The Limiting Factor of “The Endless Frontier” Is Still a Human One
In science policy circles, making science more inclusive and diverse has not received the attention it urgently needs. To remain a leader in innovation, the United States must diversify its science, engineering, and technology workforce.
Vannevar Bush’s Science, the Endless Frontier report has been invoked, quoted and misquoted, poked and dissected many times since its release in the summer of 1945, often to highlight its influence on science policy and its call for federal funding to be directed to the research enterprise through the nation’s universities. The report bears rereading—not only to appreciate Bush’s vision, but also to consider how that vision has been tested as it met the societal and political realities of the past 75 years. And, reading it with a different lens, I have found the report compelling for its messages on education and talent development, causing me to reflect on Bush’s general failure to envision how diversity, equity, and inclusion apply within that endless frontier.
In rereading the document, I was struck once again by the fact that Bush gave equal billing to setting up the infrastructure to support research and to addressing what one might call “people issues.” One of the four advisory committees that Bush established was devoted to talent discovery and development. This 14-person, all-male committee considered many approaches, including drawing talent to science via scholarships for promising students and mobilizing talent through the education of returning veterans. Both of these worthy paths would carry hidden barriers to diversity in science, technology, engineering, and mathematics (STEM). The extraordinary expansion of higher education that resulted from the GI Bill, for instance, almost exclusively benefitted male veterans, with schools generally preferring to admit male students and many women’s colleges admitting men for the first time. At the same time, students and veterans of color saw their opportunities restricted by limited access to high-quality K–12 education and the racism of Jim Crow laws mandating segregation in education and elsewhere.
Nonetheless, at least one member of the committee, James B. Conant, a chemist and president of Harvard University, saw the inextricable link between the advance of research and the development of talent. In the report, Bush quotes Conant’s statement that “in every section of the entire area where the word science may properly be applied, the limiting factor is a human one. We shall have rapid or slow advance in this direction or in that depending on the number of really first-class men [sic] who are engaged in the work in question…. So in the last analysis, the future of science in this country will be determined by our basic educational policy.”
Today, the United States faces new technological and geopolitical challenges, but the limiting factor is still a human one. Women and persons who are Black, Hispanic or Latino, and American Indian or Alaska Native remain consistently underrepresented in STEM fields for reasons that are manifold, complex, and structural. Efforts to date have been insufficient to address the roots of these issues. In science policy circles, the human resources challenge of research rarely receives the attention it urgently needs and deserves for the United States to remain a leader in advancing the next generation of science. The country is not always guaranteed to have easy access to talent from elsewhere; other nations are strengthening their own research and innovation capacities, while the United States has restricted policies for who can and cannot enter.
At a time when China’s growing number of scientists and engineers poses a challenge to US competitiveness, the country needs to embrace its diversity as an important advantage. The United States is uniquely positioned to bring the varied cultures, perspectives, and lived experiences of its population, especially when coupled with America’s ability to attract top talent from around the world, to solving problems and propelling science forward. Leaders and funders of US science should embrace this “diversity dividend” that sets the United States apart. Realizing the nation’s scientific potential will depend on finally adopting necessary—and long-overdue—systemic changes to its institutions that will allow it to build and foster a robust talent pool, leading to a more diverse STEM workforce.
A STEM workforce that does not look like the country
The potential talent pool for science and engineering is diverse and becoming more so every year. In 2021, the resident population of the United States aged 17 and under was about 50% white and 50% non-white. Women made up nearly 60% of the college-enrolled population.
However, this diversity is not reflected in STEM fields. The majority of US citizens and permanent residents who received bachelor’s and master’s degrees in science and engineering in 2019 were white; 69% of doctoral degree recipients in those fields were white. Black, Hispanic or Latino, and American Indian and Alaska Native students are vastly underrepresented in STEM degrees. For women of color, these disparities are even more apparent, and they carry over to the workforce. The precarious position of underrepresented minority scientists and engineers was highlighted by their experience during the COVID-19 pandemic, which disproportionately affected women and underrepresented minority students as they pursued STEM education and employment.
