Out of Reach

Out of Reach

1

At long last, Elsevier and the German DEAL Consortium have reached a deal. This is a milestone for Elsevier and its Chief Executive Officer, Kumsal Bayazit, who assumed the post shortly after discussions with the consortium collapsed in 2018. In a note of symmetry, Wiley’s deal with DEAL was announced (in great style at APE 2019) by Judy Verses, who was then Executive Vice President for Research at Wiley. Verses has since moved from Wiley to Elsevier, although her role (if any) in negotiations between Elsevier and the consortium was not mentioned in the press release. 
 
The agreement is notable for several reasons. For one thing, it is a much better deal for Germany’s research institutions than the inaugural 2019 agreement with Wiley. Under the agreement with Wiley, the consortium pays €2,750 per article published in a Wiley Hybrid journal. The consortium will pay Elsevier only €2,550 per article for publication in the Dutch publisher’s Hybrid titles. This is a remarkable difference when one considers inflation. To pay Elsevier less in 2024 than they paid Wiley in 2019 is a negotiating coup for DEAL. Publication in Gold OA titles will be at a 20% discount to list price (the same discount as with the Wiley deal). It is also, importantly (and unlike the deals with Wiley or Springer Nature), an opt-in agreement for DEAL Consortium members – meaning that German institutions may choose whether to participate. The extent of participation – particularly among the U15 research-intensive universities, which have previously raised concerns about the costs associated with these deals – will determine the success of the arrangement for both parties. 
 
Perhaps the most notable aspect of the agreement, however, is that Elsevier has brokered two tiers of pricing. Open access (OA) publication in most Hybrid titles will cost €2,550. However, publication in any of the Hybrid journals in Elsevier’s flagship Cell or Lancet portfolios will cost €6,450. This two-tier pricing structure offers a way forward for publishers seeking to maintain high-quality, selective titles. A per-article fee of €2,550 is geared toward high-volume titles that accept a large proportion of the papers they review and hence have lower costs per paper as compared with selective titles where all costs are borne by relatively few accepted articles. This level of per-article fee is simply not a viable price point for most selective titles. 
 
Elsevier publishes a great number of selective titles outside the Lancet and Cell families, many of which are owned by societies and are top-tier journals, peers of journals bearing the Lancet and Cell imprimaturs. Publishing in these society journals will be a bargain for German researchers at €2,550 per article. But what may be a bargain for DEAL is a problem for society journals. €6,450 is a per-article fee at which many selective titles might be viable; €2,550 is a fee at which many will be unsustainable. 
 
In the short term, societies may have some buffer as the terms between Elsevier and DEAL are not necessarily mirrored in royalties paid to the societies – some may receive more revenue than implied by the €2,550 price point due to preexisting Elsevier royalty allocation formulas. Moreover, many societies have guaranteed minimum royalty payments, so if Elsevier strikes deals that result in lower revenues to a given society journal, Elsevier may have to eat that difference – at least in the short term. In the longer term, however, new publishing contracts will likely specify royalty allocations that reflect the market value of the society’s journals – which is based on the revenues that a given journal or set of journals brings to the publisher. 
 
Imagine a near-future OA world where, across many transformative agreements, 500 papers are published in a highly selective, flagship society journal. There is a difference of nearly €2 million in annual revenue to the journal if those articles are published at a price point of €6,450 rather than €2,550. Elsevier understands the need for a higher price for premier titles as it has negotiated one for its own. The DEAL consortium understands the value of publishing in premier titles because they have agreed to a higher price point for Lancet and Cell titles. And yet, premier society titles are not included in the same tier with Lancet and Cell. 
 
Societies that publish with larger publishers are already at a great disadvantage when it comes to transformative agreements. When a publisher signs a transformative agreement, it has analyzed the terms of the deal to ensure it makes sense for the entire portfolio. If it doesn’t make sense, there is no deal or the parties renegotiate. But what benefits the portfolio as a whole may not benefit every individual journal within the portfolio. When a publisher signs a transformative agreement with a university consortium, it typically replaces any subscriptions a society may have had with consortium members. Let’s say Society Journal A has 50 subscriptions across German research institutions each valued at €2,500 and authors associated with DEAL institutions only publish 20 papers annually in the same journal. If the society’s publisher strikes a transformative agreement with DEAL at the €2,550 per-article price point, the society journal will be trading €125,000 in subscription revenue for €51,000 in OA revenue. Conversely, if Society Journal B has only 20 subscriptions (at the same €2,500 value) but receives 50 papers from DEAL institutions, it will be trading €50,000 in subscription revenue for €127,500 in OA revenue. This hypothetical transformative agreement would be a terrible deal for Society Journal A, but a great deal for Society Journal B. Unfortunately, societies cannot typically opt in or out of specific deals, as otherwise no deal would ever get done. This means that many deals will get done that are not good for at least some of the societies involved. A tiered pricing system for OA fees would go some distance to remediating the transformative agreement alignment problem for many societies with selective journals that have broad subscriber bases – titles that more closely resemble Society Journal A. 
 
