What exactly does an innovation-driven industrial policy look like?
On Wednesday at noon Eastern, Future Tense will host an hourlong online discussion about the relationship between the U.S. government, national security, and private industry. For more information and to RSVP, visit the .
When in 1791 Treasury Secretary Alexander Hamilton presented his plan to Congress to develop a U.S. manufacturing sector, he kick-started a debate about the role of the federal government in industrial policy that has run consistently through the history of the republic. In recent years, though, despite the examples of industrial policy enthusiasts like Abraham Lincoln and Dwight D. Eisenhower, the prevailing political climate—driven by the Tea Party’s influence on congressional Republicans—has been to largely reject the idea that government should “interfere” with the free market at all. In practice, the federal government has continued a wide range of measures that support industry, especially in the defense sector. Nevertheless, such has been the political climate that even the Obama administration’s investments to promote recovery from the 2008 financial crisis (especially under the 2009 American Recovery and Reinvestment Act) became the fodder for bitter partisan controversy, with 2012 vice presidential candidate Paul Ryan declaring during the campaign that “the big problems we have today … [is] … [p]icking winners and losers in the economy through spending, through tax breaks, through regulations does not work.” As a result, not only has there been a lack of public discussion, beyond a small group of academic and think tankers, about how the U.S. should run its industrial strategy, but there has been even less effort to optimize the federal government’s ability to do that.
Now, however, the United States seems on course for a historic bipartisan embrace of an innovation-driven industrial policy and with it the opportunity for government to start facing up to the challenge of delivering such a policy. Recently, the Trump administration announced a $1 billion investment in a series of academic and private sector artificial intelligence and quantum information science hubs linked to federal agencies—one more step toward embracing a set of policies that in recent years has been politically toxic in the United States. Five of the hubs will be institutes focused on quantum information science, based around five Department of Energy national labs. Notably they will include staff from tech leaders like Microsoft and Intel. While additional research funding hardly represents a wholesale shift to a command economy, these investments represent a growing recognition that the federal government needs to help U.S. companies stay ahead in the race to build world-beating technologies. Meanwhile, the Biden campaign’s “Made in All of America” economic plan indicates that it favors even bigger investments.
What is lacking, however, is any clear sense of who should be directing this innovation-focused industrial policy, where they should sit within the federal government, and what will guide their strategic thinking. While ultimately the president is responsible for everything that goes on in the federal government, a successful industrial policy will require strong leadership below that level—in particular, the right support structure to deliver that leadership across a federal government that has relatively little experience of true industrial policy. Just as we should take a keen interest in the military commander appointed to command America’s wars, Congress, the media, and others should be interested in who directs this new area of government policy and how they are set up to do so.
It is worth reiterating: It is not—and has never been—that the U.S. does not have a de facto industrial policy. Through regulation, foreign investment rules, trade barriers, and even subsidies (think ethanol), the federal government has found ways to support U.S. industry. And even the most ideological appropriators have not succeeded in removing millions of dollars of research funding channeled through the long-standing research agencies like the National Science Foundation, the National Institutes of Health, and the Defense Advanced Research Projects Agency or through programs established to support development of that research, such as the Small Business Research Innovation program (now branded as “America’s Seed Fund”).
However, we do seem to be in the midst of a broader political shift with the potential to fundamentally alter the role of the federal government. Years of free market orthodoxy and Republican-driven budget politics in the U.S. have inhibited significant thinking on how best to leverage the U.S. federal government to support industry, even among many Democrats. Meanwhile, the success of companies like Google, Facebook, and Apple over the same period, apparently the result of private investment, has helped reinforce the idea that politicians should just let the market do its thing. As a result, according to America Association for the Advancement of Science figures, while overall federal government investment in research and development increased significantly over the past decades, as a proportion of GDP, it was allowed to drifted down from 1.23 percent of GDP in 1976 to 0.94 percent in 2018. During that period rising private investment compensated, helping to keep the United States as the world’s leader in R&D investment (28 percent of the global total in 2016 according to the OECD), but relatively little of that investment has gone to supporting strategically important technologies. Instead, as you might expect, much of that money has gone to less risky, more short-term enterprises—such as commercially focused software driven companies—at the expense of more difficult research problems and the development of hardware products, which take longer to pay off. Now, however, there appears to be political convergence, if not quite consensus, on the case for greater government investment in innovation.
The reason for the shift is, of course, China, which is investing heavily in foundational technologies like A.I. and quantum information services that are set to define the 21st century. The publication of China’s “Next Generation AI Development Plan” in 2017, which said that China would be a world leader in A.I. by 2030, signaled support at the highest levels of government for building new technology, and as commentators like Kai Fu Lee have pointed out, members of China’s new generation are just as hard-working as their Silicon Valley competitors, if not more so. Meanwhile, China’s deliberate focus on making all that technology available to the military has the Pentagon understandably concerned. For example, recent advances by Chinese researchers in the field of quantum communication threaten to deliver technology that would make communications uncrackable—a revolutionary development with obvious military as well as civil implications.
As a result, some of the more national security–minded Republicans have started to question whether, left to its own devices, the private sector will fund the research and then undertake the potentially long and uncertain development of the sort of technologies that will prove necessary to match China. Most notably, Sen. Marco Rubio, a Republican from Florida, argued in a speech delivered in December at the National Defense University for a “21st-century pro-American industrial policy,” explicitly highlighting what he called the “Perils of Free-Market Fundamentalism,” arguing that “sometimes the most efficient outcome is at odds with the common good and the national interest.” Meanwhile Republicans Sen. Todd Young and Rep. Mike Gallagher both signed on to support the Endless Frontier Act, a bipartisan bill to reform the NSF by adding a Directorate of Technology and funding it to the tune of $100 billion over five years.
