Public Money, Private Shelves: The Battle Over Who Owns America's Research Record

Every year, United States federal agencies disburse tens of billions of dollars in competitive research grants — through the National Institutes of Health, the National Science Foundation, the Department of Energy, and a constellation of smaller funders. That money produces roughly a quarter of the world's peer-reviewed scientific literature. Yet a citizen who pays federal taxes, funds the work, and appears in the acknowledgments of a clinical trial or climate model often cannot read the resulting paper without paying a publisher $30 to $50 per article or subscribing through an institution they do not have.

The question at the center of this tension — who owns a research record built with public money — has been asked in Washington for two decades. The answer has begun to change.

A History of Half-Measures

The modern fight traces back to the 1990s, when the internet made photocopy-based dissemination look suddenly obsolete. Publishers, led by major commercial houses, lobbied aggressively against early proposals to require free electronic deposit of NIH-funded papers, arguing that government-mandated sharing would destroy a private industry that had, in its telling, added essential value through editing, quality control, and archiving.

The compromise that emerged in 2008 — the NIH Public Access Policy — required grantees to deposit final accepted manuscripts in PubMed Central, but only after an embargo period, originally twelve months. It was a genuine step forward and a studied half-measure. Other agencies followed unevenly or not at all, leaving researchers and the public to navigate a patchwork of inconsistent rules.

The status quo has costs that compound quietly. A rural hospital without a journal subscription cannot easily consult the latest oncology literature. A high school teacher in Ohio building a lesson on climate science may hit a paywall within minutes. Even researchers at well-funded universities routinely lose hours negotiating access across publisher platforms, and interlibrary loan offices absorb the overflow.

The 2024 Inflection Point

In late 2024, the White House Office of Science and Technology Policy issued guidance requiring that, by the end of 2025, all federally funded peer-reviewed manuscripts be made publicly accessible immediately upon publication, with no embargo — a dramatic acceleration of the earlier framework. The policy applies across agencies rather than to NIH alone, and it extends to the underlying scientific data where appropriate.

Publishers responded with alarm. Industry representatives argue that subscription revenue funds the editorial infrastructure — copyediting, typesetting, digital preservation, plagiarism screening — that makes journals trustworthy, and that immediate free release undermines the business model without a replacement in place. Several publishers have since petitioned for reconsideration, framing the mandate as an uncompensated transfer of private value to the public domain.

Open-access advocates counter that the industry's margins — frequently above 30 percent for the largest publishers — tell a different story, and that the value of a scientific article is created overwhelmingly by the scientists, reviewers, and funders, not by the publisher.

What the Alternatives Cost

The most common rebuttal to the mandate is that open-access publishing merely moves the bill. Article processing charges at fully open-access journals routinely run $2,000 to $11,000 per paper, a model that risks substituting one inequity for another: instead of readers being priced out, authors at under-resourced institutions may be priced in.

But the emerging landscape offers more than two options. Diamond open-access models — no charge to readers or authors — are supported by university libraries, consortia, and philanthropic funders, though they remain a small fraction of output. Subscribe-to-open arrangements, in which libraries collectively underwrite a journal's continued free access, have kept several humanities and mathematics titles alive. Negotiated transformative agreements between library consortia and publishers, including several signed by the University of California system, bundle reading and publishing rights into a single contract — with mixed results on cost savings.

No single model has proven sufficient at national scale. What the federal mandate does is force the experiment to run everywhere at once.

What Researchers Should Watch

For the scientific community, the transition raises practical questions that policy guidance has not fully resolved. Compliance workflows vary widely across agencies. Researchers who have built careers on publication in prestigious subscription journals face incentives that have not yet caught up to the new rules. And the long-term stewardship of the public-access repositories — who maintains them, at what cost, and with what preservation guarantees — remains an open administrative question.

There is also a risk worth naming: public access to papers is not the same as public comprehension of science. Removing paywalls does not, by itself, repair the translation gap between technical literature and civic understanding. But it removes a barrier that never should have existed, and it does so for the population that funded the work.

The deeper principle at stake is straightforward. A nation that treats its research record as public infrastructure — as foundational as highways or the postal service — must decide whether that infrastructure is genuinely public. The mandate now in force says yes. Whether the publishing ecosystem, the research community, and the repositories can build a durable system around that answer will be determined in the next several years, one deposited manuscript at a time.