A quarter of US-trained scientists eventually leave

Scale and nature of “leaving”

• “Leaving” means working outside the US within 15 years of a US STEM PhD; commenters note many scientists move back and forth, so “leave” can be temporary.
• Much of the 25% are foreign nationals on non‑immigrant student visas; they are not originally “meant” to stay unless they clear immigration hurdles.
• Some ask how this compares to inflows; others quote the paper that foreign nationals are ~50% of trainees, but only 10% leave within 5 years and 25% within 15, implying a strong net gain for the US.

Is 25% a problem or a feature?

• Several see 1/4 within 15 years as modest, especially given US dominance in tech over the data period.
• Others say it depends which 25%—losing mid‑career experts with 10–15 years’ experience is costlier than losing new grads.
• A strong view is that this is a feature: education is a major US “export,” and some brain circulation is good for US influence and global science.

Benefits to the US even when people leave

• Commenters point to the paper’s key metric: after migration, US patent citations to these scientists’ work fall from ~70% to ~50%, but remain five times higher than citations from their destination country and equal to all other countries combined.
• This is taken as evidence US industry and tech ecosystems continue to benefit disproportionately from the science of US‑trained PhDs abroad.
• Others highlight softer gains: cultural ties, goodwill, networks, and collaborative “cross‑pollination.”

Immigration policy and political climate

• Many blame complex, slow, and opaque processes (H‑1B, PERM, green cards) for pushing people out, including world‑class experts.
• There’s debate over recent US administrations: some see a sharp rise in xenophobia and anti‑immigrant policies (especially toward Chinese students) and predict a coming discontinuity; others note the paper’s claim that aggregate leave rates have been stable for decades and warn against overreading short‑term politics.

Academic labor market and exploitation

• Multiple comments stress structural overproduction of PhDs versus faculty slots: only a small fraction can remain in academia, so “most must leave” regardless of country.
• PhD education is often funded, but via underpaid TA/RA labor; many describe this as a form of exploitation, especially for foreign students who accept poor conditions to pursue a path to immigration.
• Proposals include fewer PhD students and more permanent research staff roles, or independent research organizations outside universities.

China and strategic concerns

• Some worry about “exporting” US‑trained talent to a geopolitical rival and mention Chinese programs that incentivize return.
• Others counter that many of these scientists were Chinese undergrads to begin with, that China’s own opportunities and salaries have grown, and that the US could retain more by funding science better and easing immigration.

Universities, money, and generational shifts

• Several point out that STEM PhDs at US research universities usually don’t pay cash tuition; their “payment” is teaching and research labor funded by grants.
• High‑paying foreign enrollment is seen as more of an undergraduate phenomenon, partly driven by cuts in public funding and administrative bloat.
• Some report younger generations (including US‑born scientists) now actively considering leaving the US due to politics, hostility to immigrants, and perceived decline in living standards, in contrast to parents who still see US education as the pinnacle.