Publishers of scientific journals are facing renewed threats to their business models from both sides of the Atlantic. As European science funders promote a radical new open-access (OA) publishing mandate they unveiled last month, the Trump administration is considering changes to a five-year-old directive governing the public release of research literature sponsored by federal agencies.
A delegation led by Robert-Jan Smits, the European Commission’s special envoy on OA, visited last week with officials of the White House Office of Science and Technology Policy (OSTP) and other federal agencies as part of an effort to gain broad support for the new European policy, known as Plan S. Due to take effect in January 2020, the initiative would require recipients of research grants from a dozen European national funding agencies to publish their research solely in journals whose contents are immediately available for free. It would prohibit researchers funded by those agencies from publishing their findings in most high-prestige journals, including Nature and Science, which are subscription based and make papers from those agencies available for free only after a one-year delay.
Plan S has been endorsed by both the European commissioner for research, science, and innovation and the European Research Council, which administers the European Union’s research program. Officials say it will apply to the EU’s next $100 billion, seven-year research “framework” plan, which is to take effect in 2021. “For the last 20 years, libraries, universities, and [others] had the possibility to sort this out. But they did not,” Smits says. “Now the funders have stepped in, and they now call the shots.”
Smits said at a 3 October briefing in Washington, DC, that the Plan S coalition hopes to expand its model globally, though it was too early to expect any commitment from the US. “The publishers tell us that we can only flip our journals [to an OA model] if this happens at the global level,” he said. Plan organizers have received invitations to visit with officials of South Africa, India, China, and Japan.
Meanwhile, publishers and OA advocates have rushed to meet with OSTP staff in recent weeks upon learning the agency might change its OA policy for US publicly funded research, which for now requires that such literature emerge from behind paywalls after an embargo period of up to one year. An OSTP spokesperson wouldn’t comment on the discussions. But some publishers who met with staff at the agency fear that officials may be eying shortening or eliminating the embargo and capping the fees, known as article processing charges (APCs), that publishers charge authors for their papers to be OA. Caps on these charges are also a feature of Plan S, although its framers haven’t yet established specific levels.
The developments in Europe and the US come as scholarly journal publishers gradually shift toward a substantial but incomplete embrace of OA. The immense impact of funding agencies adopting a Plan S–like approach would extend not only to publishers but also to academic institutions and individual researchers.
Federal policy in limbo
Among the physical sciences publishing societies whose representatives met with OSTP staff over several weeks were the American Physical Society, American Chemical Society, Optical Society, American Institute of Physics (AIP, which publishes Physics Today), Society of Rheology, and American Astronomical Society.
Since 2013, when then presidential science adviser John Holdren instituted the one-year maximum embargo before research was made freely available, agencies have taken various paths to OA. The Department of Energy, Department of Defense, and NSF, which collectively account for nearly three-quarters of all federally sponsored nonbiomedical research articles, enlisted CHORUS, a database established by commercial and nonprofit scientific publishers. NASA, NIST, and the Environmental Protection Agency contracted with the National Institutes of Health’s long-standing PubMed Central public repository to house their sponsored research articles. In 2008 NIH had established an OA policy that required public release after one year.
Society members attending meetings at OSTP varied as to whether they thought major revisions in OA policy were in the works. These attendees, most of whom would speak only on condition of anonymity, say interpretations depend on which staff member they consulted. “Different staff have different opinions about what should go into the policy,” says one source. Most sources believe there will be some revisions.
Publishers are most concerned that cutting back or eliminating the embargo period on OA would greatly reduce incentives for universities and other institutions to pay for their journals. “Twelve months is a natural time period that librarians consider for subscriptions,” says Fred Dylla, former CEO of AIP. The concern for embargoes is particularly great for niche journals and those that are published only quarterly, he says.
