What a National Technology Strategy Is—and Why the United States Needs One
To compete in the twenty-first century, the United States needs a nimble agency to catalyze technological innovation that delivers security, prosperity, jobs, and health—for all citizens.
Over the last half century, the global geopolitical balance of scientific, economic, and production capabilities has shifted away from US dominance. The United States is no longer in a singular position of global scientific and technological leadership, and China has become the largest producer and second largest market in the world. Meanwhile, we face equal or greater challenges than ever before on the home front, where economic inequality has increased, social mobility has declined, and political polarization is on the rise.
Unfortunately, at this moment of dual internal and external challenges, the United States’ intellectual and institutional foundations are insufficient to develop ways that can resolve them. Leading experts have made different and often-conflicting proposals. Some advocate for slowing the progress and adoption of technology. Others argue for dramatically increasing funding of science and technology, including investing in regional innovation hubs to reduce inequality and increase jobs. On their own, none of these proposals are likely to fulfill experts’ or legislators’ multiple objectives for them—improving national security, increasing the number and distribution of good jobs, and succeeding in global trade.
Missing from these debates is the recognition that win-win technology choices do exist. That is, with the right incentives, it is possible to make strategic investments in technology that achieve multiple national objectives. For example, Christophe Combemale and his coauthors have shown that not all technology leads to wage and skill polarization; indeed, many of the technologies on today’s critical technology lists may lead to better jobs for high school graduates and strengthen national security at the same time.
Likewise, in contrast to regional hub proposals that will require decades to supply the promised jobs, I have argued that equitably building the infrastructure of the future—smart high-speed transit systems, dynamic electric grids with renewables, and broadband internet access—will more quickly increase jobs in underserved areas, improve social welfare for all citizens (including health, energy access, and communications), and boost the productivity and resilience of industry. In addition, if this infrastructure is domestically procured, it could rebuild US manufacturing.
Unlike a firm, which has the single objective of profit maximization, a nation has multiple objectives, including national security, economic prosperity, and social welfare. Making transparent to policymakers where strategic win-win investments exist across these objectives will require building the intellectual foundations, data, and analytic tools necessary to inform such multi-objective decisionmaking. Acting across missions will require new government institutions capable of making such technical investments and delivering desired outcomes.
Although there has long been interest in the relationship between security and social objectives, and scholars have explored synergies and mapped trade-offs among environmental, employment, and other objectives, I am aware of no research to date that seeks to quantify trade-offs and win-wins across the full range of national objectives. US agencies and departments, including those in science and technology, typically have singular missions, such as defense, energy, transportation, commerce, and labor. These government bodies are excellent and should not be changed. At the same time, the current system leaves a hole whereby even with each agency or department perfectly fulfilling its distinct mission (say, defense, trade, or environmental protection), the country could still fail to fulfill its multi-objective role (say, for labor).
To foster win-wins across national objectives, a US National Technology Strategy Agency is needed to seed initiatives that fill gaps in the existing innovation ecosystem and to catalyze other agencies to bring their expertise to cross-cutting efforts. This new agency will need to simultaneously build the interdisciplinary intellectual foundations, data, and analytic capabilities to make win-wins transparent and inform its investments.
What National Technology Strategy Is Not
Building a US national technology strategy should not involve changing the basic structure of the departments and agencies we already have, nor should it involve imposing top-down coordination or locking the country into single technologies or policy objectives. Calls for a national technology strategy that involve top-down coordination, efforts to reduce “redundancy” across agencies, or attempts to “reduce inefficiencies” are misguided and could actually dampen innovation.
In fact, one of the strengths of the US innovation system is its diversity and redundancy. Scholars have long emphasized the importance of the diversity of the US innovation ecosystem, in which agencies and departments have different missions and can take aligned, complementary, or even opposing funding roles. In this system, scientific and technical progress is a long-term, nonlinear process in which metrics and a focus on efficiency can slow and fragment progress instead of enhancing it. The National Research Council beautifully describes how the US innovation ecosystem’s mix of mission-based agencies helped create a revolution in computing: “By funding a mix of work in universities and industry, [the United States] was able to marry long-term objectives to real-world problems. And, by channeling its funding through a variety of federal agencies, it was able to ensure broad-based coverage of many technological approaches and to address a range of technical problems.”
