Open for business: Problems facing open access scientific publishing

Published: October 22, 2022

by Gavin M. Douglas, Science & Policy Exchange (SPE)

Open access publishing has increased global access to scientific knowledge, which most would agree benefits humanity. However, the open access movement has been accused of having some negative impacts on the scientific publishing landscape. A clear understanding of these specific criticisms (outlined below, and which builds upon a previous SPE blog post) is necessary for scientists and policy makers to develop solutions.

“It’s too expensive and wastes research funds”

Author-processing charges, which are fees that scientists pay to publish open-access papers, are the foundation of open access journal business models. In contrast, traditional journals rely on subscriptions (e.g., from universities) as their main revenue stream. These processing charges are on average $US 1,626, but vary highly across journals. For instance, in 2020 the author-processing charge to make a paper open access in the prestigious journal Nature was $US 11,390. Even reasonable author-processing charges can be a burden to researchers. Although author-processing charges are frequently anticipated expenses in research grants, they still result in substantial grant funds being allocated away from research (see plot below).

Plotted values based on Table 4 in Zhang et al. 2022.

“It encourages predatory publishing practices”

The shift in journal business models from subscription-based to reliance on author-processing charges has led to a proliferation of predatory journals, which offer virtually no editorial services. Of course, the existence of these journals should not detract from the valuable contributions of reputable open science publishers, and hopefully fraudulent publishers will become extinct as researchers become better informed.

“Editorial standards are too low”

More controversially, certain legitimate for-profit open access publishers have been accused of implementing low editorial standards to ensure a high number of papers are accepted. Open access publishers primarily maintain online-only journals, which means they have low overheads for publishing papers. More papers means more author processing charges collected, so there is a clear business incentive to maximize the number of accepted papers. This bad incentive has been highlighted in several high-profile cases. For instance, editors of the MDPI journal Nutrients stepped down in protest after alleging they experienced strong pressure to reconsider rejected manuscripts. In other cases, certain for-profit open access journals have been criticized for only providing reviewers with the option to withdraw from the reviewing process rather than to reject a manuscript. After reviewers withdraw, new reviewers can be acquired until adequately favourable evaluations are produced. Accused publishers have ready-made responses to such allegations (as in the Nutrients case): corroborative results are key to scientific progress, so even if a manuscript is not strictly novel, it still deserves to be published. This is of course true, but it highlights how some publishers misdirect quality concerns about manuscripts to instead focus on novelty concerns, which are not equivalent. Regardless, there is a conflict of interest if the publishers making these judgement calls also have business models based on maximizing the number of publications.

“It exacerbates existing inequalities in career advancement”

Papers published in open access journals are cited more on average, although the effect size is controversial. Authors that publish in these journals garner several other career advantages as well, such as additional job and funding opportunities. This is a problem, because not all scientists, institutions, and countries can afford to pay high open access publishing fees. This means that researchers with less funding could miss out on key funding and career opportunities, while well-funded researchers might be disproportionately benefiting.

Journal categories, partitioned by openness and economic model. From Racimo et al. 2022 (CC BY 4.0 license).

What should be done?

These critiques are all valid, but it is important to realize that they do not apply to all open access journals. There are many different open access models, including journals run by scientific societies and non-profits (see figure above). There are also a growing number of diamond open access journals, which means that they are free for both authors and readers. If researchers shift to publishing in these journals, then these issues could be avoided. How can a systemic change like this be performed? It won’t be easy, but stay tuned for future SPE pieces on this topic.