Some publishers perceive OSTP to be gathering information on whether and how well the existing policy is working. “It didn’t seem like they were just going to throw it all out and start over again,” says Joel Parriott, deputy executive officer of the American Astronomical Society. “They were looking to see if there were certain areas that could use further attention.”
Heather Joseph, executive director of the open-access advocacy organization Scholarly Publishing and Academic Resources Coalition (SPARC), says the OSTP staff seemed to be interested in ensuring agencies were complying with the policy. “My read is their major questions are ‘Is this an effective policy?’ and ‘Is it good enough?’ ” she says.
Several sources say the OSTP review appears motivated by an overarching Trump administration drive to change whatever policies were set during President Obama’s tenure. In this case, says one, the staff appeared to focus on whether the process their predecessors had followed was flawed.
Proposed changes, if any, will likely have to be approved by Kelvin Droegemeier, whose nomination as OSTP director awaits confirmation by the Senate. His views on OA are unknown.
Plan S is detailed
Plan S was coauthored by Science Europe, an association of 37 European funding agencies and research performers with combined budgets of $23 billion. Its president, Marc Schiltz, expects a majority of his members will sign up. “We went public at the moment we felt sufficiently strong to go public,” he says.
The signers of Plan S include the national funding agencies of the UK, France, and Italy, as well as countries with smaller research budgets. Notably absent is the German Research Foundation, DFG. Schiltz and Smits say they are confident that DFG will come on board after the resolution of a court case over a faculty challenge to an OA publishing mandate by the University of Konstanz. A DFG spokesperson, however, says that although the foundation supports all forms of OA, it disagrees with Plan S’s mandate of “specific forms of open access” because it impinges on researchers’ academic freedom to decide where to publish.
David Sweeney, executive chairman of the funding agency Research England, says the framers of Plan S don’t want to specify exact publishing models. But Plan S most resembles the gold form of OA, in which the entire content of the journal in which an article is published is free of charge, without an embargo. And except for a transition period that “should be as short as possible,” Plan S won’t allow continued publication in hybrid journals, the model that now accounts for nearly three-quarters of the industry (see Physics Today, May 2017, page 24).
Hybrid journals derive most of their income from subscriptions but also accept OA articles at a charge to the authors. Gold journals are funded entirely by APCs. Some funders, including the UK government and the Gates Foundation, provide extra funding in their grants to cover APCs. For US funders, APCs generally must be paid for from grant funding; hence there is that much less money to pay for the research itself.
According to a December 2017 report by Universities UK, 73% of journals owned by the 40 top scholarly publishers in 2017 were hybrid. That’s an increase of 17% from 2015. Although the number of subscription-only journals has fallen, fully OA journals still represent only 18% of the journal count.
“Hybrids were promised to us by publishers as a transition tool for a couple years, and then everything would be full and open access,” Smits said at the Plan S briefing in Washington. “This transition tool is now being used as a business model, and that is something we absolutely don’t want.”
As to the caps on APCs for OA articles, Schiltz notes, “We don’t want to switch from excessive subscription [costs] to excessive APCs.” Caps should scale with the quality of the services being offered by publishers. Ultimately, competition among publishers could do away with the need for caps, he says.
To adapt to Plan S, publishers could also implement a variety of what’s called green OA in which the published version of an article in a subscription journal simultaneously becomes freely available in a separate collection. Many publishers now allow authors to publish some form of their articles in repositories such as author websites, arXiv, or scholarly collaboration networks like ResearchGate. But the freely available papers aren’t usually identical to the final, citable versions in journals. Sweeney, who coleads a task force that will draw up a Plan S implementation blueprint by year end, says he believes papers posted in an open repository will be compliant—but only if they are identical to the final published version.
Publishers react
Unsurprisingly, many publishers are unhappy with Plan S, particularly its exclusion of the hybrid model.
“A lot of people advance the idea of open science, which is a laudable and a necessary idea, and illogically jump to the declaration that there is one particular model of publishing that is in keeping with open science,” says Rush Holt, CEO of the American Association for the Advancement of Science, which publishes the Science family of journals. Holt says the plan “runs afoul of what has worked so well for our authors and researchers who use the material in our publications.”
