Can Europe Help the United States Avoid a $115 Billion Boondoggle?

By Michael J. Bernstein

Over the next five years, the United States may invest $115 billion in scientific and technological innovation in addition to the amount that the federal government already spends. The extra money is specified in the Endless Frontiers Act (EFA), which is included in the massive United States Innovation and Competition Act of 2021 (S.1260). This legislation will establish a new Directorate for Technology and Innovation at the National Science Foundation (NSF), and provide funding for programs and activities that will, according to the bill’s summary, strengthen “US leadership in critical technologies through basic research in key technology focus areas.”

For this investment to have the best possible impact on US innovation—and society at large—the EFA needs to infuse public values, including sustainability, gender and racial equity, and social benefits, into the research and development enterprise. Before investing billions of dollars, American policymakers should learn from the European Commission’s ongoing approach to massive, coordinated research and development investments. Between 2013 and 2020, the European Union committed nearly €80 billion (approximately $94 billion) to advance economic, environmental, human health, food, and energy goals through its Horizon 2020 funding program. Although Horizon 2020 had a broader agenda than the EFA, its successes and failures offer instructive lessons in how—and how not—to wed public values with colossal research and innovation ventures.

In particular, the European Union’s experience may illuminate how administrative structures help or hinder aspirational public agendas. Beginning in 2013, Europe’s responsible research and innovation (RRI) requirement asked all Horizon 2020 programs to adopt ethical considerations, gender equality, public and stakeholder engagement, and other concerns into proposal review criteria and other materials. The European Union then invested €1.8 billion over seven years toward some 1,100 projects claiming a degree of compliance with these goals. While that sounds like a lot, it in fact amounted to only 3.1% of the 35,061 projects that received funding. So far, despite Horizon 2020’s ambitious language, the European Union’s attempt to create a broad culture of responsible research and innovation has fallen short.

The Endless Frontier Act needs to infuse public values, including sustainability, gender and racial equity, and social benefits, into the research and development enterprise.

In the coming months, US policymakers should pay careful attention to the fine print of the EFA’s aspirations for sustainability and public value and gather lessons from comparable efforts abroad, especially Horizon 2020. As Europe’s experience with RRI in Horizon 2020 attests, the minutiae of framework organization, administrative structures, and implementation will have a significant impact on whether the EFA’s $115 billion hits its intended marks with respect to public values.

The United States now has an opportunity to leapfrog European experience by more deliberately developing selection criteria, capacity building efforts, and program coordination—and explicitly moving beyond zero-sum thinking—to advance innovation, technology, and economic gains without sacrificing US workers, the environment, or other aspects of equal opportunity that are central to the American project.

Criteria are key

For any R&D initiative to successfully deliver societal benefits, it needs appropriate eligibility and grant evaluation criteria. So far, there is little to suggest that the criteria set forth in the EFA—whether for project awards by the NSF (Sec. 2103, 2107), university technology centers (Sec. 2104), or regional technology hubs (Sec. 2401)—will significantly improve on the status quo. NSF’s broader impacts criterion, which requires that grantees describe the societal benefits of their research, remains the primary vehicle for public values to be incorporated into the funded science, but it faces criticism. For university technology centers, the EFA makes only general statements exhorting regional and geographic diversity and consideration of underrepresented populations in STEM disciplines. The bill dictates, vaguely, that grantees “take measures to prevent the inappropriate use of the research and technology of the center, including research results” (Sec. 2104.a.4.F).

To genuinely advance broad US public values through R&D, the EFA’s grant evaluation criteria should explain how awardees will consider, design, implement, and monitor technology developments for responsible research and public values. Such an effort should be specific, process-based, outcome-oriented, unequivocal, and integrated with other eligibility and selection criteria for awards.

So far, there is little to suggest that the criteria set forth in the EFA will significantly improve on the status quo.

Here the European experience with Horizon 2020 is particularly valuable. At the outset, the European Commission granted tremendous latitude to each of Horizon 2020’s 19 subprograms to tailor evaluation criteria. In most cases, especially in the first four years of Horizon 2020, this meant most calls for proposals defaulted to generic criteria about research excellence without specifically requesting consideration of how projects might contribute to broader social missions. Only when statements about responsible research and innovation appeared in multiple selection criteria did applicants have any reason to seriously consider how to integrate these social missions into their work.

