Douglas Verret
If Plan S were a person, it would be a lummox, you know, the kind of person you would never let into a china shop or, for that matter, any place having valuable objects. Oh, Plan S has good intentions alright, but the way it has chosen to go about promoting open access (OA) is clumsy and not well thought out.
Before proceeding, it might be a good idea to define a few things for the uninitiated. I have heard it said that there is nothing like a good war to teach people about the geography they never bothered to learn. Well, in that spirit, it is appropriate to take advantage of an assault on publishing to learn a few things that have been happening in that arena in the recent past. First, there is OA. In its pure version, it refers to a publishing paradigm whereby research output is made available for anyone to access online without direct payment to the author(s) or the publisher and without contrived barriers. Generally, restrictions on copying or reuse are minimized or eliminated by applying an open license for copyright. There are different flavors of OA, which are distinguished by assigning different colors to the different implementations and which have to do with how publishers are compensated, or not. More on that later.
So, one might ask, how are publishing costs covered? Publishing leaves do not grow on trees any more than does money. What one sees today is a reliance on author fees generally called article processing charges (APCs), which are remitted, out of pocket, by authors, who may be compensated by subsidies, sponsorships, or as part of research grants. APCs vary considerably. Elsevier is a dominant player in scholarly publishing and tends to charge APCs on the higher end of the price spectrum. If one scans Elsevier’s 2023 price list [1], what it will show is prices ranging from US${\$}$500 up to US${\$}$20,000. For publications in fields of interest to IEEE, an eyeball average kind of number is US${\$}$2,000 but ranges up to US${\$}$5,000.
There is little doubt that OA is changing the publishing paradigm, and to the extent that it will dominate technical publishing, it will become a major disruptor. It has the real potential to put major publishers, including IEEE, out of business if a sustainable substitute business model cannot be found. Finding, or stumbling into, a suitable business model to replace the subscription model or other reader-pay models is the pivotal issue. Stumbling into a gold mine, by the way, works for me. We can always say, post hoc, that we knew where it was from the start.
So, is OA worth the grief? I think it may be. It may seem a bit quixotic, but it has got to be good for our profession and mankind to have research results available to one and all as quickly as they are produced. The march of science and engineering (S&E) would certainly quicken with it. And in fields such as medicine, public health, pharmaceuticals, potable water, food production, and the environment, there may even be an ethical component that compels it.
From an author’s perspective, one should expect greater dissemination of one’s work. From an editor’s perspective, one should expect more citations from wider circulation, which could lead to improved bibliometrics. Librarians, it would seem, could offer a more generous catalog, with less of a budget constraint caused by reader-pay models [2]. There is also some preliminary indication that the gender gap is less severe in OA journals than in subscription model ones [3]. From a publisher’s perspective, one might expect trepidation and anxiety about the ability to reproduce, year on year, the previous year’s subscription, licensing, overlength page charge, and pay-per-view charge revenue with APCs and read-and-publish agreements alone.
Then, there is the strong connection to the principles of open science [4], which strive to make all of science transparent, accessible, and shared via collaborative networks.
Admittedly, these are all hypothetical and maybe even Edenic. No one, for example, in IEEE has been able to show clear data that demonstrate superior bibliometrics for OA titles, but to be fair, many of these titles have not matured yet, and the OA promise may yet appear. After all, the mythical town of Brigadoon did appear out of the mist. A few studies have reported better bibliometrics [5], [6] for OA journals. But, in general, results are mixed, discipline dependent, and, on the whole, inconclusive [7].
In the ideal, OA sounds rather noble. Aside from the inconvenience of despoiling technical publishing for lack of a robust economic foundation, why should we not embrace it in full and forthwith? Well, that would require a dissertation to explain. However, let me recite a few issues that have been raised. Since most of the revenue earned by OA publications is from APCs charged to authors, OA publishers are incentivized to increase earnings by emphasizing quantity over quality, which may entail taking shortcuts in the peer-review process. How many people know that they may be signing up for poorer quality in exchange for OA? In the extreme, it expands the market for predatory and quasi-predatory journals, which really use little or no peer review at all in exchange for a fee.
