Inside Trump's Science Agenda: Anti-Science Claims, Fauci's Damage, DEI & China w/ Michael Kratsios

Summary

• Michael Kratsios, the OSTP director, flatly rejects the “anti-science administration” charge — despite a Nature poll showing 86% of scientist readers backed Harris vs. 6% for Trump — arguing the scientific community “lost its way” over 15-20 years of politicization that “crescendoed and peaked in COVID.” His counter-thesis is the new report “Science: A New Golden Age”: the question isn’t whether the ~$200B annual federal S&T budget goes up, but whether it “is actually driving the best breakthroughs.”
• Asked what was behind the $3B in unspent funds on active grants that the host says were disrupted, frozen, or terminated in 2025, Kratsios does not itemize those grants. Instead, he frames the issue as resource allocation, citing Ted Cruz’s Senate committee analysis that roughly 25% of Biden-era NSF grants — ~$2B a year, $8B over four years — went to “DEI-related quote-unquote science,” and arguing that “spending more money on the wrong types of things is not the right policy action.”
• On climate, the administration’s line via Secretary Wright: yes, the climate is changing and human-burned fossil fuels contributed CO2, but “what the data does not support is that it is a climate emergency.” Kratsios points to RCP 8.5, the extreme scenario he says was pulled a few months ago because it couldn’t be sustained “with any scientific integrity” — yet it drove media focus and some funding.
• Kratsios says scientific stagnation is real: the host cites NIH’s budget rising from $14B (1998) to $47B (2024) while outcomes per dollar have fallen ~80-fold since 1950, halving every nine years — a trend people call Eroom’s law. The proposed fix is “metascience” — new units at NSF and NIH will run experiments on funding itself: Denmark-tested “golden tickets” letting reviewers unilaterally back bold grants, variable durations instead of the 18-month default, and 2,600 portable GRFP fellowships that fund the scientist rather than lock them to a project.
• The government’s proposed lane is big bets and precompetitive discovery science the private sector won’t fund: American boots on the moon in ’28, a nuclear reactor in space by ’28, moon-base elements by 2030, a “scientifically relevant quantum computer by 2028,” fusion by 2035, and the flagship Genesis Mission to “essentially double the scientific output of the United States” by applying AI to science. With hundreds of billions of private capital in AI, “the government’s not trying to compete in that space.”
• The host cites China’s spend rising from $33B (2000) to $670B (2021) — 19x growth vs. America’s 3x — and estimates that China now publishes ~50% more papers in nearly every domain except some life sciences, while noting the statistic is hard to verify. He also cites estimates of 2x-10x greater output per dollar. But Kratsios argues centralization fails: China’s single-agency model has chased EUV lithography since the 2019 export controls with “no breakthrough,” while America’s fragmented agencies are “a feature, not a bug.”
• The talent pipeline is the loudest alarm: U.S.-born computer-science PhDs inverted from ~70% American 30 years ago to ~70% foreign today, the median NIH intramural researcher is 71 years old, and NSF programs for gifted STEM students were stopped “for some reason.” The host speculates that this was part of a broader DEI effort; Kratsios calls the decision discouraging and says “that’s what’s most tragic.”
• On COVID, Kratsios is categorical: Fauci “did more disservice to science than anyone in modern history,” and rebuilding public trust “is going to take quite a while to dig ourselves out.” Implementation is already moving — an R&D priorities memo co-signed with OMB’s Russ Vought requires the top five research agencies (>$3B R&D each) to report implementation plans within 90 days.

Deep dive

1. “Not anti-science” — the politicization diagnosis

• Kratsios runs OSTP, the White House office coordinating a fragmented enterprise — NSF for basic research, DOE’s national labs, DARPA at the Department of War — after serving as U.S. CTO in the last administration. Asked point-blank whether the administration is anti-science, he points to the new report “Science: A New Golden Age”: “We really want the U.S. to be the home for the next great scientific discoveries.”
• His answer to the Nature poll (86% of its scientist readers for Harris, 6% for Trump): the scientific community has been “deeply politicized” for 15-20 years, peaking in COVID, “where suddenly there was a certain individual who, if you didn’t agree with what he said, then suddenly you were anti-science… that is the most anti-science conclusion you could ever make.”
• The lobbyist critique frames the whole episode: the median science lobbyist “is fixated” on the R&D budget — if the number doesn’t go up, science wasn’t supported. Kratsios’s harder question: is $200B a year in federal S&T funding actually producing breakthroughs, or could it flow “to different organizations, to different scientists, through different time horizons”?