The Elsevier–DEAL agreement is a missed opportunity to bring sustainable economics to selective society journals. Its two-tiered pricing structure tantalizingly points to a potential method of making transformative agreements more workable for many society journals even as it places that structure just out of reach.

Atomization

2

The latest infrastructure acquisition by a publisher was not one we at The Brief expected. While it is no surprise that ChronosHub was acquired, the purchase by the American Chemical Society (ACS) suggests a new direction for the society. The ChronosHub acquisition is a piece of general publishing infrastructure, rather than a chemistry-oriented property like ACS’s CAS database, the ChemRxiv preprint server, or at least their current use of SciMeetings as a repository strictly for ACS meeting materials. Whether ChronosHub will evolve to serve the specific needs of ACS authors or whether ACS is expanding their business into the broader publishing services provider realm is yet to be determined.
 
For those unfamiliar with the platform, ChronosHub is a manuscript submission platform of sorts. Its main function is to streamline the author submission process, particularly for OA business models. For example, it can be used to identify the transformative agreements under which an author may be covered. It also handles license management and article processing charge (APC) payments. In addition, ChronosHub automatically extracts article metadata to simplify and expedite submission. The platform is not just for publishers, however, it also provides compliance reporting for funders, as well as reporting and management tools for institutions. 
 
The acquisition of ChronosHub signals the maturation of OA business models for ACS and, by extension, other publishers. The functions it provides are becoming mission critical for supporting transformative agreements and managing a diverse array of funder requirements and author preferences. 
 
ChronosHub also serves as an example of a growing trend: the atomization of scholarly publishing infrastructure. Where once publishers expected submission systems to handle everything from upload of a submission and its related metadata through peer review, ChronosHub (along with Wiley’s Research Exchange) suggests elements of the research publication workflow might be broken up into smaller packets handled by specialized software. Examples of workflow that might be (or already are) atomized include research integrity checks (plagiarism, image manipulation), OA payment processing, reviewer recommendation, institutional agreement verification, and funder compliance – to name just a few. This allows both better workflow customization and the ability to switch out components of a workflow without impacting other parts. It also signals more holistic approaches to identity management. As identity data becomes more interoperable, marketing becomes far more effective and user experiences improve. Siloed databases of authors and reviewers (who are also society members and meeting attendees) give way to true enterprise management of identity. 
 
Unlike most other recent infrastructure acquisitions, the formerly for-profit ChronosHub is now part of a larger not-for-profit organization. This may sit better with those intent on moving away from commercially provided services toward “community-owned” (or at least, not-for-profit) infrastructure. 

Article Development Charge

3

ACS had a busy month in September. Beyond their acquisition of ChronosHub (see Item #2 above), the society brought a new TLA (three-letter abbreviation) to our collective vocabulary by introducing the ADC, or article development charge. Paying the $2,500 ADC grants authors permission to deposit the accepted manuscript version of their paper in a repository with a CC BY license. While on the surface the ADC resembles eLife’s publication fee (both are charged to the author once the manuscript passes editorial review and is sent out to peer reviewers), there is a fundamental difference: eLife authors are guaranteed publication in the journal, where ACS authors face potential rejection of their articles with no refund of their ADC payment. 
 
The ADC fits neatly into ACS’s menu of OA options, which include different price points for CC BY and CC BY-NC-ND licenses. The ADC is effectively a fee for zero-embargo Green OA as compared with 12-month embargo Green OA. By framing the APC as a “development” charge, however, the effort seems to be aimed at authors with a funder or institutional mandate for zero embargo but who seek to avoid payment (particularly through the Rights Retention Strategy). If an author declares their intention to apply the Rights Retention Strategy, peer review in an ACS journal will require ADC payment. ACS is making a clear statement here, that there are costs involved in the review process and that authors will be expected to cover those costs, by allowing ACS to sell subscription access to their work, by paying an APC for Gold OA, or by paying an ADC for Green OA. 
 
Sarah Tegan, ACS Senior Vice President and Chief Publishing Officer, informed The Brief that “more than 90% of authors whose funders require open access have a funded mechanism to make the version of record freely available” and that the goal here is to provide a sustainable publishing mechanism for the small number who don’t have these funds. ACS estimates this small number authors that could utilize the option to be “in the low hundreds” annually. Tegen confirms that the ADC “does not guarantee a manuscript will be accepted,” and adds that if a manuscript is not acceptable in its current form and is later re-submitted, no new fee will apply. Further, “the ADC will be credited if the manuscript is subsequently submitted to another ACS journal.”
 
