This week, the scientific publishing giant Elsevier, which produces thousands of academic journals, and Representatives Carolyn Maloney, a New York Democrat, and Darrell Issa, a California Republican, withdrew their support for the Research Works Act after public outcry from public-access advocates. Currently, some federal agencies require that researchers who rely on government funding make their resulting journal publications freely accessible online. The Research Works Act would have forbidden any agency from imposing this requirement, allowing publishers to retain rights to the papers. As with the recent battles over the Stop Online Piracy Act (SOPA) and the Protect IP Act (PIPA), opposition and the sudden drop-off in support for the legislation suggests that big content companies are losing some of the traction in Washington they once enjoyed.
But the Research Works Act was always largely symbolic. "It's a stake in the ground," said Allan Adler of the American Association of Publishers earlier this month, "designed to point out where the publishing industry stands." At issue is who, if anybody, should retain copyrights to-and make money off of-academic work that involves stakeholders as various as the government, universities, scientists, and publishing houses. The debate continues over whether, in the networked age, copyright will keep its place as an absolute, inviolable right. The ultimate outcome of that debate has profound implications for the future of scientific publishing-and science-in the United States.
The NIH Factor
The National Institutes of Health's (NIH) policy on public access to scientific research is at the center of this debate. In the late 1990s, publishing in biomedicine was stagnant. The research community had failed to adapt to changes in collaboration and content distribution spurred by the growth of the Internet. Publishers ruled the scientific publishing ecosystem, and scientists worked mainly within their confines.
Like other scientists, cell biologist and Nobel laureate Harold Varmus handed over the copyrights to his scientific papers to journal publishers "without pause or complaint," as he recalls in his memoir. But during his tenure as director of the NIH, he was inspired by physicists and mathematicians at a Los Alamos lab who were sharing "preprints" of their papers on the digital platform ArXiv. Varmus pushed for the NIH to start a robust open-access database of scientific papers, as well as a scientific publishing platform, in the hopes of expanding access to the universe of scientific knowledge. Facing resistance from scientists and publishers alike, he scaled back his ambitions, ending up in the summer of 1999 with a voluntary digital repository that guaranteed a one-year buffer between publication in a journal and posting online. Called PubMed Central, it was a landmark achievement. It was also empty. The producers of for-profit journals, it turned out, had little interest in making free copies of their articles globally accessible.
Jump ahead to 2007, when advocates for PubMed Central pushed through a provision in the 2008 appropriations bill that required any scientist funded by the NIH to submit copies of his or her final journal manuscripts to PubMed Central. With that provision in place, the service grew exponentially. Today, some 2.3 million scientific articles are stored on PubMed Central. NIH officials familiar with the site's traffic statistics report that some half million people visit the site daily, downloading on average a million papers each day. Since then, what was once a quirky side project has become a major threat to the academic-publishing business.
PubMed Central has its critics. Its defect isn't, they argue, that it gives the public access to research paid for with taxpayer dollars. "We obviously have no problem with the federal government doing what it can to increase public access to federally funded research," says Adler. "The only question is why they believe it's necessary and appropriate to take what publishers add to researchers' accounts." Critics say that journal publishers add value to the raw materials produced by researchers. Otherwise, the argument goes, why not simply share one's data with or provide progress reports to the public, which the NIH requires scientists to do anyway? Because this isn't just about research, Adler says. It's about the creative product of that research. In the case of the NIH, the federal government is interfering with the relationship between researchers and the publishers that cultivate and promulgate their work. As part of that relationship, critics say, authors are opting to hand over copyright to the journal of their choosing-in part to compensate the work the journals do.
Now, almost a half-decade after public access became the status quo at NIH, publishers are worried that rather than be contained there, it will spread, dragging down for-profit publishing with it.
The Legislative Background
In the United States, the federal government cannot retain copyright on works produced as a function of government. Furthermore, in its 1976 overhaul of copyright law, Congress contemplated whether government should hold copyright on works funded by government. At the time, legislators fretted over allowing a "double subsidy"-that is, American taxpayers paying once to fund research and again to access what that research produced. At the time, legislators declined to resolve the issue and instead provided guidelines: "Where, under the particular circumstances, Congress or the agency involved finds that the need to have a work freely available outweighs the need of the private author to secure copyright," reads the relevant committee report, "the problem can be dealt with by specific legislation, agency regulations, or contractual restrictions."
Under the Constitution, the purpose of copyright is to "promote the Progress of Science and useful Arts." Does the current state of scientific publishing move science forward? It's no empty question. Lawmakers or agency leaders may end up concluding that giving the public open access to government-funded works advances the march of science more than would showing absolute deference to private publishers' copyright claims.
Still, publishers have been pushing back against the NIH policy since it was first enacted. The public pressure of the last several weeks seems to have driven them away from the Research Works Act; Elsevier spoke of its wish for a "less heated and more productive climate for our ongoing discussions." But it's important to note that the Research Works Act was meant to be a firebreak, an attempt to stop the public-access philosophy that has taken over NIH from spreading throughout government.
Now, the real concern of people at the American Association of Publishers has shifted to the White House. As part of the reauthorization of the federal research-and-development America COMPETES Act in January of last year, the White House's Office of Science and Technology Policy (OSTP) was charged with developing public-access policies at other government agencies that fund scientific research such as the National Science Foundation, the Department of Energy, and others. That centralized process is making some publishers nervous. The OSTP process is ongoing and will culminate in guidance given to a multi-agency administration task force on public access.
