Editor’s Note: Today’s post is coauthored by Ann Michael and Dan Pollock. Dan is the Chief Digital Officer at Delta Think, where he focuses on data analytics, market intelligence, and strategy.

On the 28th March, the Bill & Melinda Gates Foundation (BMGF) announced a “Policy Refresh” to their current policy on open access (OA) publications. Full details were released about a week later. The headline changes are that, from the start of 2025, published research funded by the BMGF must be made openly available as a preprint, with no embargo, under a CC BY license. The BMGF will no longer pay any publishing fees (article processing charges — APCs), but grantees may choose to publish in any journal – OA or not.

This contrasts with the current policy, in effect since 2015, which mandates immediate open access publication of BMGF-funded research journal articles and covers publication fees for journals listed in the DOAJ (Directory of Open Access Journals). Preprint publication is encouraged, but not compulsory. Other elements remain largely unchanged, such as deposit in a repository, grantees’ retention of copyright, “proper metadata” to identify Gates funding, and making underlying data openly available.

We will leave it to others to pore over the details and definitions, analyze the ambiguities and debate the pros and cons. We will also let others take a position on whether they agree or disagree with what the Gates Foundation has done.

Here we aim to understand what they intend to accomplish and to ask what this might mean to scholarly publishing and the dissemination of research.

Hand holding wooden cubes with arrow icon, moving one out of position to point to a new direction

It’s not about the money

In a blog post about the changes, Estee Torok, Senior Program Officer, Malaria, noted the high costs of publication born by the BMGF – around $6 million in APCs per year since 2015. By analyzing the BMGF’s annual reports, this $6m annual spend on APCs since 2015 accounts for around 0.12% of its total expenditure on charitable support between 2015 and 2022.

Arguably this number understates the share of expenditure. A significant proportion of the BMGF’s activities include things such as advocacy, delivery of health programs, direct support for education, or translational work. These are not primary research activities and won’t lead to publications.

We don’t have figures for the subset of spending on primary research, so, for the purposes of analysis, we can try a different method. In 2022, based on Delta Think’s market data, we estimate that the total scholarly journals market accounted for around 0.43% of global spending on R&D. Again, much of this arises from activities that don’t lead to research publications. But we can use this as a like-for-like (albeit somewhat crude) comparison to understand publishing costs as a share of total activities.

To further put costs into context, the BMGF was publishing an average of 2,250 articles per year in 2017, rising to 4,000 in 2022 (or just over 0.13% of total scholarly output in 2022). Projecting article numbers against the $6m per year spent on APCs, we estimate an average APC paid of $2,162. These are crude estimates due to the gaps in the data, but according to data from Delta Think, this is a few percent lower than market average APCs for health or life sciences.

This information suggests that the costs of publication are relatively modest. As we read more of what the BMGF has written, the changes appear to be about mission alignment rather than savings.

It’s about the mission

“Open Access as it is done now hasn’t changed for ten years and in fact has given rise to some unsavory publishing practices by poor actors (paper mills, questionable quality review, unchecked pricing). This has caused the foundation to rethink how it supports Open Access to better achieve its goals of immediate access, global reuse, and equitable action.”  – BMGF OA policy refresh website FAQs.

The Policy Refresh Overview outlines the ambition behind the changes in policy:

  1. using preprints to publishing research when ready, “to prioritize access to the research itself rather than to a particular journal”
  2. withdrawing support for APCs, “to address inequities in current publishing models”
  3. supporting an OA infrastructure “to ensure articles and data are available to a wider range of audiences”.

The various FAQ pages and communications cover the BMGF’s philosophy and vision at length. There is plenty to digest, but this quote stood out to us as we examined the change from its current OA policy.

The BMGF is placing more value on early, open, and inclusive research dissemination. The method it is employing is to lead with the preprint, and de-prioritize – but not remove – the scholarly journal.

Why now?

The BMGF’s blog post notes that During the COVID-19 pandemic, preprints flourished as researchers, governments, and others raced to develop vaccines and therapeutics.”

COVID-19 saw a quantum leap in use of preprints, both within and outside the scientific community, with many scientists turning to them for the first time. It has also provided a rich source of data for numerous studies on the effects on preprints. Journal articles do not become redundant, nor are preprints on the whole a worthless wild-west of poor quality. Rapid dissemination is of great benefit and concerns about the shortcomings of preprints can be addressed. COVID-19’s preprint bump is set to have a lasting effect on research publishing.

There is always room for further study, but with sources such as The Lancet Global Health or Nature Biotechnology suggesting pre-prints are robust and even indispensable, it’s unsurprising the BMGF is betting their time has come.

Plus ça change

It’s clear from its rhetoric that the BMGF is not aiming to replace the current scholarly journals system. In the context of the wider process of research dissemination, leading with preprints is a change in emphasis from leading with journals. Whether this represents realignment or revolution is a matter of perspective.

Whatever the doubts and details, this move should be seen as an experiment.

BMGF is hypothesizing that the time is ripe for change, and that an alternative method of research publishing may prove more efficient and effective than the status quo. In this light, what might success look like to BMGF? What are they testing?

From our perspective, success for the BMGF may be contingent on three key questions.

1. Can the BMGF be a catalyst for the community to change?

“The refreshed policy operates in parallel with established publishing and discovery practices, and is meant to stir new innovation and experimentation, and give air to new models and outputs. Grantees can still publish in journals, but that output is no longer a requirement of funding.” – BMGF OA policy refresh website FAQs.

The notion of “publish-then-review” is nothing new. Preprints have been used to communicate scientific results in some disciplines for over 30 years. But outliers aside, preprint culture has not taken off when compared to the volume of scholarly publications – certainly not in biomedical and health sciences. Over the 6 years from 2017, the F1000-powered Gates Open Research platform published just 248 research article preprints – yet the BMGF funded thousands articles over that time.

