With the recommendations of an advisory committee on scientific publishing in hand, the White House Office of Science and Technology Policy (OSTP) has begun moving toward a policy that will require all federal agencies to provide free access to all scholarly articles based on the research they fund. One outstanding question that will have major economic ramifications for the publishers of scientific journals is just how long an exclusivity window publishers will have before significant contents of their journals become freely available on online platforms.
Late in January OSTP staff began poring over the 500 or so comments they received in response to the agency’s request for public input on the issue of “public access,” as it is called. OSTP also is considering a 30-page report (http://science.house.gov/press/PRArticle.aspx?NewsID=2710) crafted by the advisory committee convened by the House Committee on Science and Technology. Known as the Scholarly Publishing Roundtable, the panel sought to bridge a divide that has long existed between publishers, which charge thousands of dollars for annual subscriptions to many of their journals, and their subscribers, which typically comprise the libraries of research universities. The journal industry is enormous, with $8 billion in annual revenues worldwide, according to one estimate used by the roundtable. The numbers of journals and their publishers are also immense—25 400 and 2000, respectively, by one estimate. The costs of accessing such a volume of information, the report notes, “are challenging the strained budgets of universities and their libraries,” while the broader public has little access to the literature.
The 14 roundtable participants were drawn from all sides of the issue—librarians, university provosts, academic researchers, and publishers (including Fred Dylla, executive director and CEO of the nonprofit American Institute of Physics [AIP], which publishes Physics Today). Elsevier, the largest of the commercial publishers, with about 2000 titles, had a seat, as did the Public Library of Science (PLoS), a relatively new publisher that uses an alternative business model known as open access; it provides immediate free access to its journals and charges article authors a publishing fee. The PLoS representative, director of publishing Mark Patterson, said that he could not fully endorse the roundtable’s consensus report, a position that was also taken, for different reasons, by roundtable member and Elsevier vice chairman Y. S. Chi.
A lengthy process
The public-access movement began in earnest more than a decade ago when Harold Varmus, then director of the National Institutes of Health, pushed for the agency, which is by far the largest federal funder of basic research, to offer free online access to all the published work of NIH-funded investigators. The rationale was that taxpayers shouldn’t have to pay to see the results of the research that they paid for. Varmus later cofounded PLoS, which embodies the open-access publishing model, and he continues to chair the board of that nonprofit organization. He is also one of three cochairs of the President’s Council of Advisors on Science and Technology, which works closely with OSTP.
Following years of back-and-forth between NIH and biomedical journal publishers (see Physics Today, December 2004, page 34.), Congress finally mandated public access for that agency. Since mid-2008 all articles by NIH grantees automatically become freely available from the PubMed Central archive operated by the National Library of Medicine (NLM), but not before a delay, or embargo period, of up to one year after journal publication. Other federal R&D funding agencies haven’t followed suit, in part because there is less public demand for non-health-related research results. But other agencies also don’t have a resource like the NLM, which has an annual budget of $350 million. The roundtable report points out that to achieve public access, agencies may need to establish their own public databases in-house or with university libraries, publishers, or other external partners. Either approach will involve collaboration among all parties, the report emphasizes.
The public-access policymaking process begun by OSTP occurs against a backdrop of ongoing pressure from President Obama for agencies to take extraordinary steps to open their data and their decision making process to public scrutiny. An open government directive issued by Office of Management and Budget director Peter Orszag in December, for example, ordered each federal agency to make public within 45 days at least three “high-value data sets.” OSTP complied by providing links to previously unreleased details of federal R&D spending on information technology, nanotechnology, and climate change. Among NSF’s offerings were a Freedom of Information Act report for fiscal year 2008 and a list of its graduate research fellowship awardees from 2000 through 2009. And the US Department of Energy provided seven data sets, including lists of patents resulting from agency-sponsored research and summaries of energy-related scientific projects performed by national laboratories and other facilities since 1995.
