Diamond

Diamond

1

We were reminded of Ian Betteridge, a British technology journalist, when we read the Times Higher Education news story Can the EU make a ‘parallel publishing system’ work? Betteridge’s law of headlines states: “Any headline that ends in a question mark can be answered by the word ‘no’.” 

For readers who don’t closely follow communiqués from the European Council (EC), here’s a quick summary. On May 4th the EC issued Draft Council Conclusions on High-Quality, Transparent, Open, Trustworthy and Equitable Scholarly Publishing, which were formally approved on May 23rd. This document, which constitutes a set of political viewpoints and is not legally binding, proposes that Member States and the Commission should “step up support” for the development of a not-for-profit publishing platform that is free for both authors and readers (Diamond OA). In other words, the EC is recommending a scaling up of a Diamond OA model that would disintermediate scholarly (and especially commercial) publishers.

The European Commission dipped a toe into the water when it launched Open Research Europe (ORE) in April 2021. The platform is powered by F1000Research and is only available to researchers who are funded by Horizon 2020, Horizon Europe, and Euratom projects. Horizon Europe (the successor to Horizon 2020) is the European Union (EU)’s key funding program for research and innovation with a budget of €95.5 billion from 2021 to 2027. 

ORE has not been a runaway success. In 2022 only 185 articles (of which 110 have passed peer review) were published on the platform. The total funding pot for ORE is €5.8 million for the period 2021–2024, but it seems likely that costs will be lower than the funding given the small number of articles that have been published. 

To put the number (185 articles) into context, according to Dimensions (an inter-linked research information system provided by Digital Science (https://www.dimensions.ai), the European Commission and the European Research Council were acknowledged in more than 116,000 articles in 2022. As a comparison with the ORE platform, Nature Communications published over 700 articles that acknowledge the two funders in the same period. Based on the citation characteristics of the articles published in ORE, it looks as though researchers are choosing to publish lower-impact work there. Another Diamond OA venue, Wellcome Open Research, celebrated its sixth birthday last month; it published 363 papers in 2021 and 315 in 2022. Dimensions lists 12,844 and 11,044 articles with Wellcome funding in those two years respectively, so this Diamond outlet appears to be serving as the outlet for around 2.8% of funded articles from the organization.

Last year, the European Commission’s Directorate-General for Research and Innovation commissioned Rob Johnson from Research Consulting, “to provide advice on the organisational and financing model(s) that would operationalise Open Research Europe (ORE) as a collective publishing enterprise as of 2026.” Johnson’s report, which was published in September 2022, provides an excellent assessment of some of the challenges – both practical and cultural – that a large-scale, open-source publishing platform would need to overcome. The report models the annual financial investment required to publish a relatively modest number, between 750 and 4,500 (0.5%–3%), of 150,000 European articles each year. If a per-article cost of €2,000 is assumed, the “upper-case” annual operating cost would be €9 million.

This price point is comparable with Scientific Reports (€2,090) and PLOS ONE (€1,800), which makes the case for a new service that replicates existing services less than compelling when based solely on economic grounds. The argument becomes stronger when focused on creating an equitable publishing environment, in which neither author nor reader needs to pay. That said, researchers funded by Horizon Europe do not currently need to pay to publish in any journal themselves (as their funding agency or institution foots the bill). 

Another argument for Diamond OA is to cut out commercial publishers, as Johan Rooryck, the Executive Director of cOAlition S, laid out in an essay published this month (May). This is a particularly vexing objective for the Executive Director of cOAlition S (which represents many national funding agencies in Europe), given many of the world’s largest scholarly publishers are European companies – Elsevier, Springer Nature, Informa/Taylor & Francis, Frontiers, MDPI, De Gruyter, Emerald, and Brill to name just a few. It is unclear why European officials are actively working to undermine European businesses. 

Putting that to one side, some researchers, politicians, and policymakers may not fully appreciate just how much investment would be required to make a Diamond OA network a reality. Kumsal Bayazit, the CEO of Elsevier, told investors at the end of last year that Elsevier has 9,000 staff, of which 2,500 are technologists, and that RELX spends $1.6 billion a year on technology. European agencies would need to invest a lot more than €9 million a year to create a Diamond OA utopia that provides as much value as existing scholarly publishers create each and every day.

Research Policy Impact

2

To date, US and European researchers have largely been shielded from the direct impacts of open and public access funder policies. The National Institutes of Health (NIH)’s 2008 policy, along with the subsequent 2013 Holdren Memo, were both successful because publishers voluntarily took on the necessary work of depositing papers in agency repositories on authors’ behalf. Even Plan S has bent over backward to reduce the impact on researchers, encouraging “transformative” journals and agreements to avoid researchers (and perhaps the funding agencies themselves?) having to directly pay article processing charges (APCs). It has attempted to adapt the policy to not restrict author choice of publication venue (which is challenging when the central thrust of the policy is to restrict author choice of publication venue). But with many of Plan S’s accommodations sunsetting in 2024, and with the upcoming implementation of the US federal funders’ Nelson Memo, researchers are just beginning to feel their share of the burden of open science.
 
