Over the past 70 years, Americans from all walks of society have deeply valued the importance of scientific progress. Reflecting those values in their work, government and corporate managers have invested heavily in talent and infrastructure, demonstrating the nation’s commitment to the value of science and technology (S&T) as an essential driver for a strong economy and a safe and healthy nation. Even as the United States benefitted from the brain drains of other societies and occasionally cooperated on selective technologies with close allies, advancing science and technology was understood as essentially a national undertaking. And from the beginning of World II until the end of the Cold War, Americans prided themselves on U.S. superiority in both its S&T institutional design architecture and the output of that architecture.
Through the years, however, the context in which we pursue scientific discovery and develop scientific capital has changed dramatically. Most observers are aware that U.S. government efforts in S&T appear to be less dynamic than in decades past, when innovation in military and space technology often trickled down into the general economy. Instead, private industry now leads the way in most cutting-edge sectors, leading many to worry that the public-private synergies developed to serve national policy could decay—including when it comes to controlling key military and intelligence technologies.
This story is true as far as it goes, but it may be the wrong story. The story that arguably matters more but is far less often told begins with the observation that we now live in a world saturated with trained scientists and engineers who have essentially open access to the tools and infrastructure of science. Science and its technological applications, understood as a single linked process, have been globalized like so much else. As a consequence, we often find ourselves forced back onto our heels defending a global leadership position and trying to sustain or develop that leadership across every domain of technology, both established and emerging.
We are playing defense because many nations are now aggressively copying our approach to R&D. And many can now afford to develop national strategies, building partnership and education pipelines at home and abroad, investing heavily, and establishing technologically advanced military-industrial complexes. At least 43 nations across six continents currently have a strategy for increasing internal scientific innovation and international scientific engagement, and the African Science Academy Development Initiative (ASADI) lists an additional 11 national science academies with strategies across the African continent. Many Asian countries have increased their support of science and technology development and technical education, as seen in the increased share of tertiary degrees awarded and original research published. Unsurprisingly, the U.S. share has decreased: The United States attracted 28 percent of all globally mobile students in 2001 but only 19 percent in 2012. Countries like China are also encouraging their U.S.-trained scientists to return home to build their internal resources. Asia has led the global growth in R&D since 2003, with China expanding most rapidly and accounting for more than a third of R&D growth and more than 20 percent of global R&D expenditures. This growth across Asia is founded not only on strong public support for R&D but also by large private spending in domestic and foreign industry in the region.
At the least, this means that the institutional edifice of U.S. superiority in S&T that emerged out of World War II and was codified, in a sense, by Dr. Vannevar Bush’s seminal and protean report, “Science–the Endless Frontier,” no longer can guarantee U.S. leadership in this critical domain. The genius of Bush’s vision rested in an architecture of institutional coherence that linked government, industry, and university-based science in a flexible but mutually reinforcing discovery and application system. Scholars have called it the “pipeline” system, among other things, and have shown how it adapted over time with necessity, coming more to resemble a “connected” system.
In any event, it worked well, driving decade after decade of discovery, feeding into real world results that improved the lives of every citizen. During this 70+ year experiment, we saw basic research mature into the applications and engineering that allowed an American to walk on the moon, that drove the extension of life for the average American by over a decade, that eradicated polio and ultimately allowed the development of cell phones, the global positioning system (or GPS), and the internet helping us to communicate, navigate, and understand the world in new ways.
In a real sense, however, we have become likely victims of our own success. We have done so well with the model we built in the mid-20th century that we usually fail to acknowledge how much the foundation on which it was built has changed, and now we fail to see how much it needs to change—from a more or less exclusively national enterprise into one that weaves in and out of a larger S&T universe. Research and development (R&D) efforts have followed the same model for decades, and today we remain focused on adapting solutions to yesterday’s problems in a world that has fundamentally changed. As dissemination and understanding of information continues to grow through open data and information initiatives, as more sensors and satellites span the environment, and as the internet expands its reach, it will become increasingly difficult to protect the U.S. technical advantage in applied research, or for any nation to develop new and truly unique advantages in emerging areas. We are no longer living in a postwar period in which the competition has been decimated and everyone else remains traditionally poor and poorly educated; rather, we exist within a thriving global science community, and we must learn how to leverage this reality to our benefit, not cower in fear of a changing future.