Another national technology strategy solution that is commonly proposed is creating reports with lists of critical technologies. While these reports are a useful step, they cannot be the central foundation of a robust US technology strategy. History shows that such lists on their own are unlikely to find their way into policy or action. Between 1989 and 1999, for example, the federal government identified critical technologies through a biennial National Critical Technologies Report to Congress, with input from multiple agencies, including the Department of Defense, Department of Commerce, Department of Energy, and National Aeronautics and Space Administration. Unfortunately, the reports lacked the follow-on necessary to link criteria to policies—never mind to coordinated policy actions—in a productive way. Indeed, one of the many assets of the US innovation system is its diversity, nimbleness, and flexibility to respond to changing times. Reports don’t have this flexibility.
In addition, if a national technology strategy were about a single mission such as security, key win-win opportunities may be lost. Advanced semiconductors—which stand at the center of current US challenges in security, trade, and jobs—offer an example of the potential dangers of optimizing for only a single objective, rather than incentivizing technological win-wins across multiple objectives. For example, a policy aimed at maximizing national security and minimizing defense costs might take a three-pronged approach of funding innovations in hardware and software security, supporting chip fabrication in a series of allied nations, and funding advances in the next generation of computing (e.g., beyond Moore’s Law). By contrast, a policy giving equal weight to national security and labor might increase incentives for foreign and domestic firms to invest in fabrication facilities in the United States. A policy that added equity might also increase incentives to locate those fabrication facilities in underserved communities, while investing in university electrical engineering programs in semiconductor hardware design and vocational program training in semiconductor manufacturing in those places.
Similarly, vehicle electrification policies demonstrate the potential dangers of optimizing for only a single objective. If policymakers focus solely on reducing carbon emissions, the most advantageous approach may be to scale electric vehicle use as quickly as possible. However, if they expand the objectives of the investment to include maximizing national security, prosperity, and equity, policymakers would need to find ways to quantify the value of domestic manufacturing of batteries (for jobs, security, and innovation); identify which citizens in which places will gain and lose jobs through the transition; assess the value of various levels of cybersecurity requirements for security, welfare, and learning; and determine how shifting the source of pollution from vehicles to energy generation sites on the grid (which disproportionately have poorer populations living near them) may decrease equity.
To overcome these obstacles, in parallel to mission-oriented efforts, the United States requires a nimble institution that can work within the existing mission-oriented innovation ecosystem and identify and act upon the opportunities afforded by win-win investments. Unfortunately, for both of the above examples, right now the government lacks the data and analytic capabilities to quantify and make transparent the implications a particular technology solution has for each national objective, the trade-offs different technology solutions present across multiple national objectives, and the potential self-reinforcing benefits of certain choices for subsequent decisions (such as making it more cost-effective to locate subsequent manufacturing in the same location in the future).
Toward a National Technology Strategy Greater Than the Sum of Its Parts
Correctly implemented, a national technology strategy must be about incentivizing innovation that offers outsized returns across national objectives, without undermining the strengths of our existing innovation ecosystem.
To catalyze such technology solutions, the United States should create a small, nimble agency that can research opportunities, fund strategic initiatives independently, and work across, coordinate with, and catalyze initiatives by the existing mission-driven departments and agencies. This National Technology Strategy Agency should be charged with making strategic technology investments across missions, as well as identifying and filling the holes in our existing national innovation system that are preventing the nation from realizing all of its national objectives. This agency must have an analytic arm and an executive arm housed within the same agency.
The agency will need sufficient money for its investments to be influential and to fund platforms of technology, but its budget should be sufficiently modest so that it is forced to engage and influence efforts in other agencies to have a larger impact. Based on lessons from the Defense Advanced Research Projects Agency (DARPA), Advanced Research Projects Agency-Energy (ARPA-E), and Office of Technology Assessment (OTA), I recommend an annual budget of $3 billion for external seed funding, plus an operating budget of $500 million to employ 100 program managers and 100 analysts with an appropriate support staff and facilities. This level of funding would give the agency a budget and program manager staff roughly on par with DARPA (which today has a $3.5 billion budget) and an analytic team slightly smaller than that of OTA, which at its closing in 1995 had 143 full-time staff (augmented by contractors) and an annual budget equivalent to $52 million in today’s dollars.
