Keeping America “Science Strong”

As a member of the baby-boom generation, I have not experienced a time when US science was not the envy of the world. With US researchers claiming 60 percent of all Nobel Prizes ever awarded in the sciences, America dominates this international competition more decisively than any other nation has in any pursuit that keeps score (1). Our citizens have also benefitted enormously from our investment in research—with a robust economy, good jobs, novel therapies, and numerous products that enhance our quality of life. US leadership in science has protected our freedoms and sovereignty, provided US industry with competitive advantages through advanced technologies, allowed us to set standards consistent with our ethics and values, and served as a welcome form of international diplomacy.

Unfortunately, I see threats on the horizon to US science leadership, a few examples of which are shown in Fig. 1. China now awards more degrees in Science and Engineering (S&E) than does the United States (see figure 5 in ref. 2) and is increasing its investment in Research and Development (R&D) at twice the rate of the United States (see figure 11 in ref. 2). The fraction of US research publications is falling as is our share of top (1%) papers (see figure PBS-9 in ref. 3). China eclipsed us years ago in number of new patents awarded (see Fig. 1, Bottom) and has more corporations in the global Fortune 500 than does the United States (5). While I believe it is of benefit to the entire world when any nation uses science to improve the livelihoods of its citizens and become better steward of this planet, the United States accrues distinct economic, political, health, security, and other dividends from science leadership. Here, I present six actions that the United States should promote immediately to prevent further erosion of our science leadership.^*

Fig. 1.

Examples of measures of how international leadership of US R&D is eroding. (Top) Total investment in R&D by eight nations with advanced economies. Reproduced from ref. 2. Notes: PPP is purchasing power parity. Data are for the top eight R&D-performing countries or economies. (Bottom) Comparison of United States to China in patents awarded per year. Reproduced from ref. 4. Note: EPO is the European Patent Office. Data are from 2022.

1. Improve K-12 Education

First, I see an immediate crisis in STEM (science, technology, engineering, and math) education at the K-12 (kindergarten through high school graduation) level. While US colleges and universities are still world class, a large fraction of graduate students (more than 1 in 3) and postdocs (more than half) in science, engineering, and medicine are international students^† (see table 1 to 3a in ref. 6). Even at the bachelor’s degree level, a full one-fifth of the S&E workforce in America is foreign-born (see figure LBR-13 in ref. 7). This alone is not a problem as long as the best and brightest of those international students continue to remain in the United States. However, as other countries increase their own investments in science and technology, we are seeing those students return home to contribute to their own nations’ advancement. Our current reliance on foreign talent represents a major vulnerability to US science leadership, which can only be effectively mitigated by growing our own STEM talent.

At present, the prospects of developing sufficient domestic STEM talent do not look promising (Fig. 2) (8). Only 36 percent of US 4th grade students are rated as “proficient” in their understanding of mathematics; by 8th grade that percentage falls to 26 percent. By international standards, US students fall well below nine other developed nations in science performance, and the US ranking is even worse in mathematics. It should be no surprise, then, that the United States ranks 18th among industrial nations in the fraction of university students who receive STEM degrees (9). We simply are not preparing enough of our high school graduates to pursue STEM careers.

Fig. 2.

International comparison of preparation of students for STEM careers. US students rank below the middle of the group in mathematics and middle of the group in science. Reproduced from ref. 7. *P < 0.05. Significantly different from the US estimate at the 0.05 level of statistical significance. TIMSS = Trends in International Mathematics and Science Study. Note(s): TIMSS participants include countries, which are complete, independent political entities, and nonnational entities (e.g., Hong Kong). Advanced economies are based on the International Monetary Fund (IMF) designation of advanced economies (IMF 2022). IMF does not classify Russia as an advanced economy, but it is included in this analysis because it is a large economy with high levels of student achievement. Education systems are ordered by average mathematics score.

