There’s a growing consensus from governments and funders around the world that publicly funded research should be published openly and preferably openly immediately. Where policymakers diverge is on the mechanism for achieving immediate open access (OA). That divergence is leading to a regionally fragmented publishing sector that creates a strategic problem for publishers: the funding routes they relied on are being withdrawn, and one global model no longer fits every market. In principle, none of this is new. Publishers have run differentiated regional strategies for years, but the pace and the stakes have changed. It also sits alongside other policy fronts moving just as fast, from how AI systems use and help produce scholarly content to questions of copyright and verification. Those issues deserve separate treatment; here, we’re focusing specifically on the rapidly changing policy and funding environment around open access.
1. Institutions are hitting their OA allowance ceilings
In September Dutch researchers publishing in Springer Nature, Emerald, and Oxford University Press hybrid journals will stop having their article processing charges (APCs) covered. The country has used up this year’s allowance months before the year is out. The deals in question are read and publish agreements – a form of transformative agreement (TA), where a library or consortium redirects what it used to spend on subscriptions into a deal that also covers open access publishing for its own authors. They typically cover a set number of articles a year. Once that allowance is used up the agreement stops covering the charge, and the author can pay the APC themselves, publish behind the paywall in the same journal, or make the accepted manuscript open through a repository (should the publisher’s policy allow for this). TA caps are reached routinely, in many countries and with many publishers. The Netherlands is a clear current example of what happens when the allowance runs out.
The same pressure shows up well beyond Europe. In interviews we conducted with Chinese librarians and researchers last year, quota exhaustion was a routine feature of institutional read and publish deals rather than an exception. One library described its allocation with a society publisher as simply not enough, while another told us that once the quota is gone the cost falls to individual schools, whose entire discretionary budget for the year is smaller than a handful of APCs. India’s One Nation One Subscription (ONOS) scheme shows a similar pattern from the funding side. There’s a genuine commitment of money (roughly ~£11.75 million a year is earmarked for APC support) but a narrow scope, since only journals ranked in the top 5% by CiteScore, SNIP or SJR currently qualify.
To further illustrate the growing pressures on agreements, some institutions are simply walking away from deals with publishers. At least eleven UK universities declined Elsevier’s 2026 offer, with Sheffield calling the deal financially unsustainable. In May the University of Nottingham did not renew five agreements, covering Cambridge University Press, Oxford University Press, Ovid/Wolters Kluwer, Taylor & Francis and Wiley. The same dynamic is playing out in the US. The University of Delaware cancelled its entire Wiley package from January 2026. Only 2% of titles in the package accounted for the bulk of actual usage against subscription costs that had risen 5-9% a year while collections budgets stayed flat. George Washington University cut a swathe of low-usage, high-cost journal subscriptions as part of a wider 3% budget cut, with one Elsevier package alone costing, in one Faculty Senate Standing Committee on Libraries member’s words, roughly “ the cost of a seven-acre private island in Belize”. The backing away from deals does not necessarily mean losing all access to subscription content: institutions may retain backfile rights or individual subscriptions. But researchers can lose access to titles previously covered by the deals and, if they want to publish open access, may need to find another source of funding.
2. Funders want immediate access, but are increasingly unwilling to pay for it
In addition to the institution-publisher dynamic, funders are increasingly mandating immediate open access. In the US, the NIH has required the work it funds to be publicly available with no embargo since July 2025, and Japan has mandated immediate open access through institutional repositories since April 2025. Australia’s Research Council now requires immediate open access, and South Africa made open access mandatory for publicly funded research in December, while cOAlition S’s 2026-2030 strategy widens the routes it will accept without changing its underlying objective.
While funders are mandating immediate open access publication, many of them are also withdrawing support for an APC-mediated route to achieving it. Cancer Research UK (CRUK) will save 5.2 million pounds over three years by ending all support for article processing charges (APCs). They wrote “the problem is it hasn’t worked. At least not in its current form.” CRUK’s objection is that they pay for researchers to publish open access in journals and then their institutions pay again to read them, which they describe as publishers ‘having their cake whilst also eating it.’ Similarly, UKRI has also stopped its funding being used for charges in transformative journals and will phase out hybrid support from 2028-29. If you strip away the policy detail, the question is the same every time: who pays? Funders want immediate access without always funding the APC route to get there and institutions are hitting the ceiling on what they will cover.
