According to this week’s Science magazine, there’s some good news and some bad news regarding open access publishing. Which do you want first?
The bad news? OK, here goes.
According to a letter (free access via Sex Drugs & DNA) authored by Michael Stebbins, Erica Davis, Lucas Royland, and Gartrell White (mostly of the Federation of American Scientists), the NIH’s voluntary open access project, PubMed Central, has been a massive failure due primarily to lack of compliance:
The NIH public access policy requests that NIH-supported investigators submit final peer-reviewed primary research manuscripts to the PubMed Central database (PMC) upon acceptance for publication. The policy went into effect 2 May 2005. As of January 2006, only approximately 3.8% of NIH-funded research papers published after 1 May 2005 had been submitted to the PMC repository.
Low compliance only tells part of the story. More than half of the manuscripts available on PMC were published before 2 May 2005. Many reviews and commentaries, which fall outside of the scope of the request, and papers inappropriately made publicly available before the publisher’s public access embargo were also found in the database. This suggests either wide misunderstanding of the policy or deliberate submission of papers falling outside the scope of the database.
The policy also allows posting of papers that differ significantly from the final published version, which has the potential to create intellectual property issues as each public disclosure of the research represents prior art in the eyes of the law. Also, there is no dedicated system to guarantee that corrections made after publication, which can be significant, are made to the author-submitted paper.
The letter points out one bright spot, though. According to an analysis performed by the Federation of American Scientists, there is a positive correlation (at least roughly) between the prestige of a journal and the rate of submission of papers to PubMed Central. At least the papers appearing there should be of relatively high quality, then.
For a discussion on of open access publishing and why it’s important, check out my original post on the topic. Needless to say, though, the state of PubMed Central is disheartening. The letter by Stebbins et al. at least offers a suggestion for how to proceed:
NIH’s faltering experience so far indicates that public access policies must be mandatory and curated if they are to have any chance of success. It would also be wise for there to be a real demonstration of public desire or need before we expand it to other agencies. Unfortunately, this experiment has cost taxpayers money and the NIH credibility.
I would agree that the PubMed Central experiment indicates that mandatory open access policies would be more effective. Although the authors of this letter are critical of it, The Federal Research Public Access Act (S.2695), introduced in May of this year, would begin to address the problem by requiring free public access to most government-funded research within six months of the research’s publication. Although I would also challenge the authors’ final assertion that the need for open access policies remains to be demonstrated, overall the letter gets at the heart of one of the toughest issues facing open access publishing–the difficulties of easing into an open access system in an industry still dominated by the traditional commercial publishers.
Remember, I said that there would be good news as well.
While PubMed Central struggles in the US, open access is making gains in the UK, as a news article (subscription required) from the same issue of Science reports:
The open-access movement chalked up a victory in Britain last week, but it did not get the universal mandate for free release of research papers that some advocates want. In a long-delayed policy statement on 28 June, the executive board of Research Councils U.K. (RCUK), an umbrella organization for government funding bodies, said that all peer-reviewed journal papers produced by publicly funded research must be made available for free soon after they’re completed. Exactly what that means was not specified, and RCUK left each research council to set its own rules. In coordinated announcements, some set out hard-edged policies whereas others said they were still debating what to do.
The responses have varied, from the Engineering and Physical Sciences Research Council–which will wait for the results of a pending study before making any hard decisions–to the Medical Research Council (MRC)–which will now require mandatory open access to publications supported by its grants within six months of their publication. Interestingly, the MRC will be using PubMed Central to provide that open access:
The most stringent policy came from the Medical Research Council (MRC), source of roughly $400 million in annual biomedical grants. It declared that any papers accepted by a peer-reviewed journal must be deposited “at the earliest opportunity–certainly within 6 months–in PubMed Central,” the free public archive run by the U.S. National Library of Medicine. A mirror U.K. archive is under construction, MRC says, and when it is ready all MRC-funded papers will go there. The rule applies to grants awarded from October 2006 onward, and an MRC spokesperson explains that the 6-month clock begins to run from a paper’s publication date.
This is a major step forward for open access, one that the US should pay attention to. In the meantime, all eyes will be on the UK to see how the new policies play out, particularly when it is scheduled to be evaluated two years from now.