And Then Came The Internet
With pressure to codify federal policy on academic copyrights, open-access advocates see a thrilling opportunity to use the U.S. governments tremendous leverage-the NIH alone funds some 325,000 researchers and spends $25 billion a year-to tip the scales in favor of a new, more open model for science publishing. Champions of open access say the inability to collaborate and access scientific research freely in the past was a technical limitation-not a principled policy. "Before the Internet, you could have held [universal public access to the scientific literature] as the ideal," says open-access advocate Michael Eisen, a biochemist at the University of California, Berkeley. "But there was no way to achieve it." For the first few centuries of scientific publishing, that is, whether it was the scientist or the journals that held the rights to a journal paper didn't matter much. Publishers controlled paper, and there was no other way to distribute science. The dominant limitation was physical. Giving the scientific community and the general public additional journal copies required printing and disseminating them. Then came the Internet, bringing along with it historic efficiencies in digital reproduction and distribution. Now, says Eisen, "the costs are largely incurred with the first production of a paper."
Publishers push back, arguing that they do more than print-for example, they bring together scientists to do peer review, the backbone of the journal-publishing endeavor, and otherwise ensure the quality of the scientific literature. Eisen challenges that there are software packages now that can streamline the peer-review process, and even innovative new ways to tackle the gatekeeper role traditionally handled by journals, such as post-publication commenting algorithms that quality papers rise to the top.
Those advances, Eisen argues, make plain the role that publishers play, and the problem with journals routinely holding onto copyright. "Publishers provide a service to the community," Eisen explains, "and they should be paid as service providers." Yes, publishers help bring the scientific papers into existence. But obstetricians help bring new life into the world, he argues. They get compensated for that act of labor. What they don't get is the right to lease the baby back to his or her parents. Letting journals keep article copyrights is as crazy as letting doctors keep rights to the child. "Instead of just paying publishers once for the service they provide," Eisen says, "we're allowing them to forever garnish the final product." Where once that distinction didn't matter much, he argues, in the digital age it gives publishers an illogical power over the scientific literature.
Bringing the might of the federal government to bear is welcome, advocates say, because until now redrafting the three-centuries-old dynamic of scientific publishing has fallen on authors-the research scientists. To most of us, "A Poised Chromatin Platform for TGF-β Access to Master Regulators," the name of a recent Cell article, might sound like gibberish. But to scientists, it can sound like a career maker. Placing a paper in a top-tier journal is the key to what matters in the research life: grants, jobs, and university tenure. The stakes are so high that scientists can toil for years to mold an article into acceptable shape for the editors at Cell, or Science, or Nature. The NIH rule, as it stands, puts them in a tight spot. In an attempt to circumvent the copyright battles, the policy requires authors to submit manuscripts of their journal articles-not the articles themselves. It is a bit of a fudge, and it leaves authors to negotiate with publishers whether they're violating the terms of their publishing contract. Researchers expose themselves to the possibility of being shut out from their field's most prestigious journals.
Still, some researchers have decided that open access is worth the career risk. About a decade ago, for example, Eisen and fellow biochemist Pat Brown of Stanford launched the Public Library of Science (PLoS). It's a service where researchers share their scientific works under the Creative Commons attribution-only license-leaving them free to distribute their work through PubMed Central or other channels. PLoS has proved popular; in 2010, researchers submitted more than 20,000 articles. Similarly, some 7,500 researchers have pledged not to work with the publisher Elsevier.
"If people complain," Eisen says of the burden placed on his fellow scientists, "well, it's a sacred duty we have. We get a pretty great deal from taxpayers." Sacred duty aside, taking a hard line on public access is a great deal to ask of individual researchers trying to make a career in the competitive world of science. Advocates for open access argue that, from here on out, the federal agencies that fund science should be more aggressive in establishing an open-access policy so researchers don't have to shoulder the burden-open access should be less a scientist's brave choice more part-and-parcel of taking taxpayer dollars to do science.
In all this, it's tempting to step back and compare the Internet's effects on science publishing to what has happened with the music industry in recent years, with a content-producing industry clinging to an old and outmoded business model rather than innovating. Harold Varmus is Steve Jobs, perhaps, and PubMed Central, his iTunes. But there are real economic differences to keep in mind. Consider that musicians are making music for the masses, that recording labels support musicians with advances and publicity in the hopes of making money off their successes, and that, if all goes well, the artist gets royalties from his or her content. Science publishing is an altogether different proposition. What we're witnessing is a professional community communicating with itself-the audience for the writings of scientists is almost exclusively other scientists. Researchers are neither paid for submitting pieces to journals nor for acting as peer reviewers. Often, it's donated content and labor that is subsidized by government or by universities, the latter of which usually get large government subsidies. (Those same universities are often picking up the hefty tabs for journal subscriptions.) All of that makes it less risky for a company to do business in science publishing than for a record label to back small recording artists while hoping for a Katy Perry. Yet, the model for scientific publishing is a for-profit one, and a quite profitable one at that. According to an article in the Economist last spring, Elsevier's 2010 operating profit margin was 36 percent.
Publishers complain that open access threatens their ability to recoup their investments. Access advocates argue that perhaps the era of reaping so much safe profit is over.
"This new and innovative model appears to be the wave of the future," Issa and Maloney said about open-access science publishing, in their statement announcing the withdrawal of their support for the Research Works Act. But, "the transition must be collaborative, and must respect copyright law and the principles of open access." Issa had maintained that his sponsorship of the legislation was part of a bid to maintain such a balance. Maloney, for what its worth, represents Elsevier-the company has an office in her east Manhattan district-and counts the publisher as the second-largest contributor to her campaign this cycle.
But even with though the Research Works Act, SOPA, and PIPA are no longer viable pieces of legislation, the questions they forced into the public debate cannot be ignored. What is our copyright policy really meant to achieve? Is it working in the Internet age? Whose interests are being served? With the first journal paper about the Internet published some 40-odd years ago now, those questions form the basis for a discussion whose time has come. All of us, as taxpayers and consumers of scientific information, have a stake in the answers.