This month saw two journal editorial boards resign in protest over publisher OA policies. Much of the editorial board of Wiley’s Journal of Political Philosophy walked out in protest over the firing of the journal’s editor-in-chief because of his refusal to increase the number of articles accepted each year. Publication volume is key to the APC OA business model and Wiley’s strategy is clearly a response to funders’ preferences for OA. The editorial team behind Elsevier’s NeuroImage similarly quit in protest, this time in response to perceptions that the journal’s APC charges were too high, and plan to start their own high-prestige title with a lower publication fee at MIT Press.
 
Editorial board mutinies are not new, despite attracting overblown media coverage. There’s even a wiki page listing walkouts going back to 1989. The end result of such mutinies is rarely a death sentence for the journal in question. More commonly, they result in the launch of a new journal with seemingly little direct impact on the previous title. Case in point, the much ballyhooed resignations at the Elsevier journal Lingua and the subsequent launch of Glossa back in 2015. At this point, both journals seem to be doing well, with Lingua showing some growth in article volume and a relatively steady journal impact factor, if also a decline in category ranking. But where some past rebellions were about moving journals away from a subscription model toward OA, these latest ones have been about the mechanics of OA and the steps that publishers are taking to respond to the changing policy landscape.
 
A journal article and a new study, both out this month, suggest that this is just the tip of the iceberg, as researchers, who have largely been left out of open research policy debates, will increasingly feel the brunt of new requirements. Thomas Hostler writes that “the current discourse promoting open research fails to engage sufficiently with how additional workload impacts the practicalities of academic labour,” and suggests that the increased workload and administrative burdens of open science will lead to further burnout and increased pressure on researchers. 
 
The Council on Government Relations (COGR), an association of US research institutions, further emphasizes the impact of new open science policies through a survey report looking at the Cost of Complying with the new NIH Data Management and Sharing Policy. The report suggests that the annual costs for NIH compliance will be around $1.4 million for each large to mid-sized research institution, and a little over $1 million per year for smaller institutions. These numbers seem comically low and are based on a methodology in which dollar amounts are assigned to the level of difficulty/effort survey respondents state that each activity will require. There’s no adjustment for scale, the number of researchers at an institution within a size category, or types of research performed, which can vary enormously in terms of the quantity of data collected. Furthermore, there’s no estimate for ongoing costs to maintain an ever-growing backlog of data, which must be perpetually made available and kept up-to-date as technologies evolve over time.
 
Regardless, these figures represent only the NIH’s policy; similar requirements will soon be in place for all US federal funders. COGR’s conclusions are worrisome, as there is no new money in the system to pay for this work. Money needed for compliance will come out of existing research and library budgets. Time and effort required for compliance means more administrative work for researchers and less research being done. Perhaps most troubling is the suggestion that these burdens will end up locking out smaller research institutions from the federal research ecosystem as the cost of participation becomes prohibitive.
 
For a set of policies aimed at driving research equity, concentrating federal funding at larger and wealthier institutions seems an unwelcome consequence. With funders in the EU reassessing the negative consequences of their push for APC Gold OA (see item #1 above), librarians unhappy with the rising costs, and publishers concerned about sustainability, researchers will soon join the party in experiencing the sacrifices required to implement open research.

A Profound Embarrassment

3

Public trust in scientific research, which is in decline for various reasons, took another blow this month due to widespread publicity offered to a methodologically flawed preprint offering specious claims that one-third of neuroscience papers and one-quarter of medical papers were fraudulent. University of Washington biologist Carl Bergstrom offers a fairly succinct explanation of why the preprint’s claims are both nonsensical and have “racist consequences.” The study relies on two criteria to label a paper as fraudulent (neither of which involve evaluating the paper itself):

1. An author on the paper uses a private (non-institutional) email address and/or has an affiliation with a hospital
2. There are no international coauthors on the paper

That’s it. That’s the entirety of the analysis. Perhaps it’s not shocking to then discover that the preprint notes that their “detection tool” has a 44% false-alarm rate when applied to known non-fraudulent papers, which Bergstrom characterizes as, “That’s not a detector, it’s a coinflip.” 

The preprint might have faded quietly into well-deserved obscurity had not Science published a news story on the preprint, ironically quoting one of its authors as saying, “It’s just too hard to believe.” Even worse, the preprint hit the mainstream media via the Financial Times. All this for a study that Bergstrom suggests is “a profound embarrassment to everyone involved with the preprint and Science story alike.”