For decades, the US research enterprise has relied heavily on international talent to conduct graduate and postdoctoral research. The highest reported recipients of research assistantships are temporary visa holders. According to the National Science Foundation’s Survey of Earned Doctorates, in 2020 39% of US science and engineering doctoral degree recipients were foreign citizens with temporary visas. In 2020, temporary visa holders earned the majority of doctorates awarded in engineering (59%) and in mathematics and computer sciences (59%). This reliance on early-career international researchers is counterproductive to the nation’s interest when these scientists then face restrictive visa policies that force them to leave the United States. In any case, as more countries invest in and expand their own research and development spending and science enterprises, and as the United States is increasingly seen as an unwelcoming environment for immigrants, there is no guarantee that the currently high percentages of foreign-born PhD graduates who intend to remain in the United States will be sustained, posing risks to the future of US science.
How barriers to diversity limit the endless frontier
US higher education desperately needs to bring about the equitable system necessary to compete in a world that has changed significantly since Vannevar Bush’s time. Despite nearly a half century of calls to diversify STEM, today the institutions that are more accessible for students from lower socioeconomic backgrounds and underrepresented minority groups—such as two-year colleges—tend to be underfunded and poorly linked to STEM career pathways.
Students from underrepresented minority populations also encounter other barriers. Attending predominantly white research institutions is often challenging, whether in STEM or any other field. Aside from facing overt and covert barriers to entry, imagine undertaking the difficulty of a STEM education while being made to feel unworthy, unwelcome, isolated, and out of place in the community of scholars.
Students from these groups who are able to enter graduate education face additional significant barriers in comparison to their white counterparts: they are more likely to carry higher levels of undergraduate debt and are more likely to accumulate additional graduate education-related debt, even in fields like computer sciences and engineering that are known for providing a debt-free doctoral education. The kind of funding students receive matters because the value of research assistantships (which, according to the Survey of Earned Doctorates, support a third of all science and engineering doctoral students) goes beyond the money—they also carry attention and mentoring opportunities. Yet Black doctorate recipients, independent of field, are less likely to report research assistantships as sources of support and more likely to indicate using their own funds to pursue their research interests. In the most recent Survey of Earned Doctorates, even when field is held constant, both cumulative and graduate educational indebtedness is higher for underrepresented minority graduates than white and Asian graduates, with the highest levels of all for Black students.
All of these barriers to increasing the diversity of scientists also limit the frontiers of science by restricting the types of scientific inquiry that the United States supports and pursues. Stanford University chemist and inventor Joseph DeSimone once said that “it’s kids who came out of poverty who think first about the cost aspects of what they invent. I see that affecting technology at a big scale—you’re always looking for someone with a different idea.” Bringing such talent into STEM professions can support culturally responsive technology and innovation, and it also reduces US dependence on international sources of talent, builds community prosperity, and reduces income and wealth inequality.
Systemic change, but not enough
Fifty years ago, in the wake of the movements for civil rights and women’s rights, many institutions mounted small-scale efforts to recruit women and those from minoritized communities into science, engineering. and medicine. But after nearly a century of programmatic barriers to participation in STEM, the efforts were hit or miss at best. For women in medicine, for example, early initiatives included removing the quotas that barred them. Interestingly, removing those “do not enter” signs in the early 1970s led, over time, to women making up more than half of those enrolled in medical schools today. However, this representational parity does not translate into power parity since the policies, processes, practices, and cultures of the systems in which women study, from which they graduate, and in which they work have not been transformed. It is still expected that women will change or adapt to fit the system—as was made clear by the monumental 2018 NASEM report on sexual harassment, which showed that women still do not feel equal, safe, and seen in STEM fields.
Signed in 1965, President Lyndon Johnson’s Executive Order 11246 attempted to address discrimination. The order required equal employment opportunity and affirmative action by federal contractors in recruitment, hiring, training, and other employment practices, a move that spurred many industry leaders to mount and support initiatives to build the STEM talent base. Universities, colleges, and nonprofits generally responded by establishing small-scale efforts intended to introduce students to careers, counsel them on coursework to take, and offer them educational experiences that would prepare them to succeed in higher education and beyond. Yet little changed in the institutions themselves. And instead of pursuing strategies that would help all students succeed, most institutions of higher education created “shadow projects” to enable students to navigate the system as it already exists, never questioning that a more effective strategy might be to fix the system that failed too many otherwise-capable students. After five decades of supporting such small-scale intervention programs, higher education institutions need to shift the target to fixing the system to remove barriers to success for all.