Baking in criteria around responsibility and public interest will require researchers, innovators, stakeholders, and organizations to engage more genuinely with whether their plans can really deliver on the ambitions of what the EFA could stand for. The alternative—simply issuing ambitious but meaningless statements of intent—risks repeating already visible patterns, including companies spreading propaganda about sustainability, being “innovative” by generating temporary work forces or armies of independent contractors, and generally externalizing social and environmental fallout of their operations.

Being specific about funding criteria can drive deeper behavior changes at research institutions when criteria are backstopped by midterm and final evaluations with course-correction consequences. When trained, networked, and socialized around specific criteria and active approaches to program management, people educated by and working at university technology centers or regional hubs—whether they teach, research, start a business, advocate, or (in the language of the bill) “diffuse innovation around the United States”—stand a better chance of creating a culture of R&D aligned with broad public interest.

Patiently build the capacity of researchers and program managers

Changes in the culture of R&D, however, don’t happen overnight. To perform R&D in the public interest, program managers must be capable of selecting eligible consortia of organizations that could apply to become regional technology hubs and researchers capable of submitting such grants and doing the work. This does not come without considerable effort and support. Generally, program managers fund what they know, and what they currently know is shaped by what they learned in siloed academic disciplines. Inconveniently, for the past 80 years, these environments have justified themselves through lobbying for autonomous, curiosity-driven research. We should therefore not be surprised when the ability of researchers to suddenly pursue R&D not only for economic competitiveness but also public values such as diversity or environmental protection does not spontaneously materialize.

European experience with advancing gender equality in R&D is instructive here. Today, in the successor investment program to Horizon 2020, the European Commission requires that organizations demonstrate plans, investments, and commitments to gender equality in order to be eligible for funding. These requirements for considering gender balance across research teams and leadership did not arrive overnight. Rather, these criteria reflect more than two decades of consistent investment—and they still have a long way to go.

Generally, program managers fund what they know, and what they currently know is shaped by what they learned in siloed academic disciplines.

It’s important to note that the United States already has the capacity to support research for public values. The EFA should focus on cultivating and building on those efforts, rather than starting from scratch. Since the early 2000s, a range of NSF programs have developed ways to support responsible research and innovation that advance broad public as well as economic values. Social science and humanities, formal and informal science education, technology incubator and accelerator programming, innovation corps, and other methods for making research more useable and suited to public purposes could be better leveraged through platforms modeled on the European approach to partnerships.

Indeed, the EFA text explicitly mentions Innovation Corps as an existing NSF program to include in the proposed NSF Directorate for Innovation and Technology. A culture of science and innovation in the public interest could be integrated into many kinds of research funding. Take the EFA’s proposed regional technology hubs, for example, which would bring investment and build capacity in regional workforces and infrastructure: cross-regional collections of principal investigators, public- and private-sector partners, and graduate and undergraduate students could be encouraged to integrate societal concerns into research and innovation processes. Information gathered from such cross-regional efforts could then be used to further refine hub governance models and expectations from the managing directorate. But it will take time—certainly more than the five-year funding periods specified in the legislation—to cultivate R&D cultures and practices befitting the ambitious policy vision outlined in the EFA.

Coordinate, don’t abdicate, responsibility

One thing the EU experience demonstrates is that disciplinary silos don’t work when the goal is to implement cross-disciplinary concerns related to public or economic values. Horizon 2020 invested well over €400 million toward advancing responsible research and innovation concepts through its Science with and for Society (SwafS) program. But little was provided in Horizon 2020 to integrate or diffuse the conceptual and methodological advances gained from SwafS across other programs, which themselves spent some €1.4 billion on projects claiming to advance RRI but lacked guidance. It’s no wonder, then, that a culture of responsible research and innovation as envisioned in Horizon 2020 legislation didn’t materialize.

It will take time—certainly more than the five-year funding periods specified in the legislation—to cultivate R&D cultures and practices befitting the ambitious policy vision outlined in the EFA.

The EFA seems poised to repeat these mistakes. The bulk of the proposed new funding ($81 billion) will flow through NSF, which is defined by siloed directorates. And the proposed NSF Directorate for Innovation and Technology (with $29 billion devoted to it) lacks social, behavioral, economic, or educational components that could help it build RRI concepts into its culture and practice (Sec. 2105).