Fortunately, there are steps we can take immediately to build a robust US-born STEM pipeline. We should start by fostering in the youngest learners their innate curiosity to explore the natural world and discover science themselves, rather than teaching science as a list of facts. Informal learning centers, such as science museums, zoos, and aquaria do this well and should be well supported, including creating opportunities for low-income families to visit. A second part of the problem is that in elementary school, most teachers majored in elementary education and feel ill-prepared for STEM course content. While some very promising AI-assisted tools are being developed to assist teachers and students with math and science content, we should also consider reinstituting teacher uptraining programs at the National Science Foundation that were introduced during the Sputnik era but have since been discontinued. None of these steps should be construed as arguments for further eroding education in the humanities, arts, and other fields. We need a science workforce that understands history, communication, ethics, and other subjects and that creates informed citizens who understand the place of science in a larger context.

2. Reduce Red Tape

While we are already experiencing some erosion in international students seeking to study in the United States (10), our universities will continue to represent a beacon of opportunity for students from abroad. Recently, however, they are encountering major hurdles when applying for student visas. Inexplicably, visa denial rates are growing even at a time when the total number of visa applicants is shrinking (11). Students from nations that previously encountered few obstacles, such as Canada, Brazil, and many others, are now being denied entry. It would be well worth an analysis of what is causing these student visa denials to determine whether they are truly warranted.

Later, when those students obtain their degrees, they (and their would-be employers) experience further roadblocks in obtaining H-1 visas to work in the United States, despite the United States being unable to meet the demand from employers with US-born talent alone. Several politicians have called for providing a work permit to every international STEM graduate, which seems a low-cost approach to retaining these vital workers.

The “red tape” problem is not confined just to foreign students. Between 2013 and 2023, the cumulative total of regulations governing federal research in the United States grew by 172 percent (12). University faculty now complain that more than 40 percent of their research time is spent on meeting federal requirements rather than doing research (13), which is not only an inefficient use of their time and talent but also erodes productivity. While this is a problem, it is also an opportunity. Some technology companies are exploring the use of AI to perform some of the more time-consuming and routine aspects of compliance.

3. Create a National Research Strategy

The US scientific enterprise has evolved into a very complex system in the last 50 years. Industry now funds and performs the vast majority of the R&D in the United States. Even in our research universities, philanthropic funding (both current from foundations and legacy from endowments) currently supports almost as much research as the federal government. Despite the complexity of the system, there is no overarching plan that helps ensure that the various funders and performers are coordinated for maximum impact. In contrast, many other nations maximize their research investments with strategic plans. The Economist recently highlighted some of the impacts of China’s strategic planning on the fraction of high-impact papers in various fields (14). While United States at ~80 percent is far ahead of China at ~20 percent in high-impact papers in space sciences, those percentages are reversed in materials science, a field essential to national security, combating climate change, and other important societal challenges.

The United States invests a substantial amount of funding in R&D from both private and public sources, but surely, there are opportunities to increase the payback from those investments through strategic planning. The CHIPS and Science Act calls upon the White House’s Office of Science and Technology Policy to create a national strategy for US research that includes the private, public, and philanthropic sectors. The challenge will be to develop a blueprint for the future of American science that is acceptable to the many different missions, objectives, incentive structures, and cultures present in the United States innovation system. At the same time, such a plan also must not close out opportunities for serendipity, as paradigm-changing discoveries can emerge at any time in unexpected places.

4. Modernize University–Industry Partnerships

Universities have long recognized the importance of allowing their faculty, especially those in STEM fields, to engage with industry. Through the several pathways for such engagement, ranging from individual consulting contracts to large industry consortia hosted at universities, the faculty gains insight into the real world needs for solutions, and industry gains early insight into new discoveries that could have commercial applications. Today, however, the need for university–industry partnerships has never been greater. With the majority of R&D being produced in the private sector (Fig. 3) (15) but little of it getting disseminated through publications, the exchange of new findings should be two-way. Furthermore, with the majority of STEM graduates being employed in industry, such partnerships need to extend to the pedagogical aspects of university education to ensure that students are adequately trained for the jobs they will be seeking. Finally, graduate students and postdocs are underpaid considering their skill level and value to the research enterprise. Summer internships at commercial ventures could help supplement their meager stipends.

Fig. 3.