The Endless Frontier report proposed to solve the “people problem” by providing scholarships to attract students and spur talent development in STEM, observing correctly that funding was a barrier for most students’ college attendance in the late 1940s. The amount and mix of federal support for STEM higher education changed over the ensuing decades, expanding in the post-Sputnik days and shrinking in the 1980s with a change in all aid from grants to loans. Subsequent shifts of institutional aid from need- to merit-based support have created further barriers to students from minoritized populations who are frequently from low-income families and are disadvantaged by an often-segregated and less-than-stellar K–12 education. Compounding these challenges, changes in the structure of aid require that many students hold jobs while in school, taking away critical time to engage in intellectual pursuits, activities, and opportunities available to those with financial security.
Today’s limited scholarships and support not only fail to deliver The Endless Frontier’s promises to disadvantaged students; they also undermine other efforts to bring these individuals into academia. At the graduate level, low levels of financial support are particularly limiting for underrepresented minority students, who often enter the workforce rather than taking on further study because they lack a family financial safety net. And in families with no history of graduate education, students can struggle to justify incurring debt for more education rather than supporting their families. As a result, students who could become the very sort of faculty member that many of them have never seen are likely to look at the sacrifice in time and salary required to join the professoriate and consider it a bad bargain.
Society loses when it fails to make the science and engineering talent pool inclusive. The loss of talent is not the only problem that exclusion introduces; it also results in a loss of innovation. The business researcher Rembrand Koning and his coauthors documented that “women’s biomedical inventions are more likely than men’s inventions to focus on women’s needs.” Researcher Fiona Murray, in an article in Science, points out that innovators focus on problems that they see around them. She highlights the example of Patricia Bath, an ophthalmologist who was struck by the difference between the incidence of blindness she saw as an intern as she traveled between the eye clinics at Harlem Hospital and Columbia University. Bath, an African American woman, created many sight-saving programs and the field of “community ophthalmology.” She also created a new device and method to remove cataracts, the laserphaco probe, that could make this critical care more accessible to the communities she served. To be competitive in the global economy, and to be whole as a society, the United States cannot afford to lose innovations like hers—or the innovators themselves.
Supporting institutions that foster diversity
It is important to recognize that historically Black colleges and universities (HBCUs) play an outsized role in educating students at the bachelor’s degree level who then go on to obtain PhDs in STEM. The same is true of high Hispanic-enrolling institutions. Unfortunately, these schools continue to be undervalued and receive significantly less support than they merit. To capture the full diversity of America’s talent, these schools should be supported and emulated because they successfully attract and train diverse STEM talent. Leaders of these schools understand the necessary large-scale, structural changes that are key to building institutions that can support all students, not just some. Only by transforming government agencies, colleges, universities, and related components of the STEM ecosystem will the United States be able to capture the benefit of the different perspectives and experiences that flow from diverse voices needed to ensure a strong future for the scientific enterprise.
The successes of minority-serving institutions and initiatives demonstrate that it is time to stop relying on experiments without first putting into practice protocols that researchers and administrators already know to be effective. For example, there is a robust bank of discipline-based educational research that demonstrates the value of improving the quality of introductory courses by using active learning, reducing hyper-competitiveness, acknowledging and supporting student identity, and building a sense of belonging in the field. Centers for teaching and learning, which allow faculty to experiment with new teaching methods informed by research, already exist on many campuses and could help faculty members hone the skills necessary to engage students from diverse races, ethnicities, and backgrounds in their subjects. Faculty resistance to such educational training is often mentioned as a reason not to insist upon it; but the pandemic showed that necessity can spur many actions that people might otherwise resist. Furthermore, all research institutions could benefit from learning from minority-serving institutions about how to support student success and welcome all the identities that students bring. I urge the scientific community to seize this moment to have discussions about the quality of instruction and about how institutions can assume an asset-based rather than a deficit-based approach to students of color.
