Many scientists agree that the current scientific publication system is unsatisfactory and too expensive. The recently introduced 'Plan S', which requires scientific articles to be freely available to readers, may be part of the cure. However, despite its virtues, we argue that Plan S is not ambitious enough, as it fails to directly address the enormous cost of the current publication system built around a 'prestige economy'.
To break the monopoly of expensive paid journals, we should make use of free journals and free platforms. To demonstrate that this can work, we have recently launched a platform on which editorial boards evaluate preprint articles. The money saved by research funders and institutions as we move away from the current publication system could be used as a temporary financial incentive, providing an efficient tool for encouraging the scientific community to adopt this ambitious and radical strategy.
High costs but still restricted access
The current scientific publication system is unsatisfactory. It is so expensive (with costs around €10 billion per year) for libraries and research institutions that it has been called a "rip-off". The cost come from a combination of subscription fees to access scientific journals locked behind "paywalls", as well as from paying article processing charges (APCs, often between €1000–3,000 per article for well-known journals) that some journals charge for open-access papers.
Subscriptions and APCs ensure a mean profit margin of 35–40% for some large publishers. Despite these high costs and the fact that most research is funded by public money, access to most research articles remains restricted by paywalls. Additionally, with only a few notable exceptions (e.g. https://openresearchcentral.org/), the current publication system is not transparent, as most editorial decisions and reviews are unpublished and hidden away from any scrutiny, making it impossible for readers to estimate the quality of the evaluation and editorial processes leading to the publication of articles. On top of all this, many evaluation committees and scientists themselves judge the value of articles on the basis of 'journal impact factors' and similar journal branding, two non-scientific criteria almost entirely unrelated to scientific quality. These types of non-scientific criteria are instead often directly detrimental to research and researchers.
Open by default
Plan S aims to promote open science by mandating that scientific articles originating from research funded by a large coalition of funders (Coalition S) are freely available to readers without paywalls or restrictions. The implementation of Plan S would represent a major advance, heralding a new era towards where science is “open by default”.
Various publication models are compatible with Plan S, most widespread being open access journals ("gold OA", with or without APCs). Subscription-based journals are also compatible with Plan S, provided that the published article or the authors’ accepted manuscript is deposited in an open-access repository ("green OA").
While most OA journals are free of charge, the majority of OA papers are published via paying an APC, and authors almost always have to pay APCs (commonly between €1000–3000) to publish in the most well-known and prestigious OA journals. The gold OA route is, therefore, costly. The green OA route makes the authors' copy of their article freely available, but these are typically copies of articles published in subscription journals, and without subscriptions these journals would cease to exist. In other words, both of these routes remain costly and any future publication system following these routes will therefore remain expensive for society.
Although Plan S may lead to more open access, it provides only a limited solution in terms of the rapidly increasing publication costs for libraries and research institutions (often termed the 'serials crisis'). Instead, it is likely to result in a transfer of publication costs from readers (via libraries) to authors (via funders). However, promisingly, Coalition S has recently promised to support innovation towards the development of new and improved publication infrastructure.
Radical solutions for the digital age
We think that Plan S provides an opportunity to change the economic rules governing scientific publication, with the potential to greatly decrease overall publication costs. Indeed, Plan S could be used as the basis of a new publication system that would be much more open, radically different, highly efficient and much less expensive.
Open-access internet platforms, such as preprint servers (e.g. bioRxiv), OA institutional repositories (e.g. HAL) or e-print open archives (e.g. arXiv) provide most of the fundamental scientific services currently offered by journals and publishers at no cost to authors or readers: they make articles publicly available via the internet, archive them, provide authors with proof of anteriority, and ensure that it is easy to search for and find articles. They also allow the authors to keep their rights (e.g. authors are allowed and encouraged to use an open licenses such as a CC BY license).
Despite these advantages, one key missing element is the evaluation and validation of these articles by the scientific community. Independent peer review combined with decisions taken by a specialist editor is the standard process and a desired quality control step. While this peer review-based editorial process is regularly criticized, and there is much exciting innovation on-going in the area of peer review, the conventional editor-assisted peer-review model remains the gold standard. As a result, most of the scientific value created by journals does not come from the brand name or journal impact factors, but from the quality of the editorial process of a journal, and the researchers and research community associated with it.
Putting researchers at the centre
Building editorial boards organized and recognized by scientists themselves, outside the conventional market of paid journals (including both pay-to-read and pay-to-publish journals), for the evaluation and curation of articles made publicly available via free OA internet platforms would radically decrease publication costs while safeguarding quality control through peer review. In such a system, scientists could deposit their articles on OA internet platforms and submit them to one of many editorial boards for evaluation. Alternatively, editorial boards could invite any interesting article they find on a OA platform.
After one or several rounds of conventional peer review, the evaluation would lead to the acceptance of a final version of the article by the editorial board if deemed to be of sufficiently high quality, and the editor(s) would endorse and recommend the article. This recommended version, like all the revised versions during the evaluation process, would be available on the OA internet platform chosen by the authors. The recommended version would then be considered a final, peer-validated article of the same value as any other published article which has gone through a rigorous peer review process.
Creating change, today
We founded the project called Peer Community in (PCI), which is a group of platforms recommending preprints in evolutionary biology, ecology, paleontology, animal science, and entomology—so far, but with a view to expand into other disciplines—which provides an example of a researcher-controlled scholarly publishing system. There is no need to (further) publish the accepted/validated version of the recommended article in a conventional journal, because it already has all the fundamental characteristics of a validated scientific article: (i) publicly available on the internet, (ii) archived, (iii) proof of anteriority, (iv) searchable/findable, and (v) validated by peers.