2. Kratsios’s rationale for grant and climate changes

• On the host’s claim that 2025 saw $3B in unspent active grants disrupted, frozen, or terminated, Kratsios first separates proposed budgets from Congress’s appropriations — Congress “continued to fund science” — but does not itemize those $3B. He instead gives his best example: Cruz’s committee found ~25% of Biden-era NSF grants ($2B/yr × 4 years = $8B) went to DEI-related work after applicants were effectively told a DEI factor was “the only way we’re going to give you money.” “That isn’t science.”
• The climate position, channeled through Secretary Wright: the climate is changing, humans burning fossil fuels contributed CO2 — but “the thing that is not true and what the data does not support is that it is a climate emergency.” Kratsios says RCP 8.5, the most extreme scenario in national climate assessments and IPCC work, was pulled after they “could not with any scientific integrity substantiate continuing to have this scenario play out” — yet it was the scenario media reports fixated on for 20 years, “a real disservice to science.”
• The host’s economic overlay: China’s carbon output eclipses the rest of the world, so ±50% in the U.S. “doesn’t really move the needle,” and energy restrictions hit the poor hardest. Kratsios’s supporting numbers: ~2 billion people cook with wood, dung, or charcoal, causing ~3 million deaths a year per the UN. The host argues the fix is going “forward with new technology,” not backward 30-40 years; Kratsios agrees that extreme policy reactions would hurt Americans without solving the problem.

3. Stagnation is real — the answer is experiments on funding itself

• The host’s data: NIH went $14B (1998) → $27B (2003) → $47B (2024), yet outcomes per dollar have fallen ~80-fold since 1950, halving every nine years. Kratsios broadly concedes stagnation and adds his favorite everyday example: “In the 1990s you could fly on a Concorde and now you can’t. We’re flying slower than we were 10, 20 years ago.”
• His causal diagnosis: no incentive to check homework. The Hill’s incentive is bigger budgets flowing to districts; on the government side there’s no incentive to reveal that current methods are not working well. The answer has been “let’s just keep doing the same thing but add more money” rather than asking whether different scientists, institutions, or time horizons would do better.
• The fix: metascience units at NSF and NIH running actual experiments on allocation. First test — “golden tickets,” initially tested in Denmark and elsewhere, letting each reviewer unilaterally select a grant without committee consensus, countering peer review’s tendency toward “strike zone” proposals and attracting better reviewers. Second — breaking the 18-month grant default with fast-track six-month grants that, if successful, can lead to longer three-to-five-year grants.
• The host’s VC framing — venture works on a power law, “you want nine out of 10 failures because that means you’re taking a lot of risk,” and you fund people rather than ideas (his example: Stewart Butterfield nearly returning $3M before pivoting to Slack, sold for $30B). Kratsios’s government analogue: ~2,600 GRFP fellowships that are fully portable, so universities compete for the scientist.

4. Big bets only government can make — and where it stays out

• The portfolio: American boots on the moon in ’28, a nuclear reactor in space by ’28, first moon-base elements by 2030, a new National Quantum Initiative executive order targeting “a scientifically relevant quantum computer by 2028” at DOE, and fusion by 2035.
• The flagship is the Genesis Mission: “double the scientific output of the United States by applying AI to our hardest scientific challenges.” Kratsios’s conviction is unhedged: “AI is going to be the biggest unlock to scientific discovery in the history of the world,” across materials, chemistry, math, and physics.
• The host’s pushback — why fund AI and quantum when hundreds of billions of private dollars are flowing in? Kratsios’s line: government’s primary focus is “basic, early-stage, precompetitive R&D” the private sector won’t touch; private quantum targets commercial applications, while government wants an instrument for scientific discovery — “the government’s not trying to compete in that space.”

