Why I left the editorial board of the prestigious scientific journal NeuroImage — and helped start something new

Last spring, I joined a rebellion.

The entire editorial board of the prestigious scientific journal NeuroImage, of which I was a member, resigned after its publisher refused to take steps to ease the high costs scientists are required to pay for publication in the journal. While this decision threw the brain imaging research community into a frenzy, the rest of the world barely noticed. Yet it should. The accessibility we are striving for is crucial to combat misinformation, increase the pace of discovery, and promote innovation.

To understand why requires taking a tour through the strange, surprising, multibillion-dollar business of scientific publishing, where things function very differently from what nonscientists might imagine.

Science has lofty goals. It aims to solve problems facing humanity, improve health, and discover the fundamental nature of the universe. To make progress on such challenging problems, scientists rely on collaboration, critical review from peers, and open exchange of information. Scientific journals are the main avenues for this communication. Ideally, the more important a research finding is, the more prestigious a journal it is published in, and the more widely disseminated it becomes.

But that’s not the way it actually works.

Peer-reviewed journals represent science in the raw: the places where researchers share their latest discoveries with each other with all the technical detail needed for their work to be evaluated and reproduced by their peers. Some journals reach only a small audience of specialists; others are widely read by the larger research community. Taken together, journals form the circulatory system without which science could not function.

As recently as a few decades ago, journals were a tiny, impoverished publishing backwater. Today, after decades of mergers, acquisitions, and steady commercialization, more than half the scientific articles published worldwide appear in journals owned by just five large for-profit corporations: Elsevier, Wiley, Taylor and Francis, Springer, and Sage. One study estimated that the global scholarly publishing industry was worth $26.5 billion in 2020.

Scientific journals began as “newsletters” for scientific societies. As a result, they have some idiosyncrasies that seem perplexing to readers of other periodicals. For one, the authors write the content for free. The articles are assessed by several peer reviewers, who also work for free, and an editor (typically paid a small honorarium) decides which articles will be published in the journal. Thus, the expenses involved are minimal.

Because scientists rely on journals to keep up with the latest research, institutions such as university libraries have no choice but to subscribe to thousands of them. (It’s hard to say exactly how much the average subscription costs, given that institutions can negotiate prices, but a personal subscription might cost $750.) Perhaps surprisingly, subscription costs have been rising continuously despite the transition from physical books of articles that must be printed and delivered, to electronic collections that exist entirely online. Each scientific journal functions as a tiny monopoly, to the benefit of the big publishers, which control access to the articles. Normal laws of competition, under which consumers are free to shop for the best combination of price and quality, don’t operate. In these rarefied circumstances, it is not surprising that Elsevier, the publisher of the top-ranked NeuroImage journal, earned 37.8% profit in 2022. The cost of journal subscriptions is a burden to all scientific institutions but falls most heavily on smaller universities or those from developing countries, which frequently can’t meet the rising costs.

In recent years, many journals have responded to criticism of the subscription model by becoming open access. Under the open access model, journals make their content freely available online immediately. The Biden administration recently updated their guidelines to require this type of immediate availability for federally funded research.

Open access is one critical piece of the process of enabling free exchange of scientific research. The problem is that journals recover the lost revenue from institutional subscriptions by charging their scientist-authors thousands of dollars in so-called “article processing charges” when their papers are accepted, which simply shifts the costs to the scientists and their labs. These fees are commonly $3,000-$4,000 but can exceed $10,000 for particularly prestigious journals. For newly established labs or researchers in certain countries, article processing charges can be an insurmountable barrier to publication, causing valuable research to go unpublished. Even in well-funded labs, the costs of publication are substantial and absorb money that could be used far more productively. A lab might publish 10 to 12 papers a year, with publication costs equivalent to an entire set of experiments they now can’t perform or the salary of graduate student whom they can’t support. The vast majority of scientific research is funded by national agencies that in turn are funded by taxpayer dollars, so the high charges are, in effect, a private tax on research, paid by the public, that benefits no one but publishers. A group of scientists recently calculated that author payments worldwide came to $1.06 billion from 2015 through 2018 for the big five publishers alone.

Scientists like me are beginning to rebel against the unjustifiable waste of resources. NeuroImage went open access in 2020, and there were immediately concerns about the fees, which, at $3,400, seemed to bear no relationship at all to what Elsevier actually needed to pay the expenses of running the journal. My fellow editors and I asked Elsevier where the extra money was going. They told us some journals needed to subsidize others to make an overall profit. We reminded them of their 30% to 40% profit levels. They didn’t answer. We did our best to figure out what it would really cost to run a journal without excessive profits and asked them to lower their charges to $2,000. They wouldn’t engage, simply telling us they thought their charges were in line with our competitors and that the “market” supported them. With the monopolistic nature of the publishing industry and the complete absence of downward pressure, we were certain that these costs would only keep increasing. It was not an acceptable reply.

This was why I and the other members of the editorial board, made up of more than 40 leading scientists, unanimously decided in April 2023 to walk out, despite our qualms about disrupting a beloved and important avenue of scientific communication. To fill the resulting void, we founded a new nonprofit open access journal last spring, Imaging Neuroscience, where we’re working to keep article charges to $1,600 and waiving them entirely for researchers from low-income countries. Response to the new journal has been incredible, with 150 submissions within the first two months and more than 1,200 researchers signed up to review. Two volumes have already been published and more papers are available online. Meanwhile, NeuroImage continues to publish, but reports reaching the former editors indicate that it has struggled to find people to serve as editors and reviewers and has resorted to using some of Elsevier’s nonscientist staff editors to keep operating.

The success of Imaging Neuroscience is due to the enthusiastic support of our research community. Editors have the ability to shape their scientific field by bringing attention to exciting new approaches and highlighting stubborn challenges in need of solution, but it is the research community that makes a journal great. Our community believes that publicly funded scientific knowledge ought to be publicly available, not only to junior scientists, impecunious graduate students, and scientists from poor countries, but to interested nonscientists as well. How can we lament the spread of misinformation and distrust of science when important research findings are hidden behind a paywall? How can we justify charging patients to access the research behind the clinical trials that drive medical innovation when their tax dollars funded the studies in the first place?

With an estimated 70% of journals still paywalled and even more charging outrageous article processing charges, the ideal of open, accessible science remains out of reach, yet it’s clear the current model of scientific publishing is unsustainable and needs to change. With the first cracks beginning to show, now is the time for the scientific community, the legislature, the tax-paying public, and (most importantly) the agencies that fund scientific research to join the fight against high-profit journals and promote the widespread dissemination of research. To do anything else is a disservice to science.

Shella Keilholz, Ph.D., is a professor in the Wallace H. Coulter Department of Biomedical Engineering at Emory University and Georgia Tech, and a member of the editorial board of the journals Imaging Neuroscience, eLife, Network Neuroscience and eNeuro.