For its part, the roundtable implicitly accepts that free public access has benefits and will become ubiquitous throughout government. Indeed, its report invites other research-sponsoring agencies such as NSF, NASA, and DOE to act quickly to put their own policies in place, in consultation with their stakeholders. The group also endorses the use of embargoes, though it is not specific on what the time limit should be. For many science disciplines, the report says, a delay could range up to 12 months. But in some other fields, particularly the humanities and social sciences, longer embargoes may be warranted. Not all roundtable participants agreed. Patterson of PLoS, in a statement that accompanied the report, indicated the embargo was largely responsible for his decision not to back the report. The public, he said, should get “comprehensive public access to the research that they paid for, with no delay and no restrictions on reuse.” The other dissenter, Chi, said he couldn’t support what he termed “an overly expansive role of government” inherent in the panel’s recommendations.
New publishing models
Although journal publishing continues to evolve, the traditional subscription model remains dominant, supporting around 90% of today’s journals, according to the report. In comments he submitted to OSTP on behalf of AIP, Dylla said that author-pay models “are untested on the large scale and may not be viable for scholarly fields in which funding is scarce.” The imposition of a uniform federal policy for all science and technology funding agencies “would not accommodate the specific needs of any given agency, the rapidly changing nature of scholarly publishing, or the unique considerations of the various fields of science and the journals that serve them,” he said.
A House staffer who asked not to be named said that some federal agencies have been moving cautiously on public access in recognition of the broader implications to the scientific enterprise, particularly the financial impact on scientific societies. Diane DiEuliis, assistant director for life sciences at OSTP, said her agency is well aware of the concerns publishers have with turning the NIH requirement into a one-size-fits-all policy for other federal sponsors. Most of the comments “favored some form of expanding public access,” said Rick Weiss, a senior policy analyst at OSTP. Many urged that OSTP tailor the length of the embargo to the discipline, “particularly [to] the rate of change within the field, which in part determines the value of the information held by publishers over time,” he added.
The roundtable report has much to say about other aspects of the publishing business, including advising journal purveyors to seek out new, more flexible business models. It reaffirms the centrality of peer review in journal publishing and urges that all publication platforms, including public-access databases that may be created by agencies, be designed and built to be interoperable. The federal government, the report says, should play an important role in developing standards for applications to navigate and mine the literature from both public and private sources.
Legislative proposals
House S&T Committee chairman Bart Gordon (D-TN), who commissioned the roundtable, said its recommendations “strike a good balance by allowing public access to the results of research paid for with federal funds, while preserving the high quality and editorial integrity of scholarly publishing so critical to the scientists and seasoned science writers on whose expertise we all depend.”
A committee staffer said the report’s mention of the 12-month embargo signaled that it should apply in the “overwhelming number of cases.” While the S&T committee would like to hold a public airing of the issue, she said, it hasn’t decided that legislation is needed. The House panel has focused its attention on renewing the America COMPETES Act, the wide-ranging 2007 law aimed at restoring US technological competitiveness. The act expires on 30 September, and Gordon, one of its principal authors, hopes to extend it before he retires at the end of this year.
In the Senate, a bill introduced last July would establish a blanket requirement for free access to articles from all federally funded research after a maximum delay of six months. But no other action has been taken on the measure, called the Federal Research Public Access Act (S. 1373), despite the fact that its cosponsor, Senator Joe Lieberman (I-CT), chairs the Committee on Homeland Security and Governmental Affairs, to which the bill was referred. There is no House counterpart to that bill.
Once OSTP has completed its review of comments and the roundtable report, Weiss said, findings will be presented to an interagency committee on information and innovation, an entity cochaired by Thomas Kalil, OSTP deputy director for policy, and Andrew McLaughlin, OSTP deputy chief technology officer for internet policy. “From there, we hope to come up with some administration recommendations of how to implement some new policies,” Weiss said, adding that he couldn’t estimate how long that process would take.
