On May 21, 2026 the Government Accountability Office released a new report, Federal Research: Agencies Should Better Manage Anticipated Publishing Cost Increases Amid Shift to Public Access (GAO-26-107738). SPARC issued the following statement in response:
The federal government has one obligation in public access policy: ensure that taxpayers can access the results of the research they paid for. The government does not have an obligation to promote policies that ensure commercial publishers maintain a specific business model in the process. This report conflates public access with pay-to-publish, treating the current commercial publisher business model as an unchangeable feature of scholarly communication rather than what it is: one optional business arrangement among many alternatives.
The GAO describes a market dominated by a small number of large firms. The appropriate policy response to this broken market should not be accommodation but investment in alternatives that return control of the scholarly record to the research community and to the public, who are footing the bill.
The report states that there is currently “no practical way” for researchers to have their work peer-reviewed without paying a publication charge. This is simply not accurate. It mistakes a business arrangement for a structural necessity. Peer review has always been conducted for free by researchers, for researchers. Publishers coordinate it in some instances; they do not own it. Large commercial publishers want to control the peer review process and condition access to it by charging fees.
But publisher business decisions should not dictate federal policy, particularly when alternatives exist. Many scholarly societies and university presses support public access without engaging in pay-to-publish schemes. Additionally, overlay journals, community-led review platforms, and discipline-based preprint servers with formal review layers are growing, and the research community has the tools to organize peer review differently and without charging fees to researchers.
We urge Congress and federal agencies to:
- Distinguish public access from pay-to-publish. Public access mandates require that research be freely available, a requirement firmly grounded in the longstanding federal purpose license that gives federal agencies a non-exclusive license to use all works subject to copyright as a condition of grant funding, including deposit in federal repositories. Commercial publishers and pay-to-publish models are neither required by federal policy nor consistent with its intent. Commercial publishers have converted a public mandate into a private revenue stream.
- Recognize that pay-to-publish doesn’t just cost the government and taxpayers more; it compromises the science it purports to support. The federal government does not have a responsibility to protect commercial publishers, but it does have a direct responsibility for the quality of the science it funds. The GAO report documents the consequences of neglecting that responsibility: the proliferation of paper mills targeting open access journals, editorial boards resigning over publisher pressure to prioritize quantity over quality, and at least one major publisher retracting over 11,000 articles after discovering systematic fraud enabled by its own APC-based model. This is not incidental to the pay-to-publish model. It is the predictable result of a system in which publishers are paid per article accepted. The government cannot simultaneously assert responsibility for the integrity of federally funded science and continue to support a publishing model whose profit incentive runs directly counter to that integrity.
- Decline to treat current market conditions as fixed. Cost analysis alone is insufficient if it assumes the current publishing landscape is immutable. Agencies should also analyze how to improve the publishing landscape through infrastructure investment and coordination across the federal research enterprise. A fraction of the projected $3–4.5 billion in publisher fees over the next five years could build a durable alternative to fee-based commercial publishing.
The shift to public access is genuinely good for science and for the public, and the GAO report confirms this. The full weight of its evidence also confirms that pay-to-publish is not a viable path forward for science or taxpayers. It is more expensive, concentrates market power, and as the GAO itself documents, actively degrades the integrity of the scientific record it purports to serve.
The federal government has the legal authority, through the federal purpose license, to make publicly funded research publicly available without paying a commercial publisher to do it. The question before policymakers is not whether federally funded research should be accessible to the taxpayers who funded it. The question is whether the federal government will continue to pay commercial publishers billions of dollars to sustain their business model, or whether it will fulfill its obligation to the public whose tax dollars made that research possible.