Editor’s Note: Fred Dylla is the Executive Director Emeritus of the American Institute of Physics. With more than 30 years experience as a practicing scientist, Dylla has also taken a leading role as an advocate for the physical sciences. He serves on the Board of Directors for CHOR, Inc., the not-for-profit behind the CHORUS service described below.
February 22, 2016 marked the third anniversary of the United States Office of Science and Technology Policy (OSTP) memorandum that continues to have significant impact on the communication of research funded by the US government. President Obama’s Science Advisor, John Holdren, issued the memo as a directive to the major US federal research funding agencies to develop and implement plans for public access to the results of that research — focusing on publications and data. Three years in, it’s worth noting the remarkable progress that has been made. It is also a useful time to assess how agencies and the entire research ecosystem, including the publishing community, have reacted to the directive.
As with any issue dealing with our government, there are those who feel that not enough has been done and progress is too slow. On the other side of the spectrum, there is sentiment that too much is expected from our government agencies during an era where flat budgets are the best outcome from the annual tussle in Congress for the appropriations for these agencies from the public treasury.
From my perspective, much of the criticism of the OSTP memo and subsequent agency plans is misguided. Complaints that these are not “open access” plans miss the point – this is not, nor was it ever meant to be a set of “open access” plans. The phrase “open access” never appears in the OSTP memo, and the phrase “public access” was instead deliberately used.
Background: How did this policy come about?
The OSTP directive has a long pedigree, and it was carefully developed with significant community input. Many of its key components stem from the recommendations of the Scholarly Publishing Roundtable, which was held in 2009-2010 under the auspices of US House Science and Technology Committee with guidance and input from OSTP. The Roundtable included representatives from research universities, libraries, commercial and non-profit publishers. The group was given time and latitude to carefully consider all the issues behind the imperative of providing wider access to the results of publicly funded research. The Roundtable recommendations were used by Committee members and staff to help inform policy for OSTP and federal agencies on public access that subsequently appeared in Section 103 of the America COMPETES Reauthorization Act of 2010. This bill was passed by both Houses and was signed into law by President Obama in January of 2011. The OSTP memo was the result of the process directed by this legislation.
The COMPETES Act called on OSTP to establish a working group under the National Science and Technology Council (NSTC) to coordinate the development of public access plans. The working group was asked to: list specific objectives and the public interest that would be addressed by any proposed policies; account for the variability among the funding agencies and their respective academic disciplines; and develop standards for the identification and interoperability of text, data and the metadata for both entities, while being fully cognizant of existing international standards.
As these policies were being developed, the working group was asked to “solicit input and recommendations from, and collaborate with non-Federal stakeholders” where the public, universities, nonprofit and for-profit publishers and libraries were explicitly listed. The responsibilities of the NSTC working group were eventually transferred to the OSTP. For the next two years (2011-2012), OSTP solicited stakeholder input through two official “Requests for Information” announcements and entertained the expected office visits from many stakeholders making their case (including publishers and librarians).
On February 22, 2013, the long awaited OSTP directive to the agencies was distributed to agency heads and posted publicly on the OSTP website. When I first read the memo, it was clear why it had such a long gestation period. In a mere six pages, it covered the essential ground work on giving agencies a blueprint for developing their plans, and it was very carefully worded to encourage flexibility on some of the obvious hot-button issues: embargo lengths for access to free public text, re-use capabilities, working with the private sector, costs for implementing the plan, and the differences between publications and data.
A key requirement of the OSTP directive was the need for each agency to ensure that free versions of peer reviewed articles tagged to their research funding be made available to the public after an embargo period. The purpose was to enable members of the public to “read, download and analyze in digital form” either the publisher’s version of record or the author’s accepted manuscript. Metadata was to be made available immediately upon publication and some means of identifying the articles tagged to public funding had to be provided. Agencies were cautioned not to expend additional funds on implementing their public access plans and were encouraged to take advantage of public-private partnerships to leverage their resources.
The publishing community provides a key tool.
While Congress and the Administration were working on the COMPETES Act and the OSTP memo, the publishing community was undertaking innovative projects that would ultimately assist agencies in implementing their final plans. The development of a new funding identifier for articles was one vital new tool. The key problem of adding additional metadata to articles to identify the funding agency behind the described research was solved by Crossref ‘s FundRef initiative (now called Crossref‘s Open Funder Registry). This initiative was born from a modest presentation I gave on the need for such identifiers at Crossref’s Annual General Meeting in November 2010, and led to a successful year-long pilot project run by Crossref with seven publishers and four funding agencies from mid-2012 to 2013.
This new article metadata, identifying articles tagged to agency funding has enormous potential for widespread use throughout the research community, and it is already proving its value by enabling the service that many funding agencies are using to provide public access.
Public access progress to date.
In a communication published by OSTP on the three-year anniversary of the public access directive, Jerry Sheehan, OSTP Assistant Director for Scientific Data and Information, noted that 16 agency plans had been published, and there were several agency plans nearing approval. He further observed that the posted plans already account for 98% of the research funded by the federal government.
In addition to the agencies required to provide plans responding to the OSTP directive — those with annual research budgets exceeding $100 million — several smaller agencies, such as the Smithsonian Institution have also posted their plans. Also, more than 20 agencies have participated in periodic intra-agency coordination meetings hosted by OSTP since the directive was published.