5. Rewriting Vannevar Bush: fund the scientist, wherever they are

• The historical frame: Vannevar Bush’s 1945 Science, the Endless Frontier set the North Star — government funds basic research primarily at universities — when the federal government paid ~70% of U.S. R&D. That ratio has flipped to 70% private, and new institutions have sprung up, including HHMI, the Max Planck Institute, and focused research organizations. “Is that model still as relevant today as it was then? Our answer is it’s not, and we need a refresh.”
• Against the charge that university funding changes are politically motivated: “We should meet the best scientists wherever they are… if they’re on their own in their garage doing incredible work and can pass a merit-based review panel, we should fund them in their garage. We are pro-scientists and we don’t care where they are.”
• On consolidating science agencies, a change of frame worth keeping: after years reflecting on it, “I honestly think it’s a feature and it’s not a bug.” The extreme alternative is China’s top-down single-agency entity — which has been trying to crack EUV lithography since the 2019 export controls, with “no breakthrough,” despite nothing being more important to its economy.

6. The China scoreboard and playbook

• The host’s numbers: China’s science spend went from $33B in 2000 to $670B in 2021 — 19x vs. America’s 3x. He says that a decade ago the U.S. published roughly twice as many papers, while China now publishes ~50% more in nearly every domain except some life sciences, though he notes that estimate is hard to verify. He also cites estimates of 2x, 4x, or even 10x output per dollar given lower labor and supply-chain costs.
• Kratsios’s read of Beijing’s playbook: “copying our IP, creating cheaper alternatives… dumping them in the United States, getting our businesses to go out of business and then squeezing us” — plus threatened action on critical minerals, which demands supply-chain independence.
• The host notes that the Manhattan Project, Apollo, and the Human Genome Project were centrally planned wins. Kratsios’s response is a portfolio approach: big, government-marshalable projects alongside the free-market venture ecosystem, government science funding, and industry research. His grounds for confidence remain structural: the U.S. is still “the place where everyone wants to come work, live and study.”

7. The talent pipeline is the emergency

• The statistic Kratsios has tracked for years: U.S.-born computer-science PhD recipients went from ~70% American 30 years ago to inverted today — “seven out of 10 STEM-related PhD candidates in the United States are not American… anyone can look at that and say that just doesn’t look right.” Young post-PhD scientists are “the most underpaid people you could imagine for what they’re doing for our country.”
• The statistic that stuns both men, via NIH director Jay Bhattacharya: the median age of an NIH intramural researcher is 71. Kratsios: “Stay tuned there… a lot of interesting stuff [is] going to happen at NIH.”
• A disputed policy example: NSF once ran programs supporting gifted and talented students in American middle and high schools — “an actual decision was made by the government to no longer fund that,” Kratsios says, “for some reason.” The host speculates that this was part of a broader DEI effort and cites San Francisco dropping advanced algebra and the UC SAT fight, with Berkeley professors reporting remedial math needs and calling for the test’s return. Kratsios: “That’s what’s most tragic… we actively run programs which discourage them.”
• On immigration, Kratsios says the issue is “not in conflict” with America First, but sequence matters — “before we even start thinking about anyone else, we have to get our house in order” on American STEM participation, while those who want to stay should have legal pathways.

8. Fauci’s damage, the trust deficit, and what ships next

• The episode’s sharpest claim: Fauci “did more disservice to science than anyone in modern history” — masking students in schools while pediatric societies issued letters saying collective protest was okay even as people were not allowed to visit “your dying grandmother in a hospital.” “No regular median American can listen to that and then take science seriously… it’s going to take quite a while to dig ourselves out.”
• The host’s heuristic from that era: “if everyone thinks something, it’s probably wrong” — uniform alignment without empiricism was the tell, and questioners were branded anti-vaxxer or anti-science. His own example: publicly pushing back on RCP 8.5 three or four years ago would have made him a “Trump-er, anti-science MAGA guy”; Kratsios says the IPCC now says they can’t talk about the scenario anymore.
• Implementation is concrete: the same day the report dropped, Kratsios and OMB director Russ Vought wrote the annual R&D priorities memo putting basic fundamental research “first and foremost”; the top five research agencies (>$3B R&D each) owe implementation reports within 90 days, and NSF has already announced X-Labs funding for focused research organizations and four-year industry-partnered Ph.D.s. Congressional appropriations alignment comes next — “the Hill will always be a slog,” though he argues the reforms are nonpartisan and touch many states.
• On a possible 2028 change in administration: “I’m optimistic one that we’ll win in ’28” — but the plan is to “run really hard through the tape over the next two years and prove that they’re actually working,” so reversal becomes costly in political capital.