The New Golden Age of Science
A bi-weekly newsletter where I share news, opinion, grant opportunities, success tools and tips on personal growth.
The New Golden Age: solid diagnosis, interesting ideas and very little biomedicine
On 21 July, OSTP Director Michael Kratsios published Science: A New Golden Age, a report to the President, alongside an annexed Fiscal Year 2028 R&D Priorities Memorandum co-signed with OMB Director Russell Vought. The report itself carries no legal force. The memorandum does: federal R&D agencies must file implementation plans within 90 days, roughly mid-October, and fold the guidance into their FY2028 budget submissions.
The diagnosis & the proposed remedies
There is much to be applauded in the Golden Age report. After all it is hard to argue with its assessment that "the scientific machine is getting bogged down" (although it is hard to hear this from people who have been gleefully throwing sand into the gears of the federal funding agencies that power it). Chapter II makes a simple argument: federal science funding suffers less from the size of the pot than from the fact that nobody decides how to divide it. Agencies hand out roughly $200 billion a year in R&D, and the report says none of them runs a systematic process for asking where those dollars would do the most good. Last year's portfolio sets this year's.
From there the chapter argues that the container shapes the science. A PI-led university lab, a corporate lab and a national lab each do some kinds of work well and never attempt others. The report calls what falls through the middle "mid-scale science". This is work that needs tens of millions of dollars, a team of ten to a hundred, and five years. An R01 cannot hold it. Venture capital will not fund it, because it produces open datasets and platforms rather than products.
The chapter levels a parallel charge at peer review: panels separate good science from bad reliably enough, but they cannot separate the exceptional from the merely good, and when only one proposal in ten gets funded, they fund the least divisive one in the pile.
The report then proposes three kinds of remedy.
New performers: focused research organizations, which hire career scientists and engineers, chase one technical bottleneck, publish what they build and then dissolve — plus NSF's X-Labs, which the report describes as the first federal program designed to fund research organizations outside universities.
New ways to choose: golden tickets, which let a single reviewer fund a proposal the panel won't back; funding people rather than projects, on the HHMI and NIH Pioneer model; grants decided in days instead of months; prizes and advance market commitments that pay for results; and regranting, which hands allocation to people close enough to the work to spot it before consensus forms.
And new capacity to keep doing this: a metascience unit in every science agency, reporting to the director, holding its own budget so it can pay program offices to take part in experiments on the agency's own grantmaking. The report names the UK's Metascience Unit, set up in 2024, as the working model.
One comment. The portfolio argument is a good one, and anyone who has sat on a study section will recognize the panel described as a liability shield and that they can exhibit some of the worst aspects of groupthink But the chapter opens by promising accountability and then spends its remaining pages on mechanisms that reduce it. Golden tickets, 48-hour decisions and empowered program managers all trade procedural checks for speed and appetite for risk. That may well be the right trade. It is not the same argument. And let us not forget that the way this Administration creates change is often destructive if not vindictive - it is hard to trust the authors of these documents with the future of American science. Worth noting too that when the chapter needs an example of what the current system can't do, it reaches for the R01, which has its flaws, for sure, but has also been a powerhouse of discovery.
What is not on that list
Reading the four pillars on which the whole concept rests (Revitalize America's Science and Technology Enterprise; Secure U.S. Dominance in Critical and Emerging Technologies; Launch a New Golden Age Powered by AI for Science; and Ensure That Science and Technology Better the Lives of All Americans) and the named missions (AI for science, Quantum, Fusion, the Moon, Semiconductors) I can't help notice that not one of them is biomedical.
Biomedicine (or actually biotechnology) appears in these documents mainly as a domain to which AI will be applied rather than as a national mission in its own right. When biological sciences are discussed the push is towards basic science because"these fields underpin biotechnology and biomanufacturing, neurosciences and brain-machine interfaces, and human health and therapeutics." [page 4 of the memorandum].There is little acknowledgement of the myriad challenges that make bringing a discovery to the clinic so hard.
In the Golden Age report the summary includes the sentence"Despite massive increases in biomedical funding since the 1990s, the rate of significant breakthroughs appears to have slowed and drug approvals have flatlined."[page xi, Golden Age report].
That is a meaningful demotion for the enterprise most of us work in, and it sits alongside FY2028 priorities that tilt toward the physical sciences and engineering. If, like me, you are trying to work out what all this means for NIH, that absence is the most informative thing in the report in my reading of it.
The bottom line
The report is a vision. Your agency's implementation plan, due around mid-October, is the document that will tell you whether any of this reaches your research program. That is the one to ask your research office about.
A federal judge limits the "changed priorities" grant terminations
On Friday 17 July, U.S. District Judge Indira Talwani in Boston ruled that the administration cannot rely on an OMB regulation to terminate grants on the grounds that they no longer align with the President's priorities. She found that the termination clause does not permit agencies to end grants based on program goals and agency priorities identified after the grants were awarded (U.S. News & World Report).
The clause at issue — allowing termination where an award "no longer effectuates the program goals or agency priorities" — was added to OMB regulations in 2020. The suit was brought by 23 states and the District of Columbia, led by Massachusetts, New York and New Jersey. Talwani granted summary judgment for the plaintiffs and denied the government's motion to dismiss.
Two limits are worth noting. The order applies only to the 23 states that sued, and the lawsuit did not seek to restore grants already terminated. Inside Higher Edalso notes the ruling could be rendered moot by OMB's proposed rule, whose comment period closed on 13 July and which is scheduled to take effect 1 October.
NIH consolidates what F, T and K applicants need to know about review
On 15 July, NIH Extramural Nexus published a summary of how fellowship, training and career development applications are now referred and reviewed. Most of the underlying changes are not new; the value is having them in one place.
NIH has centralized peer review for all grants, cooperative agreements and R&D contracts within the Center for Scientific Review, including T, K and F programs — so your application goes to a CSR-managed study section even when it aligns closely with a specific Institute or Center. CSR has established new recurring Special Emphasis Panels for these awards, grouped by scientific topic and activity code, with panel descriptions posted publicly.
Fellowship review criteria are candidate preparedness and potential, research training plan, and commitment to the candidate — placing weight on trajectory and training environment, and eliminating the requirement to submit grades. For T32s, mentor training expectations are more clearly defined in the parent NOFO, and institutional training data tables have been updated to reduce burden and focus on trainee outcomes.
NIH frames all of this under its Unified Funding Strategy. If you are advising a fellowship applicant this cycle, the posted SEP descriptions are the practical takeaway — you can now see where an application is likely to land.