Although the OSTP set a six-month deadline for agencies to submit draft plans, the review process within agencies and between agencies, OSTP and OMB, meant a longer timeframe for actual announcement of plans. The Department of Energy (DOE) was the first agency out of the gate with having its plan approved and publicly posted in August of 2014. DOE, at the time of announcement of its plan, also acknowledged a partnership with CHORUS — the non-profit organization founded by scholarly publishers to assist agencies with the implementation of their public access plans. The Department of Defense released its plan in February 2015, and the National Science Foundation followed in the next month with the release of its plan. These three agencies combined represent the largest share of non-medical research and associated published research in the US. The fact that all three agencies are using a version of the DOE-designed PAGES portal — and PAGES uses the CHORUS service to link out to article content on publisher platforms — adds some needed coherence for this large segment of publicly tagged content.
At this point, six agencies are using CHORUS, which went live in August 2014 and as of this posting currently monitors approximately 209,000 articles (latest stats available here) for public access, reuse license terms, and long-term archiving and preservation. More than 46,000 of the articles CHORUS tracks are already publicly accessible. CHORUS is a no-cost service for agencies that provides access to an important collection of articles and its use no doubt will grow as experience is gained by all stakeholders. Several agencies, which have chosen other platform solutions are engaged with pilot projects with CHORUS. Agencies under the Department of Health and Human Services (DHHS), the National Aeronautics and Space Administration, the Department of Veterans Affairs, the Environmental Protection Agency and the National Institute of Standards and Technology are using the NIH’s PubMed Central platform, and five are using their own or other repositories.
The significant jump in public access to peer reviewed scholarly publications is the key accomplishment of the OSTP initiative so far.
This success rests on a clear understanding of the essential sustainability of the publishing enterprise that underwrites the publication of research results. The 12-month embargo was identified by OSTP as a period agencies should use “as a guideline for making research papers publicly available”. The OSTP memo allowed for flexible embargoes to ensure that public access policies do not impact the primary source of income for publications that rely on subscriptions.
Publishers argue that one single embargo length is not appropriate for all fields of study. Twelve months may work for the well-funded and fast-moving biomedical fields, but may not be appropriate for some fields, whose associated research articles have longer usage half-lives (basically everything other than biomedical research, from mathematics to the humanities). Access proponents argue for shorter or no embargos, completely disregarding the current economics underlying scholarly publishing, and the OSTP’s decision not to require the ring-fencing of specific grantee funds to pay for Article Processing Charges (APCs) for open access articles. The embargo controversy will no doubt continue, but the OSTP directive and most of the published agency plans have left open the option of evidence-based petitioning of agencies for embargo changes.
Another topic of conversation within the community is the extent that the directive required mechanisms to be put in place to enable full text- and data-mining (TDM) of publicly available manuscripts. The directive encourages public-private collaboration[s] to “maximize the potential for interoperability between public and private platforms and creative reuse”.
The OSTP directive encourages and provides the incentive for TDM to evolve and grow as an important research tool, from its present rather nascent state. The Scholarly Publishing Roundtable observed that new digital tools would be valuable for scholars. But, in any fast-moving field, one doesn’t want to over prescribe how it should be done, as that may impede progress, and requirements may be an impediment to unencumbered innovation. Mindful of this, OSTP’s directive did not call for any specific licensing terms for papers (again, this is a “public access” policy, not an “open access” policy).
Many US funding agencies have chosen not to directly address TDM at all in their plans. However, publishers have driven a number of TDM initiatives that encourage experiments and exercises over collections of documents from many platforms — precisely the broad brush a scholar needs to gather results from the widest data set-rather than limiting TDM to searches on one platform. Both Crossref and the Copyright Clearance Center (CCC) have instituted TDM systems, and many publishers including Elsevier and Wiley allow the Crossref, CCC and other approved tools into their systems for TDM activities.
The OSTP directive also called on agencies to set-up mechanisms for public accessibility and archiving of data generated from publicly funded research. This effort is far less mature within the research community compared to publications. Some of this is due to the lack of standards and formats for data sets, the great variability of types of data, as well as additional complications of dealing with proprietary data, confidentiality and security concerns. Despite these complications, approximately half of the posted agency public access plans ask researchers to submit “data management plans” as first steps. The OSTP directive has set in motion activities by standards organizations such as NISO and RDA, work between data citation services such as DataCite and Crossref, the SHARE Notify system, and publisher initiatives around the publication of data sets and linking of data sets to publications.
The bottom line (as of today).
So three years out from the directive, 1) public access policy is in place for 98% of the research funding from US federal agencies starting in the last year, 2) a robust article identification system is in place from Crossref that is already tracking more than 11,000 funding agencies worldwide, 3) CHORUS, a public-private partnership, is actively assisting the agencies with implementing their public access plans, 4) TDM solutions are beginning to appear, and 5) agencies, supported by various stakeholders, are making some headway on data management.
These results are remarkable, given the slow pace of government, the large number of stakeholders and the complex systems involved. It is difficult to find other examples where the US Congress, the Administration, and the private sector worked in concert to get something done that is so helpful to so many citizens in just three years.