The STEM talent pipeline (science, technology, engineering, and mathematics) is a critical resource that underpins the United States’ innovation, economic growth, and global competitiveness. Recent political decisions could throw this resource into turmoil.
The Trump administration is pushing to revoke large numbers of student visas held by Chinese entrants, and Secretary of State Marco Rubio has made it clear that STEM is at the center of this move, vowing that the United States will “aggressively revoke visas for Chinese students, including those with connections to the Chinese Communist Party or studying in critical fields.”
The stakes are high. Abruptly capping enrollment of foreign students in favor of domestic students will have unintended consequences. Statements about maintaining US leadership in artificial intelligence, quantum computing, defense, or science and technology in general are meaningless if the STEM pipeline is not continuously filled with the best.
In computer science alone, the number of students with STEM bachelor’s degrees grew from approximately fifty thousand per year in the mid-1990s to three times that number in the early 2020s. Employed students with bachelor’s degrees in this field show a similar trend, growing from one million to two million between 2003 and 2023. At the same time, the international composition of this pipeline has evolved significantly. Understanding the changing roles of both domestic and foreign-born talent is essential for making informed policy decisions and managing risks and opportunities.
Recent analysis reveals that while domestic students represent about 90 percent of STEM bachelor’s degree graduates, the landscape shifts dramatically at the graduate level and beyond. Since the mid-2000s, foreign-born students have come to constitute more than half of all master’s and doctoral degrees awarded in key fields such as computer science, mathematics, and engineering. This trend translates directly into the workforce, where foreign-born graduates make up about 50 percent of employed STEM professionals across various subfields.
Threats to the talent pipeline
Attracting science and technology talent from abroad is a phenomenal deal for the STEM ecosystem in the United States. The country of origin provides education, selects the top talent, and often pays for students’ expenses in the US education system. Chinese and Indian students represent approximately 70 percent of all foreign-born STEM graduates, with retention rates post-graduation remaining remarkably high: around 77 percent of STEM PhDs stay in the United States for a decade after completing their degrees.
It seems to be taken for granted that this influx under these—from the US point of view, very favorable—conditions will continue. The United States needs, of course, to maintain a welcoming environment for foreign talent, as they are crucial for leadership in STEM innovation. But such a reliance on international talent poses risks, particularly given the changing landscape of US immigration policy and global dynamics.
Recent shifts in policy have led to concerns about potential disruptions in this talent pipeline, underscoring the need for a more balanced approach to talent development. A majority of the foreign talent in the United States still come from China, a country that over the past twenty years has developed into what many Americans consider our greatest adversary. Much has been written about the transfer of critical intellectual property collected from both willing and unwilling foreign students learning and working in the United States.
The talent pipeline could even be closed at both ends. Potential Chinese emigration policy restrictions, alongside US immigration changes, could devastate the science and technology workforce in a few years. In the long term, increasing economic opportunities in the home countries as well as demographic trends—since 2020, China has had a declining workforce, and India’s will follow in fifteen to twenty years—will only amplify this risk.
This means that even though India during the past decade has begun to displace China in supplying students to the United States, one sole source would merely replace another. (Traditional immigration sources for science and technology talent, like Europe and the former Soviet countries, are no longer major contributors.) Geopolitical and economic tensions can evolve quickly, as demonstrated by the US-China competition. Given the number of talented STEM students needed to continuously fill the pipeline, risk management demands that US dependence be more diversified.
How America can adapt
The challenge demands a multi-pronged strategy.
First, give greater priority to investments in K–12 education to prepare domestic students for advanced STEM studies. The United States ranks twenty-fifth out of thirty-seven OECD countries in math literacy, with many high school graduates unprepared for college-level STEM courses. About nineteen million students are enrolled in undergraduate STEM education, compared to about fifty million in primary and secondary education.
Reforming and enhancing STEM education at the foundational level is crucial for nurturing a future generation of domestic talent. It will take a decade or more to achieve a measurable onshoring of US talent, but these efforts could at least partially mitigate the loss of foreign talent stemming from a restricted immigration pipeline.
Furthermore, the United States must adapt to the evolving demand for STEM graduates. Rapid technological advancements create dynamic job markets, often resulting in surpluses and shortages in various fields. Policymakers must ensure that educational and training programs align with industry needs, enabling graduates to thrive in a competitive landscape. Encouraging collaboration among government, industry, and academic institutions can create opportunities for middle-skill STEM roles, which do not always require advanced degrees but are critical for workforce development. Community colleges, apprenticeship programs, and “lifelong development” strategies (for example, veterans’ programs) can feed US talent into the high-tech sector.
Second, implement policies that facilitate the active recruitment and long-term retention of foreign-born talent once in the United States. The Optional Practical Training (OPT) program and H-1B visa pathways have been significant in allowing international graduates to transition into the US workforce. Expanding these programs can help ensure that the United States retains the skilled professionals it has invested in through education. Being able to choose the best and the brightest from a pool of twenty-seven million Chinese and about thirty-three million Indians enrolled in STEM-focused undergraduate studies results in a talent pool that the United States cannot afford to reject.
Third, encourage and support education in countries with a growing working-age population. Africa seem to be the only continent where that will be the case beyond the 2050s. Whether African states can educate and select for the US STEM pipeline remains to be seen. Nevertheless, at that point China, India, and Europe will be competing sharply to recruit top talent because of their own needs, and their universities and industries will be developed well enough to absorb them.
Fourth, continue international science collaborations to accelerate scientific advances in open science. The benefits as well as the risks of open science collaboration have been discussed in many papers, and the issues of technology transfer from American universities and industry are prominent in the news. But when carefully managed, even collaborations with declared adversaries can serve the same beneficial purposes: understanding the state of progress, interacting with the brightest minds, and boosting scientific advancements.
The STEM talent pipeline is a vital asset that requires thoughtful policy interventions to ensure its resilience. The future of the US economy and its position on the global stage depend on a collective ability to navigate these challenges effectively and strategically, developing the best talent in an increasingly technology-driven world. “Onshoring” substantially more talent with the right caliber will be possible only over a decade and with diligent attention. Ignoring the large talent pool beyond US borders will never be desirable.
The authors wish to thank their Hoover Institution colleagues—Dr. Macke Raymond (distinguished research fellow), Eric Hanushek (senior fellow), and Paola Sapienza (senior fellow)—for their advice and many helpful discussions.