That said, the combination of a required ADC payment and the uncertainty of publication makes the Rights Retention Strategy the riskiest choice, and likely the least attractive route for authors.

5th Anniversary

4

According to popular convention (or at least Martha Stewart) the fifth is the “wood” anniversary, and although there is not yet a “wood” variety of OA (paywalled online but free to read on paper perhaps?), we at The Brief would like to extend our congratulations to cOAlition S for reaching this milestone. The organization’s Plan S has been, without a doubt, a paradigm-shifting success. Although Plan S still only applies to some 4% of the literature (around 208,000 articles in 2022 according to Dimensions, an inter-linked research information system provided by Digital Science), its reach has been far wider, catalyzing a community-wide movement to OA. In 2018, the year Plan S launched, Dimensions shows 37% of the literature being published Gold OA or Hybrid, and in 2023 to date, that number has climbed to 47%.
 
With that success has come the clear realization that the long-standing concerns about both the inequities and the sustainability of the APC model for OA have proven accurate. The OA advocacy community, including cOAlition S, has shifted to an increasingly anti-APC position, a stance that would have resulted in them being labeled anti-OA in the recent past. Some disagree and are instead calling for the community to double down on the APC model, sometimes repeatedly across media outlets. To their credit, cOAlition S has exhibited a willingness to evolve, try different approaches, abandon those that have failed, and move on to new ideas. Although we find attempts to rewrite history, such as a claim that “From the start, Plan S adopted an agnostic approach to Open Access routes” a bit disingenuous given, for example, their early directives disallowing Hybrid and mirror journals.
 
The question for the next five years is whether cOAlition S will have a second act. The group is winding down support for transformative agreements, but it is unlikely this will significantly slow the growth of these arrangements as a dominant business model. The market has largely shifted toward volume-based approaches, but this shift (although accelerated by Plan S) was built on nearly 20 years of experimentation and incremental efforts by publishers. cOAlition S is now seeking a new paradigm shift, one away from the APC and toward more equitable models. This may prove a more daunting task and it is unclear whether the small number of organizations chosen by cOAlition S to approach the question offer a broad enough set of perspectives to build a new set of global-scale models to rapidly and successfully restructure scholarly communication.

People

5

Patricia Brennan’s leadership of the National Library of Medicine comes to an end this month with her retirement. Stephen Sherry will serve as Acting Director until a replacement is named.
 
Long-standing Physics World Production Editor and Product Manager Dens Milne passed away unexpectedly in July.
 
Silverchair has named Will Schweitzer CEO. Cofounder and previous CEO Thane Kerner has been named Chair of the Board of Directors.
 
Springer Nature’s Chief Financial Officer, Ulrich Vest will retire at the end of 2023 and be succeeded in the role by Alexandra Dambeck.
 
Sir Ian Wilmut, leader of the team that cloned Dolly the Sheep in 1996, died earlier this month. 

Briefly Noted

6

In other cOAlition S news, the organization has recently released a new draft on “Developing a globally fair pricing model for academic publishing.” The framework essentially divides the world up into four economic tiers and suggests that APCs be set at 100% of list price for the top tier, 80% for the second, 50% the third, and 20% the fourth. This may not prove particularly helpful in the long run for several reasons. Historically, as more and more journals have been accessed through Big Deal subscriptions, the list prices for those journals have become essentially meaningless because so few pay them. So, too, will list prices for APCs as transformative agreements become increasingly the norm. The scheme presented in the draft also creates perverse economic incentives for journals to focus on and favors accepting more papers from authors from the top tier, as those will bring in five times more revenue than papers from the fourth tier.
 
The declines in publication volume at both Frontiers and MDPI discussed in last month’s The Brief have been further quantified by Christos Petrou, who suggests that the two companies differ in how they manage their reputations.
 
Frontiers is further bolstering their efforts to weed out paper mill activity through the implementation of a new policy on authorship changes. One route for paper mills is to sell authorship on legitimate papers when they’re close to or have reached acceptance. Frontiers’ new rules, which should be adopted by all publishers if they’re not already in place, require “exceptional circumstances” and “in-depth assessment” for any new authors to be added to a paper during the review process. Frontiers has retracted 38 research papers that were linked to the practice.
 
Frontiers has also signed a “flat fee” publishing agreement with the University of California (UC) for authors to publish in 20 of its humanities and social sciences journals. This is a departure from UC’s usual multi-payer model in that 100% of the author’s APC will be paid by the school rather than the author sharing the expense (likely necessary given the low levels of funding for humanities and social sciences). No financial details have been made available as to the amounts being paid to Frontiers nor the expected quantity of articles.
 