Coordination of federal agency activities in the key technology focus areas features prominently in the EFA but leaves out key considerations (Sec. 2004.c). The legislation charges the new directorate to partner internally to study questions that could affect the “design (including human interfaces), safety, security, operation, deployment, or the social and ethical consequences of technologies in the key technology focus areas” (Sec. 2102.c.7.B.), but not in an anticipatory way or in tandem with innovation activities themselves.

This limited scope misses out on a major opportunity: to have NSF and its proposed new directorate benefit from a diverse array of practices already employed by other agencies to deliver research and innovation programs for specific social purposes. Adopting institutional innovations from top to bottom (e.g., from agenda setting, project selection, and evaluation to staffing, knowledge exchange, and integration of user expertise and citizen experience) could better empower the creation of interdisciplinary, convergent, and useable R&D through the EFA. The know-how exists. The EFA should champion the discovery of such means in a considered, inclusive, and systematic manner across the federal government. Equipping NSF and all agencies mobilized in the EFA to hone their program management capabilities so they can deliver on social aspirations seems prudent before authorizing too much of that $115 billion injection.

Move beyond zero-sum thinking

Congress should be careful to craft the EFA innovation strategy to advance truly national interests and not foster unnecessary competition between domestic regions. In the original draft of the EFA, Congress explicitly recognized that 90% of national innovation sector growth has occurred in only five major cities over the past 15 years. The current draft seeks to “avoid undue geographic concentration of research and development and education funding across the United States” (Sec. 2102); some regional technology hubs are expected to be located in rural communities and low-population states with the aim of creating jobs outside the five major cities. Still, the bill is missing explicit measures for encouraging hub-to-hub cooperation to support the national research and innovation mission.

In this respect, Horizon 2020’s European Research Area programs offer a more visionary model for supporting R&D and societal integration. Even more than economic advancement, the ambition of the European Research Area is to help support the integrity of the European Union by improving cultural exchange and the flow of people across borders. Although hampered by inequities in national budgets, the program seeks to overcome this by covering transnational convening and coordination costs, as well as supplementing funding so that national-level funders can implement cooperative programs. By offering matching investment funds, the program sponsors cost-effective ways for people from different countries to collaborate on research agendas and projects of shared purpose. Going further, the program regularly documents and shares key lessons and practical experiences associated with these partnerships.

The bill is missing explicit measures for encouraging hub-to-hub cooperation to support the national research and innovation mission.

Networks of regional technology hubs in the United States could similarly amplify their impact. Congress could modify the EFA to move regional hubs toward an approach modeled on Europe’s, in which some funds are distributed with the purpose of promoting collaboration (in tandem with capacity building) to support societal cohesion along with advancing economic growth in different parts of the United States.

Leapfrogging and the long view

In administering its R&D investments over the past 35 years, Europe has provided a wealth of lessons—both ideas to emulate and hazards to avoid. The United States should absorb these lessons in designing the EFA while avoiding obvious pitfalls. It’s important to remember that early European flagship initiatives were plagued by management challenges. Hard-won lessons with the Human Brain Project, for example, led to changes in the governance model of the Future and Emerging Technologies (FET) Flagships program that funded the project. By trading centralization for a more open, coordinated strategy, FET Flagships evolved to function as mini-directorates supporting and coordinating interdisciplinary, cross-sector, field-specific work.

Another lesson from FET Flagships is that the five years specified for the EFA’s awards is a very short term. Each of the Flagship programs was initially budgeted with €1 billion over 10 years. The original text of the EFA and its subsequent revisions build in a successive phase-out of support over a seemingly arbitrary and brief interval of time (although small and rural communities and Tribal governments are spared this diminution). More than anything, such a short timeframe will put immense pressure on researchers and administrators alike at the expense of quality.

Instead, the United States should consider a phased, adaptive approach to building, learning from, and continuously improving on any regional technology hub model. Such approaches have seen success in a range of scientific and technological domains with broad societal ambitions. This adaptive approach can be seen, for example, in how the 2012 Blue Ribbon Commission on America’s Nuclear Future sought to develop a new strategic approach to the nation’s nuclear waste management impasse.

Any sustained effort to catalyze US research and development for broad public benefit will need to last well beyond five years. Being able to improve regional and national R&D governance models over time will be a deciding factor in successful acceleration—not only of US economic competitiveness but also broader citizen well-being. Taking the time to learn from similar efforts in Europe offers a no-regrets step toward generating responsible research and innovation infrastructures for public value without boondoggles at public expense.