America’s Self-Inflicted Brain Drain –

Immigrants, who account for 14% of the U.S. population, are responsible for 36% of aggregate innovation¹ and are 80% more likely than native-born Americans to start a business.² Nearly half of all Fortune 500 companies were founded by immigrants or their children, and 59% of Artificial Intelligence (AI) PhD graduates working in U.S. industry are international students. History offers a stark warning about what happens when these flows are disrupted: immigration quotas in the 1920s caused a 68% relative decline in U.S. patenting.³ The United States now risks repeating that history as it dismantles, piece by piece, the talent pipeline that built its competitive advantage.

Consider the most recent visible blow to that pipeline. In September 2025, President Trump imposed a $100,000 fee on new H-1B visa petitions: the primary mechanism through which U.S. firms hire skilled foreign workers. The administration framed it as a measure to protect American workers and ensure the program attracts “the best of the best.” The fee adds an enormous new barrier on top of an already broken system; annual caps of 85,000 visas have been oversubscribed every year since 2004, with demand typically running at two to three times the supply. A large body of research tells us the impact of making H-1B visas harder to obtain. When U.S. multinational companies faced tighter H-1B restrictions in the past, they offshored more, hiring more at their foreign affiliates. On average, multinationals created roughly 0.4 foreign affiliate jobs for every unfilled H-1B position in the U.S.; the most internationally oriented firms replaced nearly one-for-one.⁴ Restrictive immigration policy does not keep jobs in America. It pushes them out.

Small startups are hit hardest. Unlike multinationals, they cannot easily offshore. Evidence from H-1B lotteries between 2014 and 2018 shows that startups that lose the H-1B lottery are less likely to achieve a successful exit via initial public offering or acquisition and less likely to receive additional venture capital funding.⁵ The $100,000 fee will accelerate these dynamics. For large technology companies, the fee may be an irritation absorbed into the cost of doing business even as they accelerate the offshoring of skilled work and talent. For smaller firms, for universities, and for hospitals and research institutions that rely on H-1B workers, the fee is potentially prohibitive. And it sends a signal, to firms and to workers alike, that the United States is raising the drawbridge.

But the damage extends far beyond the H-1B program itself. The pathway from international student to skilled immigrant worker in the United States is long and institutionally dependent: students arrive on F-1 visas, earn a degree, gain work experience through Optional Practical Training (OPT), and eventually transition into employer-sponsored work visas and, ultimately, permanent residency. My recent research with Caroline Fry on AI scientists emphasizes this: many PhD-holding immigrants at top U.S. AI firms arrived not through direct hiring abroad but as graduate students who transitioned into industry through university networks. The university is where the pipeline begins, and it is the part of the system now under the greatest strain.

New international student enrollment at U.S. universities fell 17% in the fall of 2025: the largest single-year decline outside the pandemic on record. International graduate student enrollment dropped 12%. These are not small fluctuations. They reflect visa interview suspensions, travel restrictions affecting nationals of dozens of countries and a political climate that signals hostility rather than welcome. The administration has signaled its intent to curtail or eliminate OPT, the bridge program that allows graduates to transition into the workforce.

My work with Robert Flynn, Raviv Murciano-Goroff and Jiusi Xiao reveals just how damaging perceived hostility to migrants can be, even without formal policy changes. We find that ethnically Chinese students became 15% less likely to enter U.S. doctoral programs after U.S.-China tensions escalated. Crucially, the decline is concentrated among the most talented, i.e. those with the most external options. And they do go elsewhere: other countries, especially Anglophone ones, absorb this redirected talent. These effects cannot be explained by policy changes alone; they reflect a climate of perceived hostility that deters talent before any visa form is ever filed.

If the United States is driving talent away, the obvious question is who stands to gain. For European policymakers, this should be both an opportunity and a cautionary tale. The opportunity is real: applications from U.S.-based researchers to the European Research Council’s early-career grants have nearly tripled in recent years, as reported by STAT News. The European Commission has launched a “Choose Europe for Science” program targeting top international researchers. France has launched its own platform and introduced a bill to create a formal “scientific refugee” status. There are early signs of Europe becoming more attractive: Nobel laureates Esther Duflo and Abhijit Banerjee have announced they would leave MIT for the University of Zurich, and Aix-Marseille University received nearly 300 applications from researchers seeking refugee status.

But Europe should be clear-eyed about the structural barriers that have historically prevented it from capitalizing on moments like this. The EU Blue Card scheme has seen modest uptake compared to Canada or Australia’s fast-track systems. Many international students educated in Europe leave after graduation, as documented by the ICMPD, citing bureaucratic barriers and limited career prospects. And compensation gaps remain large: U.S.-based AI positions typically pay 30% to 70% more than equivalent roles in Europe. The lesson from the United States is not simply that restrictive immigration policies push talent away. It is that talent pipelines are institutional achievements that take decades to build and can be destroyed with remarkable speed. The U.S. system, for all its flaws – the H-1B lottery, the low employment visa and green card caps, the employer-tied visas that leave workers vulnerable –nonetheless created an effective mechanism for channeling global talent into productive firms and innovative ecosystems. It did so because firms, universities and government programs operated as interconnected intermediaries, each reinforcing the others. Dismantling any part of this system has cascading effects on the whole.

Europe has a historic window. But seizing it requires more than welcoming statements and one-off fellowship programs. It requires building coherent, continent-wide talent infrastructure: making the EU Blue Card a genuinely attractive and well-publicized pathway with harmonized standards across member states, streamlining mobility so that a non-EU researcher does not face a fragmented patchwork of national procedures to move within Europe, and investing in the compensation and research funding that makes European positions competitive, not just available.

The global geography of innovation is being reshaped in real time. My research suggests that when the United States restricts access to skilled immigrants, innovative activity and human capital shift abroad, and with them, the knowledge spillovers and ecosystem effects that drive long-term economic growth. Countries that position themselves to absorb this redirected talent will reap outsized benefits. Those that do not will miss the opportunity. The United States is conducting a remarkable experiment: testing whether its leading innovation economy can maintain its position while systematically alienating the global talent on which that position depends. The question for Europe is whether it will treat this as someone else’s problem or as an opportunity.