For the executive arm, the Semiconductor Research Corporation (SRC) provides an excellent model of how one entity with seed funding and political capital can amplify its impact by bringing multiple funding agencies together at the state and federal levels around a common mission. Unlike SRC, however, a National Technology Strategy Agency must act to forge a technology path across the missions of the existing agencies to meet the full multi-objective role of government. Public officials with embedded autonomy—deep knowledge of the technological, social, and industrial context—are most likely to get these choices right. As in DARPA, the executive arm should have a staff of rotating program managers brought in from academia, industry, and government who are the best and brightest in their fields, able to use the position as a stepping-stone to subsequent leadership positions in their careers. Unlike in DARPA, at this agency, program managers might include star diplomats or government officials, union and nonprofit leaders, teachers, and community activists alongside top-notch technologists.
At the same time, the analyst arm will need to provide transparency for policymakers and the new agency’s program managers on the trade-offs present in different potential technical decisions for meeting national objectives. Given the current bedraggled state of the government’s analytic capabilities, the analyst arm will need to develop new data and methods to perform systematic assessments of national and global technology and production capabilities. The analyst arm should have a stellar interdisciplinary staff of PhD-level experts in each technical field (65–75% of its experts), as well as PhD-level economists, political scientists, sociologists, psychologists, and historians focused on applying their expertise to real-world technology policy problems. Similar to that in OTA, the full-time staff of the analyst arm of this new agency should leverage contracts with academic researchers to develop new data, methods, and analytic insights. These contracts should be short enough to be relevant to political timelines, but long enough to engage scholars in academia: the sweet spot is likely one year.
The full-time staff should then integrate the resulting academic insights and translate breakthroughs in data and analytics into regular government functions.
Funding academic research also plays the important role of not only bringing in stars to address the nation’s challenges, but creating incentives for researchers in academia to work on real-world technology policy problems, which require integrating technical and social science expertise.
To ensure excellence and relevance, the agency must have an external expert advisory board with leaders from academia, industry, government, and nonprofits (such as labor unions or community activists). In addition, based on lessons from DARPA’s Information Science and Technology Study Group, the National Technology Strategy Agency should have small, rotating, problem-specific expert advisory boards drawn from industry, academic, and community leaders as well as program managers. (Notably, OTA also had study-specific advisory boards.) These small study-specific advisory boards ensure that the analyst staff and academics are grounded in the science, engineering, industrial, and political realities of the problem on the ground. Subsequently, participating program managers make sure that the study’s suggestions are acted upon.
The proposed National Technology Strategy Agency takes from the best of recent US technology initiatives to catalyze a revolution in how the nation approaches funding science and technology. By incentivizing technology paths with win-wins across missions and orchestrating initiatives across different mission-oriented players, it could amplify investments across agencies and departments to deliver on not just one but multiple objectives.
Finally, and perhaps most important for its longevity, the National Technology Strategy Agency has the potential to be politically popular, particularly if it is successful in raising the employment, equity, and welfare of all citizens. Built as described above, such an agency would also be capable of teaching itself and the nation how to push forward with continuous improvement to define the future, rather than merely respond to the past.
Lessons from our National Innovation System
Catalyze coordination from the bottom up. A National Technology Strategy Agency should build upon lessons from past models that have been successful in catalyzing multiple entities to collaborate and co-seed technical initiatives. Calls for top-down coordination can misunderstand the complexity of the national innovation system and the ways that bottom-up coordination already happens within that system. In the semiconductor industry, SEMATECH, SRC, and the National Nanotechnology Initiative (NNI) offer examples of bottom-up coordination from very different stages of scientific and technology development.
SEMATECH was originally a 50-50 government-industry public-private partnership to promote near-term equipment upgrades to increase competitiveness with Japan. SRC is an industry-led public-private partnership that funds academic research three to seven years out to ensure research advances meet industry needs. NNI works to support and set priorities for more fundamental long-term research in nanoscale science and technology.
At SRC, industry leaders meet regularly with program managers from the National Institute of Standards and Technology (NIST), the National Science Foundation (NSF), DARPA, and DOE as well as state leaders to decide on funding directions and co-fund complementary agendas under a single SRC program umbrella. Likewise, NNI has facilitated working groups, an infrastructure network involving an integrated partnership of user facilities at 13 campuses across the United States, and centers to support the development of tools for fabrication and analysis at the nanoscale. It has also created NNI-industry consultative boards to facilitate networking among industry, government, and academic researchers, analyze policy impacts at the state level, and support programmatic and budget redirection within agencies.