Excitingly, a range of other initiatives propose evaluations of articles findable on OA internet platforms (listed here reimaginereview.asapbio.org), highlighting multiple potential future directions. But to our knowledge, none of these other initiatives allow authors to submit their articles for evaluation, whereas PCI attempts to bridge this gap between manuscript evaluation and peer community creation and engagement.
This new publication and evaluation system, PCI, would lead to the publication of scientific articles in open access at a much lower cost than the current system: see this blog post for details on the PCI economic model. This follows as publication on OA internet platforms is much cheaper than publication in conventional journals. The current cost of the e-print archive arXiv.org is less than $10 per article (see here and here). Whereas current APCs in some conventional OA journals exceed €2000, publishing a peer-reviewed validated article via the PCI system could cost much less: not €1000 but instead around €10. The cost of evaluation by editorial boards is very low, even in the current system, because this task is part of the academic missions of researchers: most university-employed academics who conduct peer review and are editors for journals are not paid extra for these services as it is seen as part of the job.
The difference between APCs and the actual costs of publishing lies in several costs (e.g. advertising/brand-building, lobbying, grants/prizes) which are not directly related to the publication and evaluation process. The new system we propose would be complementary to the free OA journal ecosystem (sometimes called diamond OA). According to the DOAJ (DAOJ), three-quarters of OA journals have no APCs, and some also make use of OA internet platforms, such as Discrete Analysis, or the family of journals of the overlay journal platform Episcience.
Imagine a world in which this new system—gold OA journals without APCs and free OA internet platform displaying validated, peer-reviewed articles—at least partially replaces the old one. Paid journals may still be needed, but they would instead concentrate on any additional services not offered by free systems. For example, such journals could concentrate on editorials, commentary, outreach and news coverage.
How can we convince large parts of the scientific community to embrace something so radically different from the current publishing system? Community acceptance and uptake will undoubtedly be challenging, because even if researchers are convinced that the system needs to change, they feel trapped in the current culture based on prestige and journal brand names for both career advancement and to ensure that they obtain funding. Furthermore, researchers are unlikely to change paradigms in the absence of proof that the new system works.
Our proposed plan is even more radical than Plan S, and while Plan S has attracted considerable support from funders, researchers and researcher associations, it has also encountered resistance from commercial publishers, as expected, but also from some researchers. The initiators of Plan S are currently working to build momentum and obtain broadest possible support from around the world, but Plan S remains hotly debated.
To build researcher-centric publishing systems such as PCI, we foresee two main obstacles. First, journal brand names and journal impact factors are too often used by evaluating committees as proxies for the evaluation of institutions, projects and scientists (for promotion, hiring, and funding) which fuels the status quo and disincentivizes innovation. Secondly, lack of incentive to change which will undoubtedly stall the process.
Leaders of Coalition S have clearly stated that they will modify their evaluation procedures to "fundamentally revise the incentive and reward system of science, using the San Francisco Declaration on Research Assessment (DORA) as a starting point". This is a good start.
Several members of this coalition and other funders have signed DORA and other initiatives, such as the Jussieu Call for Open Science and Bibliodiversity calling for profound reform of the research evaluation system. For example, the largest European basic science research agency recently stressed that the evaluation of researchers should focus on the content of their work. Encouragingly, good practices for funders, research institutions and professional societies have already been established. These provide clear examples and pathways for others to follow.
We need a large, clear and efficient communication campaign for a reform of evaluation practices to change the behavior of scientists and to wean them off of the journal impact factors and journal brands on which many currently depend. PCI has been gathering support and we recommend more research institutions to publicly commit to: (i) consider the PCI system as a legitimate means of evaluating and validating scientific results, (ii) treat PCI-recommended articles in the same way as articles published in conventional peer-reviewed journals, and (iii) encourage their members/colleagues/students to use PCI as a research outlet (as readers, authors, reviewers and/or editors). A more general appeal relating to gold OA journals without APCs and free OA internet platforms displaying validated, peer-reviewed articles could serve as the basis for such a communication campaign.
Putting resources back where they belong
Temporary incentives would be useful for propagating the new publication system. A financial stimulus sourced from the huge amounts of resources saved by funders and institutions that move away from conventional expensive journal subscriptions and APC-driven publishing, could be a highly efficient tool for encouraging the scientific community to adopt this ambitious and radical strategy. For example, Coalition S funders could include a publication budget in the grants they allocate, enabling researchers to decide whether to use gold OA for which they have to pay APCs, or whether to use this part of the budget for their research if they decide to use free alternative publishing systems.
This would provide researchers with a direct motivation to avoid paying APCs and encourage the use of free OA platforms. It would also allow Coalition S members to remain neutral with regard to the specifics of various OA models, and focus their efforts on promoting the general principles of immediate and full OA. This would also conform to the view of the League of European Research Universities and others that "research funding should go to research, not to publishers".
Plan S provides us with a fantastic opportunity to change the rules and promote open science, but we fear that open publishing—if not done right—would still cost worldwide research budgets billions of euros. We encourage research institutions and funders to join forces to help researchers to switch to a free and fully open-access, internet platform-based publication system.
Financial incentives combined with rapid modernization of how we evaluate research (focusing on research content and not packaging) are the two key instruments that research institutions and funders should wield to reach this goal. The use of these instruments would enable researchers to break free from the current competitive, prestige-based publishing system and embrace an open, shared and collaborative approach to publishing.
Note: This article is published under a CC BY license.