Revenue per se is not a bad thing. IEEE uses dollars earned to serve the research community, including through the launch of new periodicals and conferences, and does so quite well. But readers are the primary arbiters of quality, and they are not in the driver’s seat in the OA model. Authors are. While authors do seek to publish in high-quality journals, they are not necessarily motivated to pay for professional society benefits in addition to the APC, nor are their funders. In fact, I can’t imagine that any nonacademic entities with a subscription would have any motivation at all to pay to publish. They are already reluctant to disclose trade secrets and can access content at a modest, and likely to decline, subscription cost and preprints at no cost. IEEE, in prior decades, has experienced significant attrition in nonacademic researchers willing to publish in its journals. OA may well be an accelerant in this trend.
So, what is one to think of all these vectors pushing and pulling in many directions? My personal view is that OA is an experiment in pursuit of a very laudable goal but that has and will continue to disrupt scientific publications in a significant way. Buckle your seatbelts!
Now, with all that background, stir into the mix Plan S and its sponsor cOAlition S. You can think of it as OA on steroids. It is an initiative, launched in 2018, by a group of research funding organizations from Europe and around the world, aimed at making full and immediate OA to scientific publications a reality. The plan calls for all scientific publications resulting from research funded by public and private grants to be made immediately available on the publisher’s website, without any embargo period. These are the same goals as OA. However, the plan also includes a number of other measures to support OA, such as encouraging the use of open source publishing platforms and providing funding for OA infrastructure. The real gut punch, though, is the requirement that all articles on research results funded by cOAlition S funders must be published in pure OA journals. Publishing in hybrid journals using an OA model is prohibited. Mirror publications that use the same editorial board as a sister journal with the same scope are also banned. The apparent motivation for this was disappointment that extant market forces were not sufficient to drive a satisfactory uptake in published OA articles. However, the explicit motivation for the requirement that funded articles be published in fully OA journals (and not hybrids) was to drive subscription journals out of business. cOAlition S was not coy about this, and it is notable that the U.S. Office of Science and Technology Policy guidance conspicuously eliminates this business model requirement.
As in the unhyperventilated version of OA, the sentiment is commendable. The means, however, are extreme and wrong on many levels. First, this is a threat to publishers, particularly the small and medium-sized ones. Traditional publishing models generate revenue through subscriptions and paywalls, which can be a significant source of income for publishers. Revenue for publishing services is vital. It pays for database and server maintenance, capital costs, governance, standards, protocols, archives, compliance, printing, mailing, copyediting, and tool development, i.e., all the accoutrements of modern high-quality publishing. Societies and Councils in IEEE benefit from these services and share in the revenue for their role in generating content. In contrast, OA models rely on funding from institutions, governments, and foundations, which can be a challenge to secure.
Second, cOAlition S itself will have to divert funds in establishing the administrative infrastructure to manage and coordinate the pieces of the huge enterprise it envisions, the skills for which it does not possess. This will diminish the funds going directly to research.
Third, OA has its very own chromodynamics, and no, not the quantum kind. From Platinum/Diamond OA, where neither readers nor authors pay, to Green OA, where self-archiving is used and access is mostly free, the OA landscape can be complex. In Gold OA, publishing is covered via APCs and read-and-publish agreements, and articles are licensed for sharing. In Hybrid OA, there is a mixture of reader pay and author pay. In Bronze OA, access is free only on the publisher’s website, and no reuse is allowed. In Black OA, publishers, such as Sci-Hub and others, bear the acquisition costs and make articles available to the public gratis.
In the imagined “utopian” world coerced by Plan S and dominated by OA, publishers are impoverished, there are few resources to support Platinum OA, Hybrid OA is redundant, and Black OA no longer makes sense. That is to say, funding of virtually all peer-reviewed papers is squeezed to cOAlition S funders, foundations, and governments, most of whom, in practical terms, mandate either Gold or Green OA alone. Because subscribers are eliminated as a funding source, there is not likely to be enough money in the system to convert the total demand to OA. The pool of funders and funding is greatly reduced, which means that funders decide which research to support, which, in turn, means they decide what gets published. What Plan S is promoting, though probably unintentionally, is a publishing autocracy.