This withdrawal of support for APCs extends far beyond the UK. The Gates Foundation stopped paying article charges in January 2025, requiring preprints instead, while Switzerland’s SNSF will refuse any charge above CHF 3,500 from 2027. In the US, the NIH consulted on capping publication costs and has published nothing since, while a proposed OMB rule would make publication costs unallowable on federal grants unless specifically pre-approved. That drew much attention and cause for concern from the sector, and in August the Senate voted 90-6 to block the rule until December. Though at the time of writing, the House has yet to make a decision. So by and large, we have more open access mandates, and less willingness among major funders to pay for the APC route. The destination is converging while the funding mechanisms are not.
3. China isn’t waiting for a global consensus
While publishers are grappling with regional open access policy changes, China, the largest producer of publications in the world, is building its own domestic publishing infrastructure. In March, the Chinese Academy of Sciences (CAS) confirmed it would stop updating its journal classification list, a table that was itself compiled in China but used to rank international, largely SCI/SSCI, journals — in favour of assessing individual research outputs on their own merits rather than the prestige of the journal they appear in. Separately, two China-built alternatives to standard citation-based journal metrics, the Dongbi Index and the Xinrui Journal Ranking, have emerged this year and are gaining attention within China’s scholarly publishing community. Phase II of the national Journal Excellence Action Plan has put 1.2 billion RMB (~130 million GBP) behind 120 new journals. All the while, CAS is withdrawing support for journals charging more than $5,000 an article. The scale behind this is huge. Based on data we pulled from Dimensions, China’s research output has grown at roughly double the global rate for most of the past decade – a 12% annual growth rate since 2020, against under 6% for the world as a whole, and by 2025 it accounted for close to a sixth of everything the database indexed. While much of Europe and North America is debating how to finance openness, China is investing in reducing its dependence on Western publishers. Whether this will lead to a decoupling of the sector is unclear, but the direction of travel is something publishers should be planning for.
4. One publisher, different models by region
Publishers are trying to protect two things: the quality and sustainability of what they publish, and the ability to serve authors under whatever funding rules apply in their market. Given the regional fragmentation of policy, some publishers have come to the conclusion they can’t run a single model globally. The Royal Society of Chemistry (RSC) had previously committed to making all fifty-eight of its journals open access by 2028, and last year it moved away from that in favour of regionally tailored models. Sara Bosshart, RSC’s head of open access put it plainly: ‘one size cannot fit all’. In Europe that means its Platinum Consortia model, spreading cost across institutions; in North America it means working with the growing interest in Subscribe to Open; in China, a mix of open and subscription. Regional differentiation is one way of pursuing both, but it comes at a real administrative cost.
For publishers, the options aren’t a binary choice between one model or many models. Many opt to retain a global core and vary business models, eligibility rules and workflows by region. Adaptability is however becoming a strategic capability in its own right. The cost will not be evenly distributed and likely favour scale. Large publishers can spread the fixed cost of multiple workflows, pricing structures and compliance rules across sizeable portfolios whereas smaller publishers and societies may find it more difficult to absorb it. This could be another factor that will increase pressure for partnerships, outsourcing and acquisitions. If this is of interest to you, I’d encourage you to read our white paper ‘Safeguarding the future of society publishing’.
The sector is very much in a state of flux: CAS retired its journal classification list in March, the Australian mandate took effect in July, Dutch allowances run out in September, CRUK stops paying in October, and the US Congress either settles (or doesn’t) the fate of the OMB’s proposed rule on federal publication costs in December. This is all before accounting for growing expectations around investment in upholding research integrity, which deserves its own blogpost. The challenge facing publishers is less about finding the next universal model than being able to operate several models at once without making the experience intolerably complex for authors and institutions. What’s changing in the scholarly communication sector right now? Everything, everywhere, and all at once.
Research Consulting is hosting a panel discussion on Tuesday 29th September at 1PM BST to explore exactly this kind of regional divergence in more depth, with perspectives from Europe, North America, China and policy leadership: Cathy Foley, former Chief Scientist of Australia; Alice Meadows on North America; Ning Zhang on China; and Dylan Storan on Europe. Register here: https://us02web.zoom.us/webinar/register/WN_WXmoXw-ySemhp4hmKiG5fQ