This month The Wall Street Journal profiled the people journal editors greatly admire but hate to hear from, the volunteer detectives rooting out errors and fraud in published research. The article focuses on the work of the Data Colada group, currently facing a significant lawsuit for their review of the work of a Harvard professor. Science Editor in Chief Holden Thorpe is quoted in the article and wrote his own commentary, which links to the 2022 editorial he wrote proposing the eminently sensible notion that journals focus solely on correcting the scientific record while sidestepping the lengthy (and sometimes costly) accusations of fraud or misconduct often involved in the retraction process.
 
The 2017 lawsuit filed by Elsevier and ACS against ResearchGate has officially come to a close as the parties involved agreed a confidential settlement. Last year’s verdict essentially gave the publishers what they wanted, requiring ResearchGate to check papers for rights information at the point of upload, rather than waiting for an infringement complaint to come in from the copyright holder of the work. This appears to be enough of a victory for the publishers to end the litigation rather than carrying on through the appeals process. With this weight lifted, ResearchGate is now better able to continue its progress toward attempting to become the research community’s preferred “supercontinent” source for publications.
 
Don’t fret though – fans of publisher/copyright disputes can now turn their attention to a new lawsuit filed by Cengage Learning, Macmillan Learning, McGraw Hill, and Pearson Education against Library Genesis (LibGen), the “shadow library” behind massive amounts of piracy. It’s unclear whether this lawsuit will have any more impact on LibGen’s activities than did the case Elsevier won against the anonymous group (and Sci-Hub) in 2015.
 
The OA Book Usage Data Trust (OAEBUDT) – which is trying to solve the challenge of aggregating and curating OA book usage data by enabling community-governed sharing of quality, interoperable data – is developing an OA Book Usage Data Exchange and Stewardship Rulebook. This Rulebook specifies the principles that generate trust through OAEBUDT participation, usage data management, processing, and provision. OAEBUDT is inviting feedback on the draft principles to ensure they are developed in line with the community: you can participate by signing up to take part in focus groups or providing your feedback via online surveys. C&E will be conducting further research in support of the OAEBUDT, following on from our work to develop the OA Books Supply Chain Mapping Report in 2021.
 
Crossref has purchased the Retraction Watch database of retracted articles, as well as agreeing to fund the ongoing development of the database for the next five years (this is effectively a metadata enrichment deal). Crossref will make the information a public resource and will link it to their DOI (digital object identifier) system. The fact that Crossref is enriching DOI records with third-party information can be viewed as a significant failing on the part of publishers to properly update their DOI records when an article is retracted.
 
Another anniversary was reached this month, as Octopus completed its first year of existence. While the platform hopes to revolutionize the way research is published, it is suffering from the same problems facing any new publication venue – convincing authors to use it. A cursory examination of the Octopus website at first makes it seem robust, with 5,425 papers listed. However, a deeper dive – difficult, given the limited search capacities on the platform – shows that nearly all of these “papers” are “research problems” chosen and restated by Octopus staff from abstracts in articles published elsewhere. The 28 non-research-problem articles appear to be fragments of a small number of studies from three research groups.
 
Say what you will about the APCs for Nature journals, demand from authors remains high. FinELib, a consortium of Finnish universities, research institutions, and public libraries reports that the full allotment of Nature journal articles in their OA deal has been used up 15 months earlier than expected.
 
In other OA news, De Gruyter announced that, over the next five years, around 85% of its subscription journals will move to a Subscribe to Open model. 
 
Artificial intelligence (AI) continues to dominate the headspace of publishers and academicians alike, and this month provided no letup in the stream of AI news and analysis. Two groups of renowned authors (the first including Michael Chabon and the second George R.R. Martin, Jodi Picoult, and John Grisham) have filed lawsuits against OpenAI claiming that ChatGPT has illegally ingested and is reusing their content. OpenAI responded to an earlier pair of author lawsuits by asking a judge to dismiss most of the claims because the authors “misconceive the scope of copyright” and stating that ChatGPT outputs are new, non-infringing products. Google has made changes to its search algorithms to promote AI-generated content in search results, which has greatly angered human authors who have seen their sites tumble down the search rankings. Two pieces we at The Brief found particularly thought-provoking are Benedict Evans’ article on “Generative AI and intellectual property” and the view of AI breaking the “bargain” of the web from Kali Hays and Alistair Barr.
 
And finally, a cautionary tale on crowdsourcing and metrics has emerged from the movie rating site Rotten Tomatoes.

***
86% of Diamond Open Access journals publish fewer than 50 articles per year and therefore lack the necessary scale to make the difference. —Robert-Jan Smits (Viewpoint, Science | Business)