Similarly, two-year colleges represent unique assets in the STEM talent ecosystem, given their focus on affordability and access. These schools offer unique opportunities for learners, as well as for the scientific enterprise, which could use this system to identify people with STEM talents. These institutions must be supported and connected, so that they become genuine on-ramps to STEM degrees for the students they serve.
It is clearly time for action on diversifying STEM, but this is also a time for reflection. Why, despite decades of discussion, has so little progress been made? More than 50 years ago, Robert K. Merton coined the term “the Matthew effect” to describe how eminent scientists receive the most credit for innovations and how advantage accumulates. (The term is based on the Bible verse “For whoever has, to him more will be given, and he will have abundance; but whoever does not have, even what he has will be taken away from him,” Matthew 13:12.) The Matthew effect is often cited to explain the current system of advantage for major research institutions and the faculty stars within them. But it takes more than that to understand why there are currently no HBCUs among the highest ranking (R1) research institutions. Almost all HBCUs were established after the Civil War, and only a small number were included in funding under the first Morrill Act (1862). Although the second Morrill Act (1890) supported the establishment of public Black colleges, it allowed states to perpetuate educational segregation and systematically underfund education for Black students. Private HBCUs did not enjoy the largesse of big donors to support research ambitions. Largely situated in the Jim Crow South, these institutions struggled with deliberate underinvestment by their home states. Small and underresourced, they were not prepared to handle the influx of Black veterans who were eligible for and sought entry with support from the GI Bill. In 1994, tribal colleges received investment, but ultimately there has never been a major investment in either them or HBCUs that would enable them to become R1 institutions. The absence of a high-level research infrastructure in these institutions results in the loss of a potentially substantial network for addressing issues that differentially affect communities of color.
The pandemic has shined a spotlight on how failure to study these important issues through a lens of racism and race-awareness can affect society as a whole. Recognition of the way COVID-19 affected communities with health disparities has dovetailed with a reckoning about why highly qualified Black scientists are not funded at rates commensurate with their research grant application rates. One analysis judged that they want to study the “wrong things”—their own communities. Such research can receive a lower priority score when funding decisions are made. A researcher friend of mine described an invention by four of his Black women students: a “virtual traffic stop” at which no one, neither police nor driver, has to leave their cars. It doesn’t take a lot of explaining to see why this invention might have been seen as a positive social good. So understanding how the Matthew effect has sent resources to the richer schools and how the development of HBCUs has been constrained also reveals who controls the research and innovation agenda.
Avoiding a “tremendous waste”
When I consider the myriad ways that the US scientific enterprise has failed to diversify, I am reminded of a quote from Science, The Endless Frontier: “There are talented individuals in every segment of the population, but with few exceptions those without the means of buying higher education go without it. Here is a tremendous waste of the greatest resource of a nation—the intelligence of its citizens.”
The United States will not realize the “diversity dividend” without mindfulness in the choices of decisionmakers: requiring that researchers demonstrate the presence of diverse teams; asking major research universities to embrace partnerships with institutions with which they do not usually partner; investing in research and business development from diverse teams. Efforts must be made to find and develop talent where it is—including at two-year colleges and in workplaces. A shift in thought is required to see diversity as an advantage rather than a deficit.
And for the big institutional research powers that emerged from Bush’s vision? It may be important to temper the Matthew effect with the Luke principle (based on the Bible verse “From everyone who has been given much, much will be demanded; and from the one who has been entrusted with much, much more will be asked,” Luke 12:48). Institutions that have been the major beneficiaries of public investment in research and development over the last eight decades have an opportunity, indeed an obligation, to enable a future characterized by inclusion that delivers on the excellence and innovation needed to address the next 75 years of challenges.
Imagining science policy for the next 75 years requires embracing a broader vision of who will do science and engineering and who will invent the future. The pandemic has raised awareness of how a failure to study these important issues through a lens of racism and race-awareness can affect society as a whole. And the lessons for the future are clear: scientists, engineers, and medical professionals must look more like US citizenry. True competitiveness means recognizing and using the country’s diversity as an innovation advantage, harnessing the diverse perspectives that offer better ideas, better solutions, better science, and better and more inclusive technology.