Dealmaking

4

Sage has acquired the catalog of business and management textbook publisher Chicago Business Press.

Taylor & Francis has acquired Stylus Publishing, a US-based independent publisher specializing in Education and Teaching in Higher Education.

Thomas Telford Ltd, the commercial arm of the Institution of Civil Engineers, has sold the majority of its book catalog, and all journal titles and related publishing assets, to Emerald Publishing.

People

5

Betsy Donohue is joining Silverchair as the company’s Senior Vice President for Business Development.

Henry Levy has been named President, Life Sciences & Healthcare at Clarivate.

Jeff MacKie-Mason, a leader in the drive to open access and experimentation with new models for library–publisher agreements, has announced his retirement as UC Berkeley’s University Librarian.

Antonia Seymour, CEO of IOP Publishing, has been elected to the position of President of the Publishers Association. 

Dean Smith, Adjunct Professor of Publishing in the College of Professional Studies at George Washington University and the Director of Duke University Press, has received GW’s College of Professional Studies Faculty Excellence Award.

Emerald has appointed Terri Teleen to a newly created position, President, North America.

Project MUSE has done some structural reorganizing and has named Kelley Squazzo as Director of Library & Publisher Partnerships, tapped Melanie Schaffner to lead a newly created department in the position of Director of Communication, Marketing, and Engagement, and welcomes back Lance Tieperman as Business Operations Manager.

Briefly Noted

6

As we suggested in March’s issue of The Brief, being delisted from Clarivate’s Journal Citation Reports and losing one’s Journal Impact Factor can be a death knell for a journal. This has proven to be the case for at least four of the Hindawi journals found to have been infiltrated by rampant paper mill activity, as the publisher has announced their shuttering. Given the reputational damage caused by this scandal, it is likely an easier task to start new journals (or expand other existing journals) than to salvage those with tarnished reputations.

The havoc wreaked by ChatGPT is not just limited to professors struggling to detect artificial intelligence (AI)–written homework assignments, it is now causing “a stock-market ruckus” as education companies are suddenly seeing their products displaced. After a gloomy forecast and a drop in subscribers, Chegg, Inc., saw an overnight sell-off taking nearly $1 billion off their market valuation. The panic then spread to Pearson, as investors’ fears of AI’s impact drove the company’s shares down 15%.

34 US employees of the Public Library of Science (PLOS) have put out a call to unionize. PLOS has asked for more time to consider the impact on their global workforce.

eLife has offered a three-month progress report on its new “reviewed preprint” publication model. They are currently rejecting 73% of submissions without sending them out for peer review, and have seen a steady stream of manuscripts coming in (1,429 submissions from February through April 2023 with 337 sent out for review). So far, 51 reviewed preprints have officially been published.

The mainstreaming of former disruptor ResearchGate continues. After years of ignoring publisher requests to honor access restrictions (and after two publishers successfully sued ResearchGate), the company has now integrated the industry-driven GetFTR access control system.

The Journal Observatory, funded by the Dutch Research Council (NWO), is a new project aiming to standardize and collect journal policies and approaches to peer review, preprinting, preregistration, data sharing, metadata availability, and related issues.

As part of a larger project looking at the shared landscape of infrastructure in scholarly communication, Ithaka S+R has released a Common Scholarly Communication Infrastructure Landscape Review.

A study using an algorithm to infer authors’ race or ethnicity (apply appropriate methodological skepticism as necessary) looked at the past two decades of articles in journals from six publishers and concluded that scientists from minority ethnic groups are underrepresented on editorial boards compared with their share of authorship of papers, see longer delays between submission and acceptance, and have their papers cited less than papers by white researchers.

Spain has adopted a national OA strategy.

To support OA books and journals, Arcadia has awarded MIT Press a $5 million endowment with an additional ​​$5 million “challenge” gift to incentivize other funders by matching their contributions.

It was inevitable that peer reviewers would try generative AI tools to help draft reviews. Here is an instance of one researcher asking ChatGPT-4 to write a review for an already-published JAMA Internal Medicine paper. The JAMA Editor in Chief, Kirsten Bibbins-Domingo, jumps in to remind people that entering a manuscript (particularly an unpublished manuscript) into a large language model (LLM) such as ChatGPT makes the paper available to the LLM (and its owners), which can violate the confidentiality policies of journals. This is one more thing to keep publishers up at night (“Are the AIs reading confidential manuscripts? Is the manuscript being stored anywhere by the AI?).

***
If you think technology will solve your problems, then you don’t understand technology – And you don’t understand your problems.—Laurie Anderson, adapted from Bruce Schneier