We often hear that scientists hoard data, refusing to share information even when doing so might speed advances to patients in dire need. (We touched on it briefly in this piece and it was a major element in a recent article in The New Yorker.) It’s not just about sharing results on the fly—once a project has been completed and findings published in a journal, most of us observers outside major institutions still can’t get access due to expensive subscriptions. The situation is made all the more unpalatable since most biomedical research is funded by taxpayer dollars. Yet the average taxpayer has little ability to see what comes of that funding.
So do all these factors mean we have a community of selfish scientists? The simple answer is no. The more genuine answer is: It’s complicated. The institutional inertia of the established scientific community strongly favors researchers who go along with the data-hoarding norms.
Consider the path of your typical life-science researcher, fresh out of grad school: Jane Scientist, a newly minted PhD in a sea of PhDs so vast there aren’t enough jobs for all of them. After landing a low-level gig in somebody else’s lab, she works to improve her lot by pitching in on as many projects as possible. She hopes some of them will result in a publication—the currency of advancement in science. With her name on enough papers, Jane may eventually be able to set up her own lab, where she will immediately face the challenge of having to raise money to keep that lab, because universities don’t cover that. Getting grants is linked to her publication history, since reviewers combing through five times more applications than they can fund want to make sure a winning scientist has a reputable work history. It is so difficult to win that initial grant that some funding agencies have a special piggybank just for first-timers so they don’t have to compete against established labs. Indeed, the average age at which a scientist wins her first grant in the U.S. has crept past 40. It’s not a career for the instant-gratification type.
The publish/get funding/publish again cycle will last throughout Jane Scientist’s career. If she brings in enough money to support her lab, and if she authors enough papers, and if those papers are in top-tier journals—then our friend gets a shot at tenure at her university, which offers a bit of relief from the constant panic.
Throughout Jane’s path, there are twin themes: the need for papers and the need for funding. Papers that come out in less-respected journals don’t help, so the top publications have an essential stranglehold on the market. This handful of journals—the Ivy League of publications—gets to make rules that suit them best, and those rules often prohibit scientists from releasing data prior to publication. Offer too many details in a conference presentation, get a little too talkative with a journalist, and these top journals may no longer consider your paper. The journals want exclusivity, and they want to break your research news for you.
In a field where scientists vie for scarce funding, the threat of being passed over by a top outlet effectively shuts down data-sharing prior to publication. What about access after the fact? Indeed, many scientists want to publish in open-access journals, like those run by the Public Library of Science, that let anyone read the contents without a subscription. But there are two gotchas. First, such open-access publications are newer and therefore tend to be considered less prestigious by grant committees and tenure reviewers, so well-intentioned scientists aren’t rewarded for open-access papers the way they would be if they scored a traditional publication. Second, business models to support open-access publications are still evolving—after all, the money to fund these outlets has to come from somewhere—and today many of them charge the scientists a fee to make the paper open access. That fee is usually covered by grant money, so taxpayers are still paying to get access to these papers, it’s just less visible than a publication firewall.
Outside of the life sciences, there’s been far more success in data sharing. Physicists routinely share data before publication, and early versions of their papers are made publicly available even before they’re accepted by a journal. (It is perhaps worth noting that life science advances are more likely to become useful intellectual property than physics advances.)
We are starting to see some biologists adopt these methods, often led by those who have tenure or guaranteed institutional funding to insulate them from the publish-or-perish cycle. Individually, no single scientist can upend a system that has been established over centuries; those who try may wind up starving their own labs and leaving the field.
What’s really needed is system-wide change. Some funding agencies have begun to require grant winners to ensure their publications become publicly accessible after a certain period of time, but those requirements need to be strengthened and should take effect immediately upon publication rather than six months or a year later. More importantly, the science community needs a better way to evaluate journals—one that doesn’t put such a premium on history—so that open-access publications and outlets with more open-minded approaches to data sharing are finally on an even playing field with traditional journals.
Only when they can fairly choose to share data without risking their careers will we see what scientists are really made of.