If I were to distill all that is wrong with Plan S into one major blind spot, it is this. It fails to look at technical publishing as a system. There are publishers, professional societies, libraries, researchers, teachers, authors, students, consumers, foundations, trusts, and governments, all of whom contribute and benefit from the enterprise. All the stakeholders work in concert with one another. However, cOAlition S is so focused on making the research it funds available to all immediately that it stiff-arms its partners and throws the system out of equilibrium. To mix metaphors, it acts as a lumbering oaf in the proverbial china shop, knocking over whatever is in its way.
Thinking of technical publishing as a system, it is easy to see what will happen. In the scenario where OA is dominant, it will be the total funding from cOAlition S, foundations, and governments that will pace the publication rate, as explained in the preceding. It is not likely that researchers will want to use more than 10% or so of their funding for publication costs. Based on the Scopus database [8], with 3.8 million S&E papers published annually, and assuming an average US$2,000 APC, funders will need to provide research funding to the tune of ∼US$75 billion (a lower bound) every year to cover an estimated US$7 billion in publication cost. To that should be added whatever will be used to subsidize the publishing of articles from developing economies. Furthermore, the growth of technical publishing will be tied to the growth in total funder revenue. It is ironic then that OA may evolve to mean diminishing access.
Not all disciplines will benefit equally. In the United States, for example, health, biological, and biomedical sciences published 51% of all S&E papers in 2020 [9], which accounted for more than all the other papers in all other disciplines combined. All engineering fields combined accounted for 11%. A similar profile was replicated in Europe and Japan. This suggests that as the constrained funding is depleted, there will be fewer opportunities for OA articles in computer science, information science, physics, chemistry, social science, and materials science. Forget arts and humanities and the less well-off societies.
Finally, this systems view would be wholly remiss without discussing the elephant in the room, which is the actions anticipated by the People’s Republic of China (PRC). Between 2018 and 2020, China published 23.4% of the world’s scientific papers, eclipsing the United States [10]. As compiled by Clarivate, China now publishes the highest number of scientific research papers yearly, followed by the United States and Germany. Furthermore, Chinese research accounts for 27.2% of the world’s top 1% of most frequently cited papers. Also in 2020, China’s Ministry of Science and Technology and Ministry of Education released two policy documents, which defined guidelines for academic publishing and research evaluation that have begun to rock the scientific publishing world. Specifically, in basic science, a limited number of “representative works” will be used for assessing research quality. “Individual scholars can choose no more than five representative papers each year, one third of which must be published in Chinese journals, if national grants or awards are in pursuit” [11]. In this scheme, bibliometrics become almost irrelevant.
The PRC has a centrally planned economy, and virtually all research funding comes directly or indirectly from the government. Funding priority, then, will be determined by national interests and not necessarily by what is best for the global health of science and technology. Clearly, how these priorities are set will have a nontrivial impact on OA funding internationally. Of paramount importance will be incremental funding beyond what has already been reckoned in the OA funding pool described previously. A joint policy document released by the Ministry of Science and Technology and the National Natural Science Foundation of China officially forbids using national grants to encourage paper production [11]. This may mean that the PRC will not be a large funder of OA publishing.
Undoubtedly, there will be, and indeed already has been, a significant decrease in submissions from China to Scientific Citation Index (SCI) publications. Since almost 30% of OA SCI articles have originated from Chinese institutions, this is where the impact will be the greatest.
But all is not gloom. There are some upsides to this story. In addition to the potential for incremental funding for OA, the cognizant ministries have provided guidance to researchers to grow their international presence. This provides a thrust that may further mitigate the mayhem in English language journals, provided they are top-tier ones. Of course, one must pretend that there is no contradiction with the guideline that downplays journal bibliometrics and simultaneously seeks global recognition for its articles.
