The Steep Price of Free Science Access
Governments and funding agencies have promised the public unfettered open access to research. Scientists could foot the bill.
Conducting high-quality research is expensive. Salaries and stipends must be paid. Equipment must be purchased and maintained. Reams of data must be collected and analyzed, sometimes by costly supercomputers. Disseminating the results of that investment also isn’t cheap, and the cost of scientific publishing has traditionally been shared between publishers and readers. But scientists increasingly find themselves adding new line items to their budgets: publication fees.
Like other producers of print media, scientific publishers have traditionally sold access to their content through subscriptions, primarily to universities and other research institutions, which in turn provide access to their faculty, students, and staff. But over the last two decades, the open access (OA) movement has sought to do away with subscriptions entirely and provide free and unrestricted access to the world’s scientific knowledge. On Aug. 25, 2022, the White House’s Office of Science and Technology Policy announced that, by 2026, all peer-reviewed journal articles about research funded by the United States government must be freely available immediately upon publication. The announcement follows similar initiatives in Europe, notably Plan S.
The cause is noble in principle. Taxpayers fund a great deal of research through federal grants, and the argument that citizens should have access to the resulting papers is a strong one. Advocates of the OA model also point out that it benefits researchers: Not every scientist works at a university that can afford pricey journal subscriptions, and OA papers are also more widely read and highly cited.
The trouble is this: Publishing high-quality, peer-reviewed journals is expensive, and there is not yet clarity on who, in the OA ecosystem, should bear the cost.
While OA mandates have exacerbated this dilemma, OA itself is not new — especially for physicists, who, after all, created ArXiv, one of the first online repositories of scientific preprints. APS has been a supporter of arXiv since the early 1990s, when journals were only accessible via subscription. Authors who publish in the Physical Review journals are welcome and encouraged to self-archive the peer-reviewed version of their manuscripts on arXiv.
APS launched its first open access journal in 1998, four years before the Budapest Open Access Initiative set the modern OA movement in motion. The journal, Physical Review Accelerators and Beams, continues to publish advances in accelerator science, technology, and applications at no cost to authors or institutions thanks to industry sponsors. So-called diamond OA journals, however, are more exception than rule, because relying only on external support is unsustainable.
Instead, publishers have largely moved their journals to the gold OA model, where the revenue that traditionally came from subscriptions is offset by article processing charges, or APCs, paid by authors, whose research is then immediately and freely available to read. Some APCs are reasonable, as in the nonprofit space. Others run upwards of $11,000 per article.
There are consequences to this move. Given the pressure researchers are under to continually publish papers in prestigious journals, these fees can quickly add up and divert funds away from research itself. A recent survey of more than 3,000 physicists conducted by APS and fellow physical societies AIP, IOP, and Optica revealed widespread support for OA, but respondents consistently reported APCs as a barrier to publishing in non-subscription journals. Early career researchers in particular said they struggle to secure funds for publication charges.
Moreover, attaching a monetary value to each paper incentivizes some publishers to publish more, and not necessarily the best, science. All the while, scientists face enormous pressure to publish for career advancement. In a 2020 survey, 12.5% of early career APS members said they have at times considered engaging in unethical behavior to meet the demands placed on them — up from 7.7% in the original 2003 survey. Rates of data falsification also doubled during that period.
A shadow industry of predatory publishers has risen to take advantage of this vulnerability, charging fees but offering little or no review. In one example, an author (read: not a scientist), in response to an invitation from a urology journal, published a fake paper about a fictional disease inspired by an episode of Seinfeld. My own inbox is littered with flattering invitations to submit or review for such-and-such journal, and it’s increasingly difficult to distinguish between legitimate publications and scams.
The burden of “pay-to-play” publishing falls especially hard on researchers from less-resourced institutions and lower income countries. As the journalist Shi En Kim wrote recently for APS News, “the cost to publish in top-tier journals can be as much as, or more than, the yearly salary or entire research budget of a professor in a developing country.” APS — acutely aware of this peril — has made great strides in subsidizing the costs of publishing and accessing research for scientists around the world. But the basic OA equation still needs balancing.
So how do we get back on the road to a sustainable OA future, where publishers like APS can maximize access to research while continuing to invest in the scientific community? Experiments abound — including around how to preserve excellent peer review against the pay-to-play incentives that could undermine it. Some publishers have done away with the acceptance and rejection of papers altogether, releasing only peer-review reports on preprints instead; some scientists have sought to separate the peer review process from scholarly publishing altogether. I am watching these experiments with interest. For now, APS should continue to offer a variety of publishing options to meet the diverse and changing needs of the broader physics communities. Hybrid journals give authors options. APC waivers for researchers from lower income countries help level the playing field. Rigorous journal standards protect the science.
Above all, APS should be guided by its “by scientists, for scientists” creed. The organization, as a nonprofit membership organization, does not have commercial interests. It publishes for the sake of good research — for the people doing, reading, and applying science. “Science is our only shareholder,” its recent Purpose-Led Publishing coalition says. As the industry changes, and as the promises and perils of open access become clearer, this mantra will light the path ahead.
Robert Rosner
Robert Rosner is a theoretical physicist and the William E. Wrather Distinguished Service Professor at the University of Chicago. He served as the 2023 APS president and currently serves as 2024 past president.