When an NIH funded epidemiologist tried to comply with the 2024 NIH Public Access Policy, Nature Medicine charged her $12,850 for a paper that had cost $0 to publish a year earlier. The fee is the unfilled half of the policy.
The 2024 NIH Public Access Policy was supposed to settle a long-running fight over who gets to read taxpayer-funded science. Starting July 1, 2025, any peer-reviewed paper arising from NIH funding was supposed to be made freely available to the public immediately on publication. Elizabeth Selvin, a clinical research professor at Johns Hopkins Bloomberg School of Public Health, did everything the policy asked of her. She reports, in a June 11, 2026 STAT First Opinion column, what that compliance cost her: $12,850.
That was the 2026 figure. In June 2025, Selvin had published a different NIH-funded randomized clinical trial in the same journal, Nature Medicine, and reports paying nothing. The swing, from $0 to $12,850 in twelve months for the same journal, was not driven by new services or new copyediting. It reflects the structure of the open-access market that the NIH policy never managed to override.
Selvin directs the Welch Center for Epidemiology at Johns Hopkins Bloomberg School of Public Health and serves as a deputy editor of Diabetes Care. Both studies were funded by a mix of NIH grants and philanthropy, framed by her as taxpayer-funded research. The 2024 NIH Public Access Policy, which took effect on July 1, 2025, is the mechanism she identifies. Its stated intent is free, immediate public access. The way it fell short, in her account, is that for-profit publishers including Springer Nature, Elsevier, and Wiley — which she notes together publish more than 7,500 journals — retain the right to require full copyright transfer from the author and to embargo the author-accepted manuscript in PubMed Central for six to twelve months after publication.
The result is a structure in which NIH compliance is technically possible without ever making a paper free. The author can deposit the manuscript in PubMed Central on a delayed schedule, satisfy the letter of the policy, and the publisher can still sell the published version of record at subscription rates. The article processing charge (APC), like the $12,850 Selvin reports paying, is the price of immediate open access. The embargo is the price of compliance on the cheap. The choice between them is not made by the funder or by the public. It is set by the publisher's contract.
Selvin's experience is a single data point from a researcher with a direct financial stake in the answer. Her figures are testimony, not audited accounts, and the framing of the 2024 policy's failure reflects one researcher's view, not consensus. The structural point, though, survives that discount. The 2024 NIH Public Access Policy was a Congressional mandate backed by decades of advocacy from patients, librarians, and taxpayer-funded researchers who objected to paying twice, once in taxes and once in subscriptions, for the same work. The mandate landed without an enforcement mechanism aimed at the publishers most able to resist it. Springer Nature, Elsevier, and Wiley, the three publishers Selvin names, control a large fraction of the high-impact journals in which NIH-funded researchers want to publish. None of them is obligated by the policy to alter their standard contract.
Defenders of the existing model argue that subscription revenue funds the peer review, editing, and distribution that make journals valuable, and that APCs are simply the price of a different distribution choice. The 2024 policy does not, in its text, endorse that view or reject it. It requires deposit on a specific timeline. It does not require a specific contract with the publisher. The cost of immediate open access therefore behaves like a hidden levy on the NIH grant: a charge the funder does not budget, the institution does not negotiate, and the public never sees on the front end.
What the 2024 policy did not do, and what the Selvin comparison makes visible, is constrain the price of compliance. A mandate without a price ceiling is a permission slip. In Selvin's account, the Nature Medicine APC was set by the publisher and presented as the only path to immediate open access, not as a fee to be negotiated. The next round of policy debate, whether at NIH, in Congress, or in the negotiations around the next reauthorization of federal research funding, will have to decide whether the public is buying access to its own science or renting it back from the publishers who carry it.