Quietly but perhaps even more significantly, the Trump administration seems to have taken this point too. As author and blogger on big business Matt Stoller has written, despite consistently submitting budget proposals that massively cut R&D budgets across the federal government, the administration has repeatedly agreed to funding increases. For example, as Stoller shows using figures collected by the America Institute of Physics, while the administration’s proposed fiscal year 2020 budget request would have reduced the Department of Energy’s funding for renewable energy by 86 percent, the actual FY20 appropriation saw an increase of 20 percent, a pattern repeated across other science-related funding lines. Meanwhile, additional funding for areas like artificial intelligence and quantum science appear to show that, put simply, President Donald Trump’s administration is more scared of China than of the Tea Party.
This shift in Republican sentiment has left the Democrats emboldened to make an even more bold and explicit case for an innovation-driven industrial policy. Joe Biden’s former (and maybe future) national security adviser Jake Sullivan recently wrote with Jennifer Harris in Foreign Policy, “[A]dvocating industrial policy … was once considered embarrassing—now it should be considered something close to obvious.” With that in mind, in August, the Biden campaign released an economic plan promising a $300 billion investment in R&D, as well as a further $400 billion for federal procurement, which would be used to boost funding in key technology areas.
In truth, a closer partnership between the U.S. government and U.S. industry has a strong historical precedent. There has been a strain of Jeffersonian resistance to government support of industry running through the history of the country. But that partnership has always come to the fore, especially when the country has come under threat. Each time, however, such policies have raised additional public policy questions. This is no exception. If such large amounts of government money are to be applied to supporting industry, how should that best be done? The problem is that with the speed with which the taboo on discussing a U.S. innovation-driven industrial policy has disappeared, there has been little time for a public debate about what such a policy should look like and still less consideration of how it might be implemented across the federal government. We now need to be ready to have those discussions.
It hardly needs be said that it is difficult to predict what a second-term Trump administration would do in this or any other area, but it seems likely that it would continue the recent practice of devolving responsibility to federal agencies, especially the Department of Defense and Department of Energy, with what is left of the Office of Science and Technology Policy in the White House driving specific high-profile initiatives like the A.I. and quantum information sciences institutes. That can work, assuming that the appropriate senior government leaders are engaged, but it does carry the risk that relevant decisions affecting other departments (for example, the various executive orders targeted at Chinese tech firms like Huawei, TikTok and WeChat) are pursued in isolation from a wider innovation strategy.
The Biden campaign has been much more explicit about what it seeks to do, but there is still an absence of any indication of how it plans to implement its agenda. To be fair, aspirational org charts do not win elections. But with a commitment to spend hundreds of billions of additional taxpayer dollars in just a few years, the details of the implementation plan will matter. How should funding be apportioned across the various sectors mentioned in the Biden plan (such as advanced material, health and medicine, biotechnology, clean energy, autos, aerospace, A.I., telecommunications, and more)? Or should funding be allocated around DARPA-style “missions” as advocated by Italian economist Mariana Mazzucato, whose ideas on industrial policy have recently been influential in Europe? Such missions would likely include “meeting the China challenge” or “addressing climate change,” but what about pandemic response, or even other social challenges, such as dealing with an aging population or addressing mobility (as recommended in a recent Mazzucato-led commission on an industrial strategy for the United Kingdom)? What is the optimal balance between investment in research and measures to spur development of that into usable technology? And what is the right suite of policy tools to deliver the investment? Equally important, what is the right balance between investing in creating jobs and creating skilled workers to fill them? More broadly, the Biden plan at least suggests R&D investment should be distributed more evenly around the country, but how should such a policy be implemented in practice? And who is monitoring what is working and what is not, and who takes action to change course in light of new information?
No doubt the clever people who wrote the Biden platform are busy writing their transition memos right now, with a view to addressing at least some of these points. Whatever that plan looks like, however, it should be built on a few key principles, which of course would be relevant to a second-term Trump administration, too.
Whoever wins the election, his administration should take a cross-government approach to implementing these policies. Effectively delegating responsibility for innovation strategy to a few big departments risks, at best, a suboptimal misallocation of resources and, at worst, creating policy dynamics (such as conflict with China) that might be avoided by a more holistic strategy. So campaign policy committee plans notwithstanding, the best way to do that might well be through a cross-government public/private review process. There will be real pressure for quick action, but as Eisenhower said, “Plans are useless, but planning is everything.” The benefits of building national buy-in through a deliberate strategy process should not be underestimated.
Senior leadership of this effort is obviously critical, but perhaps even more important will be to get the right team to oversee policy implementation. As a practical matter that will need to leverage a diversity of backgrounds: There are very few people who can rightly be called “industrial policy practitioners” with the right blend of government, business, economic and political know-how. A Biden administration might feel pressure to keep business people out of the mix, but without that perspective, it will be much harder to develop and implement the right balance of new policies that help key industry sectors develop research into usable technology. At the same time, the team must be built from a good mix of government insiders to be bureaucratically effective. Similarly, the team culture of the implementers will be crucial: They need to be adaptable and willing to learn.
Beyond that however, and perhaps most importantly, the federal government should not just be left to build this on its own: An industrial policy will need to be a whole of nation effort. If indeed we are moving to a period where the United States embraces industrial policy, as I believe it should, it will require a national effort—from Congress, industry, academics, the media, and ultimately the electorate—to make sure that it is done right. Quite apart from the hundreds of billions of dollars potentially at stake, it may not be too much to say that the future of the country literally depends on it.
Future Tense
is a partnership of
Slate,
New America, and
Arizona State University
that examines emerging technologies, public policy, and society.