Science funding policies could help resolve issues in open access publishing

Open access science publishing faces many issues, which are challenging for individual researchers to address. In contrast, funding organizations are well placed to drive systematic changes in science publishing, as they can directly mandate where grant recipients publish. If funding organizations were to require authors to publish in journals that did not have misaligned incentives compared to scientists, then many issues in open access publishing could be resolved. But before discussing this point further, an overview of the issues facing open access publishing is first needed.

The current science publishing landscape

The traditional business model for journals is based on subscription fees. This results in paywalls blocking scientific knowledge, and means that many researchers cannot afford to access relevant articles. Consequently, some organizations and scientists have pushed for peer-reviewed articles to be made open access (i.e., available for free, generally online). This approach is gradually becoming common in science publishing. For instance, in 2015, 45% of published articles were open access.

This rise in open access publishing has been mirrored by new open access mandates from many science funding organizations, including the Canadian Tri-Agencies. In this case, researchers that receive Tri-Agency funding (based on the current version of the policy, last-modified 2016-12-21) have two options for disseminating their findings: either publish in an open access journal and/or upload their work to an online repository (with permission from the journal).

This requirement, which has been echoed by many funding organizations, could improve the dissemination of scientific knowledge. However, there is also little doubt that increased open access has resulted in massive shifts in the publishing industry. The key driving factors are author-processing charges, which are fees paid by authors (generally to cover journal losses due to missing subscription fees) and are on average $US 1,626. These charges, and the fact that scientific papers can now be published online with low overheads, mean that open access publishing can be extremely profitable for publishers. This is reflected by major growth in for-profit open access journals, including a disproportionate number of predatory journals. Under this system there is clearly the potential for bad incentives to drive for-profit open access publishers to behave against the interests of scientists, who spend grant funds to cover the fees underlying their business models. Most pertinently, these publishers have little incentive to have high editorial standards, and are instead incentivized to maximize the number of papers accepted.

                Key scientific publishing types (as defined here)

Alternatives to traditional journals

The Tri-Agency policy statement described above suggests an alternative to paying author-processing charges: to upload manuscripts to online repositories instead (often referred to as ‘Green OA’, see above table). The exact repositories referred to are university library databases in this case. Unfortunately, in many fields it is not the norm to upload manuscripts reviewed and published elsewhere to these databases. Indeed, most scientists would likely be unsure whether they have copyright permission to do so, and in many cases might be totally unaware of this option.

Journals have clearer policies on whether authors may release their draft manuscript as a preprint. These are generally initial versions of manuscripts that scientists post online at the same time as they submit their manuscript to a journal. Preprints can also be posted without the authors having any intention of publishing in a traditional journal. Peer Community In is a non-profit organization developed by scientists where preprints can be reviewed (disclosure: I have been both an unpaid reviewer and editor for this organization). Preprints that pass review can optionally be published as part of the Peer Community Journal. This is one example of a diamond open access journal (i.e., free for both authors and readers). Despite the clear benefits of this option, these journals made up only 9% of all total publications in 2021.

The prestige trap

So why do scientists not publish in diamond open access journals more frequently? One key reason is the prestige trap: scientists are most motivated by the perceived prestige of a journal when selecting where to submit a manuscript. This often outweighs any author-processing charge. And for good reason: many scientists believe their career advancement is most dependent on where they publish. Thus, there is a catch-22: diamond open access journals will only become prestigious once many scientists start publishing in them, but scientists will only start publishing in them once these journals are considered prestigious.

A possible solution

Publishing in these journals could be normalized if funding organizations mandated that grant recipients publish in diamond open access journals. This would be controversial, and there are several foreseeable criticisms

First, Canadian scientists would be less competitive for international grants and global recognition. To some extent this would be true, but could be partially mitigated if the process was transparent. In other words, it could be made clear that reported publishing patterns were the result of a mandate from funding organizations rather than reflective of the quality of the work of Canadian scientists.

This then raises another criticism: how would we evaluate scientists and their work? This is less of a problem then it seems, as the current standards for evaluating scientists are already commonly criticized. Surely the scientific community can develop improved evaluation standards that are not dependent on for-profit journals. In addition, if diamond open access journals became widely adopted, then they would eventually be partitioned into journals of high and low impact, which is what current evaluation standards are based on.

Many scientists might not agree with my arguments, and there would undoubtedly be other obstacles to widespread adoption. Nonetheless, the current science publishing landscape needs reform, and Canadian funding agencies have an opportunity to be world leaders on this issue. I urge Canadian policy makers to seriously consider this proposal.