Fund solutions, not industries. A National Technology Strategy Agency must undertake policy tailored to technological and sectoral nuances, while explicitly avoiding policies that support industries. Policies focused on sustaining established firms or specific industries rather than catalyzing solutions to problems will fail to achieve important national objectives. For example, a challenge like the end of Moore’s Law in advanced semiconductors will require enormous quantities of funding to solve, yet has implications for economic prosperity, national security, and social welfare. It would be easy to misallocate funding in an attempt to address this problem—indeed to misunderstand the nature of the challenge itself.
The system of developing silicon-CMOS chips (the kind of integrated circuit that underpins computing), which has flourished for 40 years, is coming to the end of its physical limits. It would be foolish to simply fund established firms to continue this soon-to-be-defunct trajectory. Instead, we should fund the advances in new material systems (beyond silicon-CMOS) to ensure computational capabilities continue to advance and that the United States leads in those advancements. Here, I am not proposing choosing technology winners; no one knows which innovation in beyond CMOS devices will be the solution. Rather, I am emphasizing the importance of spending our limited national dollars on the right problem.
Think beyond moonshots. A National Technology Strategy Agency must avoid the lure of using “moonshots” as a one-size-fits-all solution. Although they’ve become increasingly popular, moonshots, competitions, or contests are unlikely to work well in all contexts, particularly where significant platform coordination is necessary. Consider, for example, the challenge of inventing the next generation of underlying transistor technologies. This challenge is an extremely difficult problem requiring advances in the underlying physics with implications for security, prosperity, and society. But trying to solve the problem through a moonshot or prize would be problematic. First, it requires coordination across the computing technology stack, including new chip architecture, new software, and new equipment. Therefore, it would be difficult for a single innovator or new entrant to manage this coordination, especially with such high uncertainty early on about which new technological solution would win. Second, the required capital investments are considerable‑—more than a billion dollars is likely needed even for a device-prototyping foundry. These considerations speak to a need for coordination, rather than individualized competition. Here, a government arm similar to DARPA, in coordination with other agencies and private industry, would be best suited to lead a technology revolution. Such an agency would be able to achieve the necessary coordination and overcome issues preventing private firms (new entrants and established corporations) from making the leap on their own, including fragmentation of technology trajectories, declining profit margins among established firms, and profitability of short-term solutions for other private stakeholders.
Orchestrate outcomes without choosing winners. A National Technology Strategy Agency should take lessons from DARPA on how to successfully orchestrate technology revolutions. My research on DARPA demonstrates that, rather than forcing policymakers to choose between the extremes of free markets or the heavy hand of government to select successful technologies, DARPA offers a third alternative: embedded network governance. Program managers work to identify and influence new technology directions through constant contact with the research community. By understanding emerging themes and matching them to military needs, those managers then bring together discrete researchers, bet on the right people, stand up competing technologies against each other, and fund platforms (or “pyramids”) of technologies to address interdependencies across components in the system. As they do their job, program managers maintain the bird’s-eye perspective critical to integrating and orchestrating disparate research activities spread throughout the national innovation ecosystem to achieve military, scientific, and technological goals. This goal-oriented, program manager-level orchestration of technologists has led to technological breakthroughs as wide-ranging as the internet, mRNA vaccines, and artificial intelligence.
Leverage the whole ecosystem. A National Technology Strategy Agency needs to leverage the entire innovation ecosystem, understanding the variety of models within it and the role each plays in advancing science and technology. While DARPA may play an important role, it plays only one role in this complex system.
Although DARPA is a model for funding and commercializing breakthrough technologies, it is not a model for funding breakthroughs in basic science. In medicine, the Howard Hughes Medical Institute has funded breakthrough discoveries in basic science by providing substantial funding to worthy young investigators with few restrictions. This open-ended approach contrasts starkly with DARPA’s mission-oriented milestones and orchestration. Likewise, in the computing revolution, the Office of Naval Research made critical early investments in advance of DARPA.
Neither mission-oriented nor focused on the funding of eminent scientists, NSF has repeatedly come through on essential aspects of technology development. In the case of the internet, while DARPA funded the early breakthroughs, continued efforts sponsored by NSF to develop CSNET and later NSFNET (a program of coordinated, evolving projects linking university-based supercomputer networks to be able to share information and resources with each other) demonstrated the value of internetworked communication systems and led to the internet’s eventual commercialization. With its funding spread among researchers at a wide range of institutions, generally universities, NSF also plays an important role in the broad scientific and technological education and dissemination of knowledge needed for the development and commercialization of revolutionary technologies.
Furthermore, the mission-oriented departments and agencies play a critical role in providing incentives to focus on and solve real-world problems. For example, given the centrality of computing to each of their missions, DOD, DOE, NASA, NIST, and NIH all played essential roles in the computing revolution.
