New eLife editor Michael Eisen wants to shake up scientific publishing

Michael Eisen, the new editor-in-chief of the journal eLife. (Photo by Jim Keeley, HHMI)

The University of California system’s recent decision to walk away from negotiations with scholarly journal publishing giant Elsevier highlights once again the many problems within the scientific publishing business, a $10 billion-per-year worldwide enterprise that is the bedrock of modern science. Publishers like Elsevier, Springer — which publishes the high-impact journal Nature —and dozens of other for-profit companies and nonprofit scientific societies are an essential part of the give-and-take of science, offering a place to publish and share new results. But they also charge for scientists and the public to read those results, much of which the public originally funded through federal agencies such as the National Institutes of Health (NIH) and the National Science Foundation (NSF). The UC most recently paid Elsevier $11 million for a year’s worth of access to its journals, which include the well-known medical journal The Lancet and more than 2,500 lesser-known titles, from Poetics to Fungal Biology.

Michael Eisen, a professor of molecular and cell biology and a Howard Hughes Medical Institute investigator, has done his part to disrupt the stodgy business, which he thinks not only takes advantage of authors and universities, but distorts the process of science. As a founder 19 years ago of the first open access journal, PLOS (Public Library of Science), he sought to establish a new business model where scientists pay to publish, while anyone can view the results for free. Other journals slowly moved in that direction, but even today, only about 20 percent of all published research is open access, and almost none of the papers appearing in high profile publications like Nature, Science and PNAS (Proceedings of the National Academy of Sciences) can be read by the public without charge.

Appointed last month the editor-in-chief of the open access journal eLife — Berkeley Nobel Laureate Randy Schekman is stepping down as founding editor — Eisen has a new platform to shake up the field of science publishing and help make it serve scientists and the public.

What was your reaction when UC ended negotiations over a contract with Elsevier?

I think I wrote on Twitter, “Good. 25 years too late. But good.”

It has been clear forever that this (paying for subscriptions) was a problem, but universities just keep going back to the well and making one more deal with the publishers. There was a lot to like in the way UC handled this — in particular, pushing for universal access to UC-published papers. Not just for UC, but to make sure that everybody can read UC papers for free.

You can understand why Elsevier is wary of such deals, because they have the potential to have a broad economic impact on its business. Nobody is going to cancel their Elsevier contract because they can get free access to just UC papers. But if it is UC and the University of Michigan and big universities in Europe and Asia and South America, then subscriptions have less economic value. It is good UC is pushing things in the right direction.

Here is what would have to happen for the Elsevier thing to really matter: UC would have to stick to its guns and not cave — just agree to a lower price with Elsevier. The second thing is that other universities have to follow. UC is big, but it is not big enough. The simple loss of papers and subscriptions from UC isn’t going to be enough to drive Elsevier to change unless it thinks this is a harbinger of other things. A lot of people are looking at what is happening with UC now, and they’ll either be emboldened to take on Elsevier, or they will realize that resistance is futile, depending on what happens next with UC.

Elsevier is only one publisher, though a big one. What about our subscription agreements with other publishers, like Springer or Science?

Maybe this will lead to what really needs to happen, which is for UC to go into these negotiations with an ironclad policy that says, “We are never paying for subscriptions.” I really think that UC and the state of California should require that people who work for the California government or receive funds from it have to make their work freely available to the public. That should be an irreducible first principle of publishing.

Then, UC has some challenges. It needs to make sure that the authors publishing out of UC have access to effective means of publishing their work, which is going to necessitate spending some money, whether it’s coming directly from UC or from grants or whatever.

Elsevier was trying to have it both ways. It wanted UC authors to pay it to publish, but then it also wanted UC to pay for access to the content. I understand why Elsevier is trying to get that bargain, but it is also totally appropriate for UC to tell it to take a hike.

In a tweet after the Elsevier decision was announced, you wrote: “I’m all for boycotting Elsevier — I’ve been doing it for 20 years — but I do wish people would realize that the real culprit here is not Elsevier itself, but rather a publishing industry (that includes many nonprofits) addicted to profits from subscription journals.” Could you elaborate?

To me, journal publishing in its current form is terrible for science. When we started PLOS (Eisen, Pat Brown of Stanford University and Harold Varmus, at the time the NIH director), we had one big target: the economics of journal publishing. From my point of view, science is getting a raw deal out of this arrangement, because it is providing all the money, but not getting access for everybody on Earth. The open access model that we and others initiated 20 years ago just changes the nature of the transaction so that you are paying them (the journals) up front to publish an article. There is no longer any rationale for denying anybody access.

But access is only one of the pathologies in publishing — all, in some broad sense, legacies of the historical accident that the printing press was invented before the internet.

The subscription model is a print business model, a model that presumes that the primary expense in producing a journal is in printing and shipping. So, it makes some sense not to mail a free copy of the journal to anyone who requests it.

Of course, that logic disappears when you are sending electrons instead of pages. Open access is meant to address that problem.

But other things are left behind by print. The format of papers, the bundling of papers together into journals. The very existence of a journal as a concept itself is somewhat nonsensical, especially now, when we are pushing 30,000 discrete titles in science publishing.

Then, there is the idea of peer review and rejecting papers. In a print world, pages in a journal were a scarce commodity and, therefore, you had to ration them by some form of assessment that decided which ones should be included. Again, with the internet, that is not necessary. The system we have where a paper goes to one journal, it gets reviewed, sometimes it gets accepted, but often it gets rejected and it goes to another journal, and so forth. It doesn’t even make sense.

And then, authors who send (the journal) papers get paid in the coin of the realm, which is citations. As long as the whole scientific endeavor is addicted to using journal titles as an indicator of quality and import, the people who control those indicators have incredible power not only to dictate the way that publishing works, but to dictate what type of science gets done. It is very distorting and another area where the interests of science and the interests of the public are not being served by the publishing industry.

As you’ve pointed out, universities make hiring and tenure decisions based on scholars’ citations in premier journals.

To some extent, UC outsources to journals decisions about who gets hired and who gets tenure. The price we pay for that relationship is literally billions of dollars in subscription costs, but more fundamentally, journal publishing is not trying to figure out who should be hired and tenured. The journal system was not designed to do that; the way papers are peer reviewed is not meant to figure out who should be hired. This affects the type of work that (scientists) do, and the way they sell their work is to entice journals. The whole journal publishing system has its talons in science in so many different ways that make it incredibly important that we fix it and get rid of all these pathologies, but also make it incredibly difficult.

You are now joining that industry as editor-in-chief of eLife, a seven-year-old open access journal dedicated to the biomedical and life sciences. Is this an opportunity to help change the way scholarly publishing works?
The vision I pitched to eLife was to build a new ecosystem of publishing. Just as a thought exercise, imagine that we set up a little operation at eLife where people send their papers to us, we review them and, instead of deciding thumbs up or thumbs down for eLife, (the editors) look at the paper and say (to the researcher), “You know what? This is a Nature paper. Or a PNAS paper.”

If, instead, you render the judgement as, “Here’s who we think the audience of the paper is, here’s how important we think it is, here is why we think it’s important, here are the aspects of the paper that we think deserve attention” — that is a more useful and robust way of recording the output of review.

That is not necessarily what I am proposing. But the whole appeal of eLife is that it is high functioning, it has lots of dedicated people, it’s got a reputation, it’s got lots of papers coming in. We can work with authors, editors and others to do the kind of experiments with mechanisms of peer review and other aspects of publishing that they and I want to see happen.

It is a challenging problem to try to move science publishing from its 17th century home to one in the 21st century.