Organizing a community around peer review and providing the infrastructure to support it has been one of the key contributions of scholarly publishers. However, publishers operate predominantly within the context of journals. If BMGF’s vision is to take off, something will have to fill this gap for preprints.

The announcement of VeriXiv – “a new verified preprint platform that will enable the rapid availability of new findings and promote research integrity” – suggests the BMGF is thinking about just this issue. VeriXiv will run integrity checks over content, consistent with the new BMGF’s policy of mandating preprint servers with a “sufficient level of scrutiny to submissions”.  While this is not peer review, and the details are not final, it is an attempt to address infrastructure and quality.

Only time will tell if the BMGF’s policy will corral sufficient levels of engagement from the community.

2. Is the pre-print good enough as compared with the VoR?

“To meet the urgency demanded by today’s greatest global challenges, there must be a culture shift in publishing that ensures the prioritization of equity and access over prestige and personal interest.” – Open Access Policy Refresh 2025 – One Pager.

The debate about the value of the journal Version of Record (VoR) is ongoing. Whatever the arguments, the adoption of preprints will depend on researchers’ views of their relative value to the VoR. As we note above, COVID-19 may have proven to be a game-changer. Preprints do add value – and traditional, high-quality, highly-selective publishing is not infallible, as the well-known retractions of papers about treatments for COVID-19 demonstrate. But decades into the digital age, the journal article has proven remarkably resilient, and formal peer-review remains of value.

If researchers do work with preprints more widely, then we should be able to gather further evidence about the comparative value of preprint and VoR, and of the roles that each might play as any new system evolves.

3. Will this result in more equity and inclusion in research communication?

“It is not sustainable for the foundation to continue supporting inequitable models that operate counter to its mission and priorities, and disadvantage directly the populations the foundation’s investments are meant to help.” Bill and Melinda Gates Foundation

BMGF sees the current journal publishing model as denying many access to the research process. Moving to APC-based open access has failed to address the fundamental issues of affordability and other open models (such as S2O) have not gained traction. While BMGF can afford to pay fees for its own authors, it positions its new policy as mitigating issues of access to the sharing of knowledge for all. Discussions about problems with the APC model are current, with organizations such as OASPA and cOAlition S (and others) also raising concerns about APC-based models.

The focus of BMGF’s work puts it in a good position to test how preprints may help with equitable access to the publishing and reading of research information. It should be able generate data to assess if there is more access and uptake on BMGF funded research because of its policy.

Are researchers ready for a change?

While the BMGF may be all-in on this approach, from an industry perspective the BMGF Policy Refresh represents a small but potentially valuable experiment.

As with all new processes, there will be bugs and details to iron out. In many ways, addressing these is the easy bit. To do so requires commitment, resource, and expertise – all of which we can reasonably assume the multi-billion-dollar BMGF can supply or acquire in abundance. It will be interesting to see what metrics and KPIs BMGF develops, what data it produces, and what interventions it deploys to refine and steer its process.

The harder bit is to determine whether the market – the ecosystem – will be receptive to a new approach, and if that approach will achieve BMGF’s objectives. The only way to resolve this is through experimentation, by pushing ahead with something new and seeing what happens.

The BMGF may be a small player by volume of articles published. But it is a large funder, with an influential voice, and not alone in its desire for change. If the answers to the questions we pose above are “yes” then it would suggest that the world is ready for change and may follow BMGF’s lead. If not, then the world isn’t, and it won’t.

Which of these outcomes is defined as “success” will depend entirely on the agenda of those making the assessment. Our perspective is that any outcome from the process of experimentation – even an inconclusive one – may be deemed to be successful. We’ll have learnt something.

Ann Michael

Ann Michael

Ann Michael is Chief Transformation Officer at AIP Publishing, leading the Data & Analytics, Product Innovation, Strategic Alignment Office, and Product Development and Operations teams. She also serves as Board Chair of Delta Think, a consultancy focused on strategy and innovation in scholarly communications. Throughout her career she has gained broad exposure to society and commercial scholarly publishers, librarians and library consortia, funders, and researchers. As an ardent believer in data informed decision-making, Ann was instrumental in the 2017 launch of the Delta Think Open Access Data & Analytics Tool, which tracks and assesses the impact of open access uptake and policies on the scholarly communications ecosystem. Additionally, Ann has served as Chief Digital Officer at PLOS, charged with driving execution and operations as well as their overall digital and supporting data strategy.

Dan Pollock

Dan Pollock is the Chief Digital Officer at Delta Think. His background includes executive leadership and strategic advisory roles, and senior positions managing products, operations and change in organizations ranging from startups to complex non-profits and commercial corporations. At Delta Think, Dan focuses on data analytics, market intelligence, and strategy.


2 Thoughts on "Gates Policy Refresh: What Would Success Look Like?"

Nothing much quoted here from authors who were funded by the Gateses.

But then again, maybe they have spoken very loudly. “Over the 6 years from 2017, the F1000-powered Gates Open Research platform published just 248 research article preprints – yet the BMGF funded thousands articles over that time.”

In social sciences (esp. in Economics) there is a decades-old continuing practice of hosting discussion/working papers. Quite a few of these end up in journals as published articles. This practice never hurt the business and neither did they harm science. This being called as “preprints” and considered new was due to natural scientists discovering ArXiv and other platforms. If the industry had paid equal attention to trends in Social Sciences, everything will not seem like a new experiment! BMGF is pretty much creating a discussion paper repository in that sense (SSRN, CEPR, Munich Papers, etc have all working models for such a thing already).

Comments are closed.