One way to describe what has changed is to focus on scale and saturation. Since 1945, the U.S. population has more than doubled and global population has tripled. Global investment in R&D has gone from amounts counted in millions of dollars to more than $1.77 trillion, with well over 70 percent of that coming from outside of the United States, and with 75 percent of U.S. funding coming from non-Federal sources. Trends in the United States show industrial and philanthropic funding of science on the rise, and the Federal funding of science flattening for the foreseeable future. But ironically enough, most state-supported universities have shifted focus from teaching to competitions for federally funded research and, with state resources declining and tuitions skyrocketing, Federal grants now support many students to such an extent that universities are now more dependent on Federal than on state funding.
Meanwhile, the scientific knowledge being produced globally has by some counts doubled every nine years in the post-World War II era. A study by Lutz Bornmann and Rüdiger Mutz give the number solely for cited publications documented in the Web of Science at 1,859,648 just in 2012. This is only a portion of global publications, and with this exponential growth of scientific knowledge a new set of challenges arises with respect to effectively turning this knowledge into capital. Trajectories are stable, rate is steady, but growth is therefore exponential. So our structures for understanding what our investments are producing, what new discoveries are being made, and what breakthroughs are ready to be harvested for security, health, and the public good have become inadequate. Communicating new discoveries to the full scientific and engineering community through publications and conferences designed for leveraging that knowledge cannot scale indefinitely. Transitioning the knowledge created out of basic science into technology that can affect human health, security, and wellbeing requires new structures and institutions that can work at the dramatically larger scale and speed of this new globalized science environment.
To give an idea of the scale of the challenge, note that scientific capital in the form of usable concepts has increased tremendously as measured by patents, with annual patents granted growing from about 50,000 in 1976 to more than 300,000 in 2014. A recent publication by Mohammad Ahmadpoor and Benjamin Jones shows a significant correlation between the two. When the researchers assessed 4.8 million U.S. patents and 32 million research articles for the minimum citation distance between the patents and prior scientific publications, they found that 80 percent of scientific publications and more than 60 percent of patents link to the other. One new question raised by this successful experiment is whether the efficacy of these processes continues to scale as the information and knowledge created grows exponentially. Likely, it does not.
To whom would U.S. leaders turn for an answer to such a question? Most Americans are unaware of it, but the U.S. government not only employs a huge number of scientists, mathematicians, and engineers, but it also has developed over the years a fairly ornate set of advisory bodies to assist leaders in both the Executive and Legislative branches of government.
Whereas in 1945 there were extremely limited sources for trusted science advice to leadership, we now have more than 215 Federal Advisory Committee Act (FACA) boards alone, tagged as Scientific Technology Program Advisory Boards, across the Federal government. A coordination function envisioned already in 1945 by Vannevar Bush now resides in the Office of Science and Technology Policy (OSTP), created in its current form in 1976 to advise the President and the Executive Branch. This broad-based advisory process goes on regardless of the comprehension of or the attitude toward science and expertise that resides in the person of the President. But other problems exist all the same.
The policy-focus of the OSTP office is largely separate from the budget and resource prioritization functions that occur in each Executive Branch agency policy office. Somewhere, somehow, a balance need be struck between the pressures on agencies applied by Congress and the very real budget constraints under which they labor, on the one hand, and the need to coordinate Federal S&T strategy in accordance with White House direction of presidential intent and priorities. Without direct influence on either funding or the independent strategy offices in each Federal agency, the OSTP has uneven influence in coordinating a national research agenda.
Additionally, Congress now has no official source of science/technical advice since the Office of Technology Assessment was abolished in 1995 during the Clinton Administration. So Congress tends to rely instead on lobbyists and local connections for trusted advice— a reliance that may not actually provide the depth and context needed for our Legislative Branch to reach fully informed decisions. Finally, we have more than 100 national and defense laboratories, federally funded research and development centers (FFRDCs), and university-affiliated research centers (UARCs) representing their own organizational priorities and seeking funding to sustain their livelihoods. This produces hundreds of sources of “trusted” advice, often competing and sometimes in conflict. So the balance has shifted from too little advice available to too much, with a significant portion of it conflicted and some of it compromised by special interests. New concepts for how to create objective, trusted, and non-partisan input in this environment are scarce.
