public health, and ensures a nation’s security. That formulation has always been far too simple: those social goods require much more than new technologies, as important as they are. In addition, it is no simple thing to figure out who should fund R&D, how much money should go into which topics, or which features of the institutional landscape transform research results into practical benefits.
The history of technologies such as photovoltaic (PV) solar panels teaches us that simply inventing new technologies in the United States is no guarantee that American workers and firms will benefit from making and selling them. American firms did the earliest research on photovoltaics, and the US government greatly expanded its funding for PV R&D starting in the 1970s. Despite this first-mover advantage, the United States did not retain its dominance of the industry, as manufacturing and even important R&D began moving to Europe and Asia. By 2019 only 9.8% of PV modules available in the United States were manufactured here. It takes more than R&D spending to make a country prosperous, healthy, and secure.
Thus, I think rather than trying to put the age of Sputnik on rerun entirely by increasing government spending, this legislation will be successful if it promotes experimenting with new ways of targeting and disbursing federal R&D funds, actions that will be both welcome and overdue.
One fundamental part of experimenting with such processes will be reaching an understanding of how much more funding science needs. That would require having a clear understanding of how much or in what sense the United States is “behind” other countries. By aggregate measures, combining private and public funding from all sources, the United States is doing quite well in terms of R&D spending. In late March, the Organisation for Economic Co-operation and Development (OECD) released a short “R&D Highlights” note reporting that in 2019 the United States’ total national spending on R&D reached 3% of gross domestic product for the first time. This level of spending does not make the United States the most R&D-intensive economy; it trails far behind countries such as Israel and Korea and is ninth overall. Nonetheless, the note calls the 3% level a “milestone” and presents data showing a steady upward trend since 2015 in national R&D spending. That 3% R&D intensity number is larger than, or at least close to, that of other large industrial countries; and, in fact, the absolute amount of American spending on R&D exceeds that of any other country—although China is moving up quickly.
What is distinctive about US R&D spending is that about 71% of it comes from the private sector, a higher rate than that of most other countries except, notably, China. This high percentage could be a problem if it means that the overall enterprise of technological innovation has an excessively short-term focus, but if so, just putting more money into government funding will do little to solve this problem of balancing short- and long-term R&D investments.
Managing increased flows of government money into this complex R&D system will not be a small task. The somewhat ironic result of the “new Sputnik” framing is that we may be on track to exceed even “old Sputnik” levels of funding. President Biden’s proposed fiscal year 2022 budget includes increases for R&D funding for health, energy, NASA, and the National Science Foundation (NSF). If the president were to get his request funded in full, federal R&D budgets would increase by $13.5 billion, a 9% increase over FY2021. In addition to President Biden’s request, S. 1260 authorizes roughly an additional $200 billion in R&D funding spread over five years. Adding these two funding boosts together would make for a 23% increase in federal R&D spending in the first year. To put that into historical perspective, in the first full fiscal year after Sputnik, federal funding for R&D increased by “only” 17%.
There’s no doubt that researchers, especially those at universities, would welcome these unprecedented budget increases, but the situation could raise a new set of problems down the road. Five years may be a long time in politics but it’s short in research, and it is not clear that funding will be reauthorized when the five years runs out. Will the continuing threat of China cause even larger appropriations to flow? If not, will the research teams that the new money created then be disbanded? If the surge in funding induces ever-greater demand for R&D funds, the easing of the pressures on the R&D system will be short lived.
The most promising part of S. 1260 is where it creates new structures and rules, not merely new budgets. As others have observed, this bill nudges federal funding toward a more applied orientation, and it does so while experimenting with different modes of funding. The core of the bill is a heavy focus on applied research on technologies that senators believe will be central to newly emerging industries. NSF gets a new directorate for innovation and technology, and the bill enlarges on the DARPA and ARPA-E model by applying it to other policy goals. The bill also proposes to set up a series of technology innovation hubs around the country aimed at linking research universities to industries and catalyzing economic growth, especially in regions suffering from deindustrialization. In addition, although both Republican and Democratic advocates for the bill deny that it constitutes “industrial policy,” there can be little doubt that it does, and even the bill’s Republican sponsor acknowledges that it breaks with conservative orthodoxy on economic policy.
The advantage of the original Sputnik threat was that it consisted of a discrete technological achievement. Once the United States successfully launched the Vanguard 1 satellite, even after a string of embarrassing failures, it regained some sense of parity with the Soviet Union. Little more than a decade later the country laid claim to technological superiority when it put astronauts on the Moon and brought them back safely.
By contrast, Sputnik 2.0 poses a much more difficult problem. It’s not enough to build the first working model of a particular technology, or even a suite of technologies. If there is one thing we know from the post-Sputnik experience, it is that spending more money on R&D will not solve all the problems science policy confronts. And it will not deliver the security, prosperity, or public health that the public expects from these investments. An overhaul of American science policy is long overdue and will require innovative experiments with funding and a built-in mechanism for learning from the experience. Those features of S. 1260, more than the eye-popping budget numbers, could be a good place to start that process.