Publication costs – "we are on the edge of the abyss"

What is your assessment of the situation with article processing charges?
Alain Schuhl: We are currently observing a dual increase. Firstly, the unit rates of APCs are rising steadily. Secondly, there has been an increase in the number of articles published following the payment of publication fees. The amounts involved depend on the scientific field but the increase concerns all disciplines. These worldwide and national trends are also reflected at the CNRS.

In tangible terms, what are the actual amounts being paid?
A. S.: In 2020 the CNRS paid over €3 million in APCs compared with €1.8 million in 2017. A recent prospective study in France run by the French Ministry of Higher Education and Research1  reported that France paid almost €30 million in APCs in 2020. The same study suggests that at this rate of increase France could be paying €50 million in APCs by 2030 or even nearly €200 million if all scientific publications have switched to the author-pays model by then. This amount is far higher than the budget of the subscription-based model which is already excessively expensive.

As well as these €3 million, we can observe that the CNRS paid nearly €1.3 million in other types of fees such as 'colour charges' (to display images in colour) or more general 'page charges'. Moreover, publishers seem to have a limitless wealth of imagination when it comes to coming up with new charges - 'submission charges', 'fast review charges', 'delayed Open Access charges' and so on.

The current trend is therefore worrying and we need to put a brake on this downward spiral.

Which publishers does the CNRS pay the most charges to?
A. S.: The leading publisher is by far the Springer Nature group. This is particularly due to the success of its two mega-journals Scientific Reports (currently a €2090 APC per article) and Nature Communications (currently a €5190 APC per article).

Next, we have observed a significant increase in the sums paid to two native open access publishers, Frontiers Media and MDPI (Multidisciplinary Digital Publishing Institute). These two publishers account for the most marked increase in APC expenditure in the last four years. We observed a 139% increase in charges paid to Frontiers Media and 746% for MDPI2  whose average APC charges per article are now around the same as Springer Nature's.

But all the major scientific publishers like Elsevier, Wiley, IEEE, the American Chemical Society and so on have increased their charges.

We hear a lot about predatory publishing. Does this affect the CNRS?
A. S.: By offering authors the possibility of being published in return for a fee, the author-pays system has increased the possibility of publishing research and this is the antithesis of all notions of scientific ethics or integrity. It paved the way for the emergence of predatory publishing based on an economic model that encourages the existence of a race to publish. This brings two contradictory systems into opposition - a financial system geared towards profit on one side and a scientific validation system intended to improve the state of knowledge on the other. Worse still, if predatory practices continue to creep further into publishing, it could suffice for authors to pay to be published with no guarantee of the scientific quality of the results of their work. Research assessment bodies really need to be aware of this.

Predatory publishing is a real problem, and researchers are not safe from this, if only because they are frequently solicited to publish, serve on peer review committees for these journals or are asked to direct a special issue. Among the most common predatory practices we can cite slipshod or even non-existent peer reviewing, very high manuscript acceptance rates, abnormally quick turnaround times between the submission and acceptance of a manuscript, a frenzy of special issue launches, stealing the names of existing journals, identity theft and so on. Scientists must be extremely vigilant about such practices. These are already being taken into account by certain research assessment bodies which take predatory journals off researchers' publication lists or even consider publishing in predatory journals negatively.

However, the issue is complicated by the fact that many journals are in a kind of a grey area in which they are neither completely fraudulent nor completely honest.

What can you tell us about Frontiers Media and MDPI more specifically?
A. S.: They are typical examples of the open access publishers we need to be vigilant about if only because of their practices encouraging researchers to publish, the quantity of new journals they launch on the market and the short times they offer between an article being submitted and being accepted. For some years now, MDPI and Frontiers Media have also distinguished themselves through the sheer volume of their special issues.

MDPI has become the world's third largest publisher in terms of the annual number of articles in just a few years. For the CNRS, this is now the fourth largest publisher in terms of numbers of articles published by our research units and the second largest publisher in terms of the amount of APCs paid. Frontiers Media's growth dynamic is similar and its editorial practices are also comparable.

Researchers have the right to publish in the journal of their choice but we encourage them to be particularly vigilant about the editorial practices of the journals they submit their manuscripts to. Tools like Compass to Publish which was created at the University of Liège and Think Check Submit enable researchers to become informed about this issue.

What effects will this have on the assessment of scientific research?
This is certainly the most important initiative in our action on this issue. Researchers' practices will change if they come to realise that the evaluation of their scientific work only takes into account its quality and the real advances in knowledge their work contains and totally disregards quantitative criteria.

Reforming research assessment is essential if we are going to successfully transition to a better science that necessarily has to be open. The CNRS has demonstrably adopted a proactive policy on this issue by modifying researchers' assessment forms and also now the requirements for applications to the CNRS. To be fully effective, this action must be an integral part of an international dynamic. For this reason, in 2022 the CNRS contributed to the creation of the international Coalition for Advancing Research Assessment (CoARA) that is dedicated to the reform of research assessment and made up of universities, research organisations, academic societies and so on. The CNRS is actively involved as one of the founding members and Sylvie Rousset, the Director of the Open Research Data Department (DDOR), has been elected to CoARA's board.

What do you think are the future perspectives for scientific publishing?
A. S.: The current system is currently replacing inequalities in people's access to scientific literature maintained by the subscription system with inequalities in researchers' capacity to publish scientific articles. Only teams with a sufficient budget will be free to choose where they publish which will further exacerbate North-South inequalities. We are currently on the edge of the abyss because subscriptions still exist and the payment of APCs is becoming increasingly widespread. It is up to us not to take that extra step over the edge. If all publications switched to immediate open access with APCs, given the budgetary resources this already requires, this model would not be financially sustainable. As I said in April 2022, open access cannot be reduced to paying to be published.

• 1Antoine Blanchard, Diane Thierry, Maurits van der Graaf, Retrospective and prospective study of the evolution of APC costs and electronic subscriptions for French institutions, Committee for Open Science, 2022. (hal-03909068)
• 2Increases observed at the CNRS on its expenditure for 2020 compared to the organisation's expenditure in 2017.