The result is that the expansion of our advisory boards and sources of “trusted” advice for science has created a situation in which government leaders must determine who among the hundreds of competing sources of input is most trusted, a situation that in effect takes us back almost to where we started 70 years ago when we had no structures for providing sound advice. This challenge, too, calls for us to rethink our strategies for sustaining research, translating that research into solutions for our challenges and economic competitiveness, and fostering a pipeline of talent needed for future economic security.
Today we find ourselves engaged in many conversations about these challenges, but the solutions these conversations typically produce are largely centered around two aspects: protecting the institutions and successful approaches of the past, and stimulating an entrepreneurial culture. Both seem necessary, but insufficient.
The reason is that the entire landscape of research institutions and funding dependencies have shifted dramatically since 1945, and especially since about 1990. The primary challenge for remaining competitive in a world saturated in technical knowledge, a broad diversity of funding sources, and widely accessible tools and infrastructure is not primarily a lack of sustained support for research in academic institutions; nor is it a lack of worthy small projects and ideas. These are, of course, important functions to sustain. It can be a good thing that researchers are expected to seek funding, drafting hundreds of proposals for small amounts of funding. This drives a competition for ideas, and can result in solutions to smaller, more incremental, fundable, and manageable problems. It’s also important because the diversification of both performing institutions and funding sources gives us new opportunities and will of necessity often require diversification in our approach beyond the standard Federal funding model, which remains nation-focused.
But the real challenge is how to perforate our old national model in such a way as to be able to monitor and use scientific-technical innovation resources on a global scale, yet do so without relinquishing control over technologies critical to national security. The old problem, back in 1945, was how to avoid having Federal money drive out or marginalize efforts from local government, foundations, and private donors. The problem now is how to make do with relatively less Federal money in a resource mix as a whole that spills across national borders like water spills across the squares and shapes of a parquet floor.
What to Do?
So what should we do? In broad terms, the emphasis of policy going forward should rest in five basic themes: diversifying the focus on sources of funding; empowering the use of science to solve problems at every level; reconsidering what science advice should be and how it should be organized; leveraging modern tools to keep track of and to understand emerging knowledge so as to facilitate its application; and creating a sustainable commitment to Federal funding of basic science.
Diversify the focus on sources of funding. About 75 percent of U.S. R&D funding is non-Federal, yet the primary focus of most policy discussion is on the 25 percent that is Federal. On one level this doesn’t make sense, but on another level it does because the government can influence its own institutions far more readily than it can those in private hands. There have been attempts to address this shift in the epicenter of funding but it is difficult. The Department of Defense leaned forward in an effort to innovate the government’s way into a better relationship with Silicon Valley as a solution for a more innovative military—the Defense Innovation Unit experimental (DIUx)—but it got off to a rocky start. The incentives across these communities will need to be better aligned to make this a functional new business model going forward.
But beyond a focus on Federal funding we also actively prioritize the work it produces. We stress the importance of independence and external review for sustaining quality, rigor, and reproducibility in Federally supported science, but as a nation we do not enable access to peer review for non-Federal funds, or pioneer alternate pathways for assessments that could lift trust for the results of the broader 75 percent. How do we elevate the results of the majority of funding, and make the rigor and independence of industrial funding transparent while protecting intellectual property rights? How can we help philanthropic and peer-to-peer funding access approaches to allow all funders to leverage analytics and understand what already exists, so as to highlight unique new discoveries and to link existing knowledge to pathways for transition and harvesting?
Empower the use of science to solve problems at every level. We have a clear discussion across America at the moment about the value of science, and the trend is clear: Increasingly, the average citizen does not trust science or expertise in general. Some of this has to do with a flattening of social authority in the culture in nearly every respect. But much of the lack of trust may be associated with a general disillusionment with the Federal government as the sole source of solutions.
Additionally, with the adoption of test-focused curricula in the K-12 classroom, increased emphasis on reading and math has decreased students’ broader exposure to science and experiential learning. By making science the purview of the highly educated and elite, the nation is reducing the opportunities for the application of scientific solutions to local problems. Instead we should enable and facilitate effective, high-quality approaches by empowering people at every level—individual, local, state, and global—with all variations of education to leverage science to solve their own problems.