Invest in crossover capabilities. Science and technology investments have systemic implications and complementarities that, if invested in thoughtfully, could enhance opportunities to succeed across multiple objectives. My and my colleagues’ research on responses to COVID-19 at the national and firm levels underscores the importance of national competencies in technology and production and the ways science and technology capabilities can reinforce each other across sectors.
On the one hand, national domestic capabilities can support a nation’s ability to effectively invest in and rapidly regulate the introduction of new technologies. As just one timely example, DARPA was among the agencies that provided early funding for mRNA vaccines. This investment, coupled with the tacit knowledge of US manufacturers that had previously developed complementary intellectual property and domestic manufacturing capability, enabled the development and manufacture of Moderna and Pfizer-BioNTech’s mRNA vaccines on a remarkably short timeline.
Likewise, where companies quickly and successfully pivoted into producing scarce COVID-relevant medical supplies, the centrality of US-owned businesses and the domestic manufacturing ecosystem was notable. US manufacturers already in the business of personal protective equipment (PPE) production (e.g., Honeywell, 3M, Prestige Ameritech) were able to leverage their intellectual property, sourcing networks, and domestic manufacturing capabilities in other sectors (such as aerospace and advanced materials) to rapidly shift to domestic manufacturing sites. US domestic manufacturers in other businesses, such as automobiles (e.g., General Motors, Ford, Tesla) and filtration materials (e.g., American Melt Blown and Filtration, Berry Global Group), were likewise able to leverage their substantial technical, sourcing, and production expertise to rapidly begin mask, ventilator, and melt blown material (used in facemasks) production.
On the other hand, a lack of national domestic capabilities can hurt a nation’s ability to respond. Ongoing research by Afonso Amaral and colleagues suggests that European countries without domestic manufacturers of ventilators struggled disproportionately in the regulation of new domestic entrants. Likewise, although some components of the domestic manufacturing ecosystem supported the US response, the dilapidation of the US manufacturing ecosystem also hindered some manufacturing companies attempting to pivot into producing badly needed medical supplies. Research by Nikhil Kalathil and colleagues suggests that, during the pandemic, small- and medium-sized companies could have been particularly responsive to the need for PPE and other protection at local dental clinics, nursing homes, and small hospitals. But these manufacturers faced problems including a lack of access to domestic sources of intermediate inputs, equipment that was largely built in China, and a lack of skilled workers to run, fix, and adapt equipment. They also faced information barriers and high financial costs of passing regulatory hurdles.
These examples show that an effective national technology strategy must attend to cross-cutting investment that spans sectors and layers of the national innovation ecosystem, while ensuring that there are no holes in critical domestic capabilities—whether in production or innovation.
Toward a National Technology Strategy
The COVID-19 pandemic has highlighted deep global interdependencies in health and manufacturing as well as national challenges in racial, geographic, and income inequality and job safety. As the United States began attempting to respond to the pandemic, the nation found that we had undervalued various aspects of social welfare, including health and equity. We had also undervalued resilience and domestic manufacturing. We lacked the data to know who all our own manufacturers were and which of those could possibly help respond to the pandemic, never mind where bottlenecks might exist elsewhere across the globe. If our institutions continue to address only singular missions, such as national security without health or equity, these problems will be repeated, whether during the next natural disaster, the next pandemic, or the next war.
To build a stronger and more secure, prosperous, equitable, and resilient nation, we must create an institution that can design investments to realize all of legislators’ objectives for them. With a National Technology Strategy Agency—whose mission is to identify technological solutions with win-wins across all of the nation’s objectives and to leverage the complementarities across sectors and the systemic nature of investments in technology—we can create this future.
Jobs and equity are as central to our sovereignty and security as weapons. The last 75 years has demonstrated that, without a portion of government assigned to designing technology for win-wins across all national objectives, the impact of technology choices can be uneven, reducing jobs and equity for some while increasing productivity and wealth for others. But, done right, technology investments can address—and indeed have outsized returns in addressing—national security, economic prosperity, jobs, health, the environment, and equity. The United States cannot afford to get these decisions wrong. While the excellent existing departments and agencies will continue to fulfill their specific missions, we must act now to found the new institutions that will identify, catalyze, and orchestrate technological paths across our innovation ecosystem to ensure that our technology investments are designed to create security, prosperity, and welfare for all citizens.