White House’s open-access research directive scrambles long-entrenched models, raising key questions

In August, the White House’s Office of Science and Technology Policy released a memo directing all federal agencies to form plans to make all federally funded research publications and data publicly available without embargo by the end of 2025.

Most people who heard the news likely envisioned making a cup of coffee on Jan. 1, 2026, opening up JAMA’s or Nature’s website, and being able to read any article for free.

But that’s almost certainly not going to happen.

The OSTP memo specifies only that new embargo-free manuscripts must be publicly available “in agency-designated repositories” like PubMed Central, not on journal websites. Would-be readers are still going to have to trace new papers to the repository of the agency that funded the work in order to read them.

But key issues around the OSTP guidance remain unresolved. To what extent will journals shift their business models as a result? And what approach will they take to open access? Such issues, experts say, are weighty, with implications not only for scholarly publishing, one of the largest-margin industries in the world, but also for science itself.

Many people who don’t have an affiliation with a research institution — and thus don’t have institutional access to scholarly journal subscriptions — struggle with getting access to the latest paywalled research: nurses and doctors at health care facilities without research centers, people navigating their family members’ or their own diagnoses, or those who are simply curious about the research their tax dollars have funded.

Currently, federally funded research is made available through public repositories but is subject to a 12-month embargo.

“For people who were using these public access repositories — and there are a lot of people doing that already — that embargo meant that they were second-class citizens who were living in the world of 12 months ago, as far as science was concerned,” said Ana Enriquez, scholarly communications outreach librarian and interim head of the department of scholarly communications and copyright at the Pennsylvania State University Libraries.

The situation used to be even more restrictive. But over the past few decades, and increasingly over the last 15 or so years, readers and taxpayers have been gaining greater access to federally funded research. In 2007, a spending bill directed the National Institutes of Health to require all research publications funded by the agency to be publicly available at a maximum of 12 months after publication. Under a 2013 directive from the OSTP, the rest of the federal agencies with research budgets larger than $100 million followed suit in 2016.

While open-access advocates lauded these steps, these efforts didn’t fundamentally change the stability of the scholarly publishing enterprise. The new move by OSTP, experts agree, could have more far-reaching effects.

It’s also not without precedent.

In 2018, a group of largely European funders announced Plan S, an open-access initiative that stipulated that research funded through its supporting organizations would have to be published in an open-access journal or platform or be immediately available in an open-access repository without an embargo, among other stipulations. In addition to the European funding agencies, private funders including the Wellcome Trust, the Howard Hughes Medical Institute, and the Bill & Melinda Gates Foundation signed on.

The Plan S policies compelled many publishers to create new open-access journals. However, some publishers merely allowed Plan S authors to have exceptions to their normal rights licensing contracts, which both kept Plan S authors in compliance with their funders and maintained the status quo for the journal’s other authors.

This patchwork approach to funder compliance may have worked up to this point, but the U.S. expansion of public-access policies will undoubtedly affect more research. Whereas only 20% of the authors published in the journal Science were subject to Plan S policies from 2015-2022, for example, 44% of authors over the same time period were funded by U.S. agencies, according to the American Association for the Advancement of Science, which publishes the journal.

The result is that journals that have traditionally relied on charging for access to new research will have to find new ways to make up lost revenue. After all, without an embargo, why should institutions continue to subscribe to a journal, especially as “many libraries — probably safe to say all libraries — are under extreme budget pressure,” according to Enriquez.

“I opened a special bottle when I read the [OSTP] press release,” said Robert-Jan Smits, president of Eindhoven University of Technology, who founded Plan S when he was working as the European Commission’s open-access envoy. “A bottle of good French wine!”

Smits pointed out that the taxpayer currently pays for research three times: when the researcher gets a research grant, when the publisher asks a different researcher to peer-review the work for free, and when institutional libraries pay for subscriptions for the journal. The OSTP’s economic analysis, published the same day as the memo, identifies two additional times the taxpayer pays: if the researcher is assessed a fee for publishing their research, and if taxpayers themselves want to access the research.

If the scholarly publishing tides really will turn upon this new guidance, the research world now has an opportunity to figure out whether the Jenga tower of research should simply be reconstructed with different groups of people at the top and bottom as financial winners and losers, or if it’s possible to build a more equitable structure.

Popular alternative models for making research free to access often revolve around preprints, the non-peer-reviewed manuscripts researchers can post to subject-specific repositories like bioRxiv, medRxiv, physics’ original arXiv, and others. Over the last 30 years, fields like physics and math have adopted preprints as part of their publishing workflow, which both rapidly disseminates their results and functions as a stop for community feedback on the way to traditional peer review.

When people ask about making peer-reviewed research free, Richard Sever, co-founder of the preprint servers medRxiv/bioRxiv, says he always asks, “What do you mean by ‘peer-reviewed’?”

“Five years ago, 10 years ago, [people] weren’t really thinking about that,” said Sever. “They basically went, ‘Well, peer review — the stuff that’s in journals.’ But we all know that that’s very varying. The quality of peer review in some places is not as good as peer review in others.”

Sudip Parikh, CEO of AAAS, said that he’s not “a stick in the mud” about what a research journal should be, but also that the question “if you could design the system from scratch, what would you do” is not a “terribly valuable” question because we can’t design the system from scratch.

“We have a $45 billion funding agency for biomedical research sciences, we have a $9 billion funding agency in the National Science Foundation, we have the $99 billion Horizon Europe … there is no starting from scratch,” Parikh said in an interview. “There is ‘what do we do that both evolves and revolutionizes the current system?’ What that requires is a thoughtful process about how do you maintain the things that we care about in this system that have worked? How do we change the things that haven’t worked? I think that’s the better question.”