We can work to develop approaches that reject the exclusivity of past processes that drive non-scientists into situations of blind trust or distrust, while ensuring that rigor and quality are maintained. And in the process we can improve our ability to harness the entire American R&D investment for true competitiveness in a new world saturated in knowledge, funding, and S&T infrastructure. Peer-to-peer funding sites are one interesting opportunity that might be evolved into “match-making” services between problems and solutions, allowing communities to identify affordable approaches to solve challenges.
Reconsider what trusted advice looks like. In a world where we have hundreds of “trusted” Federal scientific advisory boards, federally funded research and development centers (FFRDCs), university affiliated research centers (UARCs), national laboratories, NIH laboratories, Defense Department laboratories, and more, all vying for a limited Federal funding pot, how do decision-makers know who really to trust? Combine this challenge of scale in our advisory system with the challenge facing the OSTP to exert specific influence across the expanse of Executive Branch agencies. We have a problem.
Perhaps it is time to consider the formation of a team within OSTP that is focused on analysis, not budget—a group that has deep access to data, both open source and government proposals and grants, whose members are trained in ops research with true analytic skills and no competition for the budget in specific technical area. This would be a group that could look at the pros and cons of emerging scientific areas to articulate both the benefits and challenges that will inevitably follow discovery.
Leverage modern tools to understand current knowledge and to translate that knowledge into solutions. In a world where millions of articles are published annually, can we believe that the full results of our support will ever be understood simply by scientists reading journals and attending conferences? Currently, government program managers have limited access to external data sources and accompanying analytics to ensure an objective view of the state of science within their fields. We stand up panels based on the expertise of individuals without arming them with the very tools scientific funding has developed over the past decade: tools that can objectively query and question the uniqueness of the projects funded. These resources are less accessible to philanthropic and peer-to-peer funding and lack external transparency and acceptance when used by industrial sources.
There are commercial tools available and many bespoke analytic products produced internally by organizations across industry and government that allow the analyst to digest, parse, search and analyze relationships or similarities and differences among millions of scientific journal articles, patents, and funding documents for public and private companies, helping to navigate the full array of ‘innovation’ information and find the highest value opportunities. An analyst may search on topics across the full array of data to find imaging technology of value to Defense only published within colonoscopy research that would never be noticed by scientists if domain expertise were the sole foundation for decisions. But data is the key – currently the analysis is only as good as the data you can access, and data is expensive and protected. Large publishing houses make it difficult to afford access to the full set of publications and patent data is published without any regard to machine readability. The government could provide significant benefit to the public by finding ways to make data and basic analytic tools available to organizations that want to bring science to bear on our national and global challenges.
Sustain commitment to public funding of science. If the past has taught us anything, it is that Federal support for basic science has tremendous value. We have created generations of scientists and engineers who want to make the world a better place, and they have done so time and time again. We know, too, of times when we have believed a challenge to have been solved only for it to reemerge: Antibiotic resistance is an excellent example; we declared victory, but nature adapted, and now we are in a struggle mirroring that of the early 1900s. Funding for basic research must therefore persist at stable and hence predictable levels regardless of the success or failure of specific applications.
However, when 75 percent of the funding for science comes from outside government, perhaps we need a broad shift in our concept of the government’s role. Currently, the U.S. government tries to cover the entire landscape of scientific fields from basic math to cures for cancer. It makes more sense for the Federal agenda for science to focus on areas where industry and foundations are not supporting scientific development, or areas where there is no dual-use potential, and on areas where the nation cannot depend on industry and philanthropy—such as critical vaccine development or weapons systems. By better focusing the 25 percent of the scientific budget that the Federal government controls, we can advance these areas faster and leverage innovation from outside government better than we do today.
Alas, we have more questions about the way forward than we have answers. But good questions are more important than and are prior to good answers. We have ideas, we have suggestions, but we are concerned about the lack of robust questioning in science policy today. We must rise above a protectionist stance of our history and ensure that national leadership to understands that science policy is a deadly serious and important domain within national policy, including our national security policy. There are few signs that either major party really gets that. That’s a problem that may spawn many questions. We had better find some good answers before it’s too late.