Finally, the National People’s Congress, in March of this year, approved a plan that will refocus China’s Ministry of Science and Technology on key efforts to meet top national priorities. It also created a powerful Central Science and Technology Commission intended to enforce consistent policies across government agencies—and hold them accountable for achieving their objectives [12]. It is unknown how this shake-up will affect previous guidelines. They are, without a doubt, the wild card in this game.
In summary, OA is a growing movement in publishing. It has a group of followers whose exuberance would make Beyoncé envious. Its primary goal is meritorious and compelling, but it has been and continues to be disruptive of the publishing world. You can judge whether that is a good or a bad thing. The looming and unanswered question is whether its underlying business model is sustainable. Plan S, by contrast, takes an altruistic and potentially benevolent idea and morphs it into a coercive autocratic scheme that discounts market forces and may well produce harmful effects for researchers in developing countries, professional societies, and scientific publishing overall. When it comes to China’s growing impact on scientific publishing, only time will tell whether the country’s nationalistic version of OA will nurture it or starve it.
Douglas Verret (verret@ieee.org) is a Life Fellow of IEEE.
[1] “Article Publishing Charge (APC) price list.” Elsevier. Accessed: Mar. 17, 2023. [Online] . Available: https://www.elsevier.com/books-and-journals/journal-pricing/print-price-list
[2] S. Guy and S. Carroll, “Integrating OA collections in a small academic library,” Comput. Libraries, vol. 42, no. 5, pp. 14–17, Jun. 2022.
[3] Y. Zhu, “Who support open access publishing? Gender, discipline, seniority and other factors associated with academics’ OA practice,” Scientometrics, vol. 111, no. 2, pp. 557–579, Mar. 2017, doi: 10.1007/s11192-017-2316-z.
[4] “Open concepts and principles.” Open Science Training Handbook. Accessed: Mar. 17, 2023. [Online] . Available: https://open-science-training-handbook.gitbook.io/book/02opensciencebasics/01openconceptsandprinciples
[5] H. Piwowar et al., “The state of OA: A large-scale analysis of the prevalence and impact of open access articles,” PeerJ, vol. 6, Feb. 2018, Art. no. e4375, doi: 10.7717/peerj.4375.
[6] E. C. McKiernan et al., “How open science helps researchers succeed,” Elife, vol. 5, Jul. 2016, Art. no. e16800, doi: 10.7554/eLife.16800.
[7] A. Langham-Putrow, C. Bakker, and A. Riegelman, “Is the open access citation advantage real? A systematic review of the citation of open access and subscription-based articles,” PLoS One, vol. 16, no. 6, Jun. 2021, Art. no. e0253129, doi: 10.1371/journal.pone.0253129.
[8] “Publications output: U.S. trends and international comparisons,” Nat. Sci. Found., Alexandria, VA, USA, Dec. 2019. [Online] . Available: https://ncses.nsf.gov/pubs/nsb20206/executive-summary
[9] “Publications output: U.S. trends and international comparisons,” Nat. Sci. Found., Alexandria, VA, USA, Oct. 2021. [Online] . Available: https://ncses.nsf.gov/pubs/nsb20214/publication-output-by-country-region-or-economy-and-scientific-field
[10] D. Lu, “China overtakes the US in scientific research output,” Guardian, Aug. 2022. Accessed: Mar. 19, 2023. [Online] . Available: https://www.theguardian.com/world/2022/aug/11/china-overtakes-the-us-in-scientific-research-output
[11] S. Qingyang Li, “The end of publish or perish? China’s new policy on research evaluation,” Max Planck Inst. Hist. Sci., Berlin, Germany, Nov. 2020. [Online] . Available: https://www.mpiwg-berlin.mpg.de/observations/1/end-publish-or-perish-chinas-new-policy-research-evaluation
[12] D. Normile, “China rolls out ‘radical’ change to its research enterprise,” Science, Mar. 2023. Accessed: Mar. 17, 2023. [Online] . Available: https://www.science.org/content/article/china-rolls-out-radical-change-its-research-enterprise
Digital Object Identifier 10.1109/MED.2023.3282436
Date of current version: 15 September 2023