Steven Inchcoombe, chief publishing officer at Springer Nature, one of the largest journal publishers, says the hybrid model continues to drive the transition toward full OA. He notes that more than 70% of the company’s authors in four European countries make their research available immediately upon publication. A company-sponsored study early this year showed that without the hybrid option being available to UK authors, only around 30% of their published research would have been OA, instead of the actual 77%.
The subscription model also “sustainably works for highly selective journals with substantial news, analysis, and review content like the Nature research journals,” Inchcoombe says. Since selective journals reject most of the submissions that are evaluated, they have to review many manuscripts to fill their pages. The result is high processing costs and, ultimately, a cost per article that is much higher than for less selective journals. A former Springer Nature executive says the cost per published article in Nature is around $39 000.
Smits says Plan S members want to ensure the continued health of nonprofit society publishers, which use the revenues from their publishing operations to finance educational, journalistic, outreach, and other activities. “They are an essential part of the science and research infrastructure, and we recognize they add value,” Sweeney adds. “But it may not be done the same way in the future.” The two say they will consult with society publishers as they draft the implementation plan, but they wouldn’t say what sort of accommodation might be made.
Smits says European governments and universities are fed up paying what he says is anywhere from $10 billion–$25 billion each year in subscriptions worldwide to publishers, some of which are making profits of 30–40%. “There’s something very rotten in the system, and it has to change big-time.”
Neither Holt nor Inchcoombe would estimate what percentage of their respective content is supported by the Plan S funders. Holt would only suggest that the figure is in the single digits. But one society source says as much as one-fifth of some physics journals’ content could be subject to Plan S mandates.
Researchers caught in the crossfire
Some European researchers also are unhappy with Plan S. Lynn Kamerlin, a biochemist at Uppsala University in Sweden, says the hybrid ban could prevent some European researchers from publishing in high-quality journals. “If the rest of the world continues to submit their best work to those journals, then researchers outside Europe may think twice about collaborating with researchers who are forbidden from publishing in them,” she says.
Britt Holbrook, a New Jersey Institute of Technology assistant professor and member of an OA advisory panel to the European Commission, says Plan S was essentially drawn up by funders and directed at publishers. “The part that’s missing is … input from the researchers,” he notes. Holbrook says the plan infringes on academic freedom. “Research includes dissemination of the results,” he says. “Part of academic freedom is that faculty give themselves the rules on that.”
Sam Hay, a University of Manchester chemist, says he agrees with the goal of OA. “But the problem with most open access as it stands right now is it’s really expensive for the author,” he says. “We couldn’t afford to publish multiple papers a year.”
Chemists funded by UK Research and Innovation, a Plan S signatory, won’t be able to publish in prestigious journals such as the Journal of the American Chemical Society or most of those published by the Royal Society of Chemistry. Indeed, Plan S would bar publication in most of the higher-prestige journals in any of the disciplines, unless they switch to an all-OA publishing model. “That’s an issue because many of our colleagues have drivers to publish in these journals,” Hay says. “You start having issues with mobility and other things as well,” he adds, referring to the repercussions Plan S will have for academic scientists’ reward system.
That system currently prizes publication in high-impact—and mainly subscription-based—journals. The rest of the world is unlikely to change the incentive, flawed as it may be, in response to Plan S, notes Kamerlin. An ambitious student or postdoc might reconsider a position at a lab or research group funded by a Plan S agency, fearing that it could hamper future mobility and career prospects.
Smits acknowledges that the academic reward system must change from what he calls its “obsession” with the journal impact factor. And Sweeney says Plan S “will require steps to be taken by universities in their appointments and promotion procedures, and by funders in their grant assessment procedures.” But he adds, “If we wait for that change to happen, we will be here in 10 more years, if not longer.”
