NEJM says open access is unnecessary. Right.

Surprise: the New England Journal of Medicine thinks open access is a bad idea. Open access is the model of scientific publishing in which all results are freely available for anyone, anywhere, to read.

This week NEJM published an editorial by one of their correspondents,  Charlotte Haug, that purports to present an objective look at open access publishing, and finds that the "experiment" has failed, and that free access to scientific publications hasn't delivered on its promises.

What is NEJM worried about? Their expensive, exclusive model of publishing–where everyone has to pay high subscription fees, or else pay exhorbitant fees for each article they read–is threatened by scientists who want all science to be free. Pesky scientists!

NEJM is especially worried about "Plan S", a proposal in Europe to require that all scientists whose work is funded by the public be required to publish their results in open-access venues. Plan S is due to take effect very soon, in 2020 for 11 research funders in Europe.

The NEJM article is a clever but deeply flawed effort to prove that open access isn't working. It's full of fallacies and straw men, so much so that it's hard to know where to begin. Since they're not playing fair, though, I won't either: I'll cherry-pick three of Haug's arguments and explain why she's wrong about each one.

But first, to set the stage, let's remind everyone of what we're talking about. Scientific papers are written by scientists (like me), who are largely funded to do their work by governments, non-profit organizations, and occasionally by commercial companies. The writing is done by the scientists themselves, who submit papers to journals for peer review. The peer reviewing is also done by scientists (again, like me) who do this work for free. The journals pay nothing for all this work.

In other words, we do all the work for free, using funding provided by the public, and the journals then take that work and sell it for a very tidy profit. (Richard Smith estimates that NEJM itself has an income of $100 million with a 30% profit.) The vast majority of scientific and medical journals are owned by five for-profit corporations, as the NEJM points out:

"The five largest publishing houses (SAGE, Elsevier, Springer Nature, Wiley-Blackwell, and Taylor & Francis) continue to grow, with high profit margins."

For the past two decades, scientists have spoken out more and more over the outrageous practices of for-profit publishers, whose subscription fees and profits have grown while the costs of distribution have plummeted. Virtually everyone gets scientific papers online now. Why sign over copyright when we can distribute our work so cheaply now? The open access movement was founded to provide an alternative: open access journals allow everyone to read all the content for free, and the authors retain their copyrights.

Now let's look at the NEJM article. Haug starts by pretending to agree that open access is a good thing, writing:

"The idea — that the results of research should be available to be read, discuss, and examine... — has few, if any, opponents in either the scientific community or the public."

Reading this, you might think that Haug (and NEJM, by extension) are fans of open access. They are not.

Haug then proceeds (she thinks) to dismantle the arguments in favor of open access. First, she states that publishing costs have not dropped, but have increased. As evidence, she asserts that "Electronic production and maintenance of high-quality content are at least as expensive as print production and maintenance." This claim is, frankly, nonsense, but since Haug doesn't cite any evidence to back it up, there's nothing really to refute. It's obviously much cheaper to post a PDF on a website than to print thousands of hardcopies and physically ship them to libraries around the world. If costs are going up (and again, Haug cites no evidence), that could be simply because publishers are paying themselves higher salaries (NEJM reported compensation of  $703,324 for its chief editor in 2017), or hiring large staffs, or renting luxurious offices–who knows? Haug doesn't explain.

In any case, the costs of publishing at NEJM, a closed-access, subscription-based journal, have little to do with whether or not scientific and medical research should be freely available.

Her next argument against open access is that the most highly-cited journals are subscription-based, like (ahem) NEJM. My response: so what? Everyone within academia knows that it takes a very long time to establish a reputation as a "top" journal, and young scientists will always want to publish in those journals, regardless of how expensive they are. This has given closed-access journals like NEJM (and Nature, Science, JAMA, and Cell, to name a few more) tremendous power, which they have wielded to fight against open access at every opportunity. This editorial represents another example of that fight. The fact that many scientists still want to publish in these journals doesn't mean they should keep the results locked behind a paywall.

Setting aside this tiny number of "prestige" journals, open access papers do get cited more, as was demonstrated by this study from 2016. The evidence shows that open access does lead to higher impact: papers that are freely available are read more and cited more.

Finally, let's turn to Haug's coup de grace, which she wields near the end of her piece, as a sort of "proof" that open access is really unnecessary. Here she argues that NEJM is already open, mostly:

"About 98% of the research published in the Journal since 2000 is free and open to the public. Research of immediate importance to global health is made freely accessible upon publication; other research articles become freely accessible after 6 months."

First, let's acknowledge that merely by pointing this out, Haug is admitting that the main arguments for open access are legitimate; i.e., that it's a huge benefit to society to make research freely available. I'm going to agree with her here.

What Haug doesn't mention here is that there is one reason (and only one, I would argue) that NEJM makes all of its articles freely available after some time has passed: the NIH requires it. This dates back to 2009, when Congress passed a law, after intense pressure from citizens who were demanding access to the research results that they'd paid for, requiring all NIH-funded results to be deposited in a free, public repository (now called PubMed Central) within 12 months of publication.

Scientific publishers fought furiously against this policy. I know, because I was there, and I talked to many people involved in the fight at the time. The open-access advocates (mostly patient groups) wanted articles to be made freely available immediately, and they worked out a compromise where the journals could have 6 months of exclusivity. At the last minute, the NIH Director at the time, Elias Zerhouni, extended this to 12 months, for reasons that remain shrouded in secrecy, but thankfully, the public (and science) won the main battle. For NEJM to turn around now and boast that they are releasing articles after an embargo period, without mentioning this requirement, is hypocritical, to say the least. Believe me, if the NIH requirement disappeared (and publishers are still lobbying to get rid of it!), NEJM would happily go back to keeping all access restricted to subscribers.

The battle is far from over. Open access advocates still want to see research released immediately, not after a 6-month or 12-month embargo, and that's precisely what the European Plan S will do.

With Plan S looming, I've no doubt we'll see more arguments against open access in the coming months, but scientists have at least one ace up our sleeves: we're the ones who do all the work. We do the experiments, we write the papers, and we review the papers. Without us, the journals would cease to exist. The journals will have no choice but to go along with plan S, because without the scientists, they'll have nothing to publish. Let's hope the U.S. will follow suit in the very near future. It's long past time to change the archaic, closed-access policies that have kept medical and scientific results–results that were funded by the public–locked behind the paywalls of for-profit publishers.