US total R&D expenditures by source of funds from 1953 to 2020. Industry now funds (and produces) 75 percent of the total output from the United States. The federal government funds about 20 percent of research, primarily at universities and government labs. The remaining 5 percent is supported by foundations, legacy philanthropy from university endowments, and state and local governments. In terms of research conducted at universities, foundations and legacy philanthropy is nearly equal to the federal government investment at this time. Reproduced from ref. 14. Note(s): Some data for 2021 are preliminary and may be revised later. The data for 2022 include estimates and are likely to later be revised. Federal performers of R&D include federal agencies and federally funded R&D centers. R&D funding listed as other combines data from nonfederal governments (state and local) and nonprofit organizations. For more information, see tables 2 and 6 of National Patterns of R&D Resources (2021 to 2022 edition).

Several universities, particularly those with long-standing geographic connections to major high-tech industries, are starting to revamp their relationship with industry (16), such as relaxing restrictions on time spent on industry engagement. Keck Graduate Institute (17) is one example of an entirely new sort of educational institution created to effectively train students for high-paying jobs in the biomedical sector. Conflict of interest policies will also need to be reexamined to ensure appropriate disclosures and to avoid even the appearance of impropriety.

5. Ensure Access to Major Research Facilities

Fifty years ago, it was possible for an individual university to host a major research facility, such as a particle accelerator or a telescope. Today, the frontiers of science have been advanced so far beyond what they were then that it takes an entire nation, or more likely, many nations working together to build, maintain, and get maximum benefit from a major science facility. Recent examples of international research facilities are CERN, the International Space Station, and the International Ocean Drilling Program, all international collaborations for which there is little concern that collaboration threatens our economic competitiveness or national security. Nevertheless, there have been recent examples of the United States shying away from involvement in international programs. Even when we are engaged, we are often a fickle partner. Unfortunately, the United States cannot be a science leader if it isn’t even at the table.

To reverse this trend, the United States could begin by inviting international partners to join us in US-led initiatives. We should at the same time become a collaborator of choice by committing to international projects for defined periods of time. It would help if there were agreement on policies for where and when we will enter partnerships, as well as procedures for evaluating the success of collaborations periodically, including criteria for shutting down a project that has run its course.

6. Build Public Support for Science

Finally, none of the above can be accomplished or is even worth accomplishing if we do not have public support for science. Several high-quality surveys (18) demonstrate that public trust in science has been on a downward trajectory since the COVID-19 pandemic (Fig. 4). Public opinion surveys by the Annenberg School of Public Policy at the University of Pennsylvania demonstrate that public trust in science rises when the research is credible, prudent, unbiased, self-correcting, and beneficial to average Americans (19). These qualities are closely aligned with the norms of science that we all aspire to, but often fall short in demonstrating.

Fig. 4.

Results over the last 5 years from a Pew survey on public opinion on the value of science to society. Since the pandemic, trust in science has eroded. Reproduced with permission from ref. 17. Note: Respondents who did not give an answer are not shown.

Actions that have the potential to reverse the downward trend in trust include rewarding researchers at all levels for producing research that is excellent and trustworthy, not just splashy. The reward system should also recognize excellence in public engagement by scientists, to create two-way dialogs to better inform both citizens and scientists. Universities should also provide training in research ethics, and many are now doing that. Finally, we simply must find effective and scalable strategies to counter misinformation, an insidious vehicle for eroding trust in science. Historians such as Will and Adrian Durant have noted that a great civilization is not conquered from without until it has destroyed itself from within. The rejection of scientific and evidence-based findings is the first step in that internal decay.

Since I presented the State of the Science address, I have been overwhelmed and very pleased to hear from many others across the STEM enterprise offering ideas, solutions, and partnerships. I was gratified to hear Dario Gil, in his inaugural address as the new chair of the National Science Board, echo many of these same observations and solutions. Of course, the hard work begins now: To investigate exactly how to implement these six actions to promote US leadership in science and to determine what other steps will be essential. Reversing the troubling trends indicated above will require new thinking and creative solutions across all government, industry, academia, and philanthropy. We at the National Academies look forward to working with all our partners to ensure that science continues to support the American dream.