In an editorial in early September, Parikh and other senior executives at the AAAS warned that one of the most cited open-access publishing models, in which authors pay a fee called an article processing charge that allows their research to be published without a paywall, would create inequities.

Parikh described this approach, often called the “gold” open-access model, as “just plain, outright bad for the enterprise” because it weaves structural inequity into the tapestry of the scientific enterprise.

“If you’re at the top of those heaps, this is easy,” said Parikh. “You can afford that APC [article processing charge]. And so you’re going around going, ‘Well, this is a great system. The system works and everybody can read it now. I’ve made it fair and equitable,’ except for the fact that you’ve suffocated early-career scientists that are where all the diversity of thought, experience, geography sits.”

For more selective journals that publish fewer papers, APCs can vary from $4,500 to publish one article in AAAS’s gold-OA journal Science Advances, to $11,390 to publish gold OA in a hybrid journal like Nature, which has a subscription fee but allows authors to choose to publish their articles open-access for an APC fee.

It’s easy to imagine the kinds of people who wouldn’t have easy access to $10,000 or more to publish their research: early career researchers, researchers from less-well-funded disciplines, researchers from lower-income countries. Carrie Webster, vice president of open access at Springer Nature, the publisher of the Nature family of journals, said there are waivers available for anyone who can demonstrate that they don’t have the funds to pay the APC, though she acknowledged that having to request a waiver in the first place isn’t a perfect solution.

Sever said that, even with waivers, functionally, different institutions and disciplines would functionally end up subsidizing others under the gold OA model. He cited the example of molecular biology, a discipline in which researchers are well-funded and “can easily find $10,000 to pay an APC in Nature.” But in math or sociology or literature, “do you then have a situation where you say, ‘Oh, well, all the molecular biology APCs cost a little bit more because we’re having to give so many waivers to all the people who work in mathematics’?”

In addition to contributing to inequity, Sever also pointed out that instead of incentivizing publishers to be selective about which papers they accept so they can keep costs low, an APC-driven publishing economy can incentivize “low-touch” open-access journals that process as many articles as possible since they are receiving a fee for each one.

Webster pushed back against the idea that “pay-to-play” vanity or so-called predatory publishing could gain popularity. “If you publish rubbish, the community will see it. You will be named and shamed and the papers could be discredited and withdrawn,” she said. “You could only do it for a very short period of time, because then everyone would know that this journal is not a credible place. It would get shut down or just get blacklisted, so it wouldn’t be possible.”

In an interview, Jedidah Isler, OSTP principal assistant director for science and society, emphasized that the agency’s guidance does not prescribe a publishing model. Rather than mandate open access publishing where papers are freely available from the publisher, the public access policy outlined in the memo and past federal guidance can be fulfilled through something that looks closer to what is known as the “green” OA model, wherein researchers are allowed to post the peer-reviewed, accepted manuscript to a repository for public access (a “post-print” or “author accepted manuscript”). In this model, the open repository version satisfies any public access requirements without levying any additional fees.

While publications like Science and Nature have editorial and journalistic content that they could potentially count on to drive subscriptions in a green OA model, that’s not the case for the majority of journals. Because only the accepted version of the manuscript is allowed to go onto repositories under a green OA route, not the fully typeset publisher’s version, researchers’ desire for the traditional, fully published version of a manuscript may drive subscriptions for a time. The publisher’s version of a paper is also simply much easier to access, and the barrier to finding the freely available version might prevent people with institutional subscriptions from regularly going to the publicly accessible versions of papers.

But Enriquez, of the Penn State library, said that long term, the internet will be easier to search. Tools like Unpaywall and Open Access Button are already making it easier to find the legal open-access versions of articles. Journals won’t be able to rely on the fact that their paywalled version of a paper is easiest to access for long, especially when libraries are under tremendous pressure to reduce their budgets.

Besides the gold and green OA models, other approaches include “transformational” agreements, or “read and publish” agreements, whereby a library agrees to pay a sum to a publishers in exchange for rights to read their content and have their institution’s researchers publish open-access without an extra APC. But Enriquez said that transformational agreements are supposed to be transitional and eventually unnecessary as publishing shifts to fully open access, though that transition has been extremely slow.

Many sources said it’s too soon to tell what’s going to happen without the official guidance from U.S. research agencies. The issues that need to be addressed are myriad. Will the guidance be aligned across agencies, or perhaps consistent internationally with other open-access movements like Plan S? Since the memo also encompasses research data, how will “research data” be defined (the final, cleaned spreadsheets, or the raw data exported from the microscope)?

Though the OSTP memo is agnostic as to publishers’ business models, will Congress or funding agencies allot extra financial support for author APCs, larger publishing agreements, or for supporting repository infrastructure?

The NIH has a history of leading U.S. research policies and has noted in the past a commitment to “shift[ing] the culture of research” around open data. Along with reassurances that the the 22 federal agencies in the OSTP’s Subcommittee on Open Science are having conversations around what policies federal agencies should set, there are indications that the NIH and other agencies are trying to produce policies that are functional for science in the long term, not just ones that fulfill current requirements.

“The future of open science, the seamless integration of data and information flow is a transformative future that we’re trying to get to at the end of the day,” said Lyric Jorgenson, the acting associate director for science policy at the National Institutes of Health. “We want to make sure all these steps that we’re taking aren’t fulfilling a requirement as much as they are really taking actual steps towards this new future that we’re trying to envision together.”