Comments on the weakening of the NIH public-access policy
SPARC Open Access Newsletter, issue #82
February 2, 2005
by Peter Suber
The day after I published this article, the NIH released the final version of its public-access policy. However, the final version of the policy is essentially the same as the version I wrote about here. The NIH was set to announce the final language of its public-access policy on January 11, but abruptly postponed the announcement. One theory, published in several newspapers, was that Mike Leavitt didn't want any controversial announcements in the days leading up to his confirmation hearings on January 18-19. Leavitt was the White House nominee to head the Department of Health and Human Services, the cabinet-level department that contains and oversees the NIH. Some government insiders support this theory, but others report that Leavitt had nothing to do with the postponement. (Leavitt was confirmed on January 26.)
The more important story is how the policy had changed. The version that NIH would have announced on January 11 differed in some significant ways from the September 3 draft that was the subject of public comments. The September 3 draft asked NIH-funded authors to deposit copies of their journal articles in PubMed Central (PMC) as soon as the articles were accepted for publication. PMC would then provide free online access to those copies six months after publication. The PMC copy would only be released sooner than six months with the publisher's consent.
The "January 11" version reportedly made three important changes. First, it let authors decide when the PMC copy would be made freely available to the public, from the time of publication up to 12 months later. Second, it eliminated the need for publisher consent if authors chose PMC release sooner than six months. Third, it added an explicit NIH exhortation that early release was better than later release. (The language of the January 11 version has not been made public. This description is based on published interviews with the principals and insider reports.)
Here are some notes on the recent revisions to the policy.
* The NIH is retreating. It had its reasons, and there may be some kind of silver lining. But even after exploring both (see below) I have to conclude that the NIH has weakened its policy and that the weakening is unjustified and harmful.
* This is just the latest in a series of concessions to publishers that take us further and further from the public interest in the free and immediate dissemination of publicly-funded medical research. The original House of Representatives instructions to the NIH already included the six month embargo on public access, a compromise with the public interest whose only rationale was to preserve the revenue stream of publishers. But at least the House asked the NIH to "require" free online access to articles based on NIH-funded research. In the September 3 draft, the NIH softened the requirement to a request. Congress endorsed this softening in the November 20 conference report from the House and Senate appropriations committees. But even then, the conference report stood by the six-month embargo. The House also wanted immediate public access to any papers for which NIH paid part of the publishing costs, but the NIH removed this provision in the September 3 draft as well.
* The chief problem with the January 11 version of the policy is that free online access could be delayed up to 12 months after publication. This is a significant delay, more serious in biomedicine than in most other fields. It will slow down research and slow down advances that promote public health.
* It's even worse. The 12 month figure is an illusion. Since deposit is now voluntary, not mandatory, some authors will never deposit their work, especially if their publishers demand that they don't. This creates two problems: the delays could be indefinitely long, not limited to 12 months, and the NIH is deceiving itself and the public by suggesting that there is a 12 month deadline when there isn't any deadline at all. The policy is not only a retreat from the previous policy, but a retreat from clarity and coherence.
* There's a potential up side here. The NIH will exhort grantees to ask for early or immediate public release through PMC, and some grantees may do so. In a piece in the January 21 Washington Fax (not online), Janet Coleman offers this quotation from Dr. Zerhouni:
We're going to tell the scientists, "look, you have the right to specify when your paper can be made public by NIH. You can tell us right away, three months, six months, nine months. If you scientists feel it is going to damage your society or scientific publishing or your relationship with your publishers, then you can go up to 12 months." We expect 12 months to be the exception, not the rule....I'm convinced that people will see this as win-win.One way to read this is that Zerhouni believes that the new policy will provide free online access sooner than the old policy. The language may allow delays beyond six months, but the average delay will be shorter than six months.
This might happen, and critics of the recent policy revisions must admit it. Moreover, if it does, that will not only be good, but a better outcome than we would have had under the September 3 version of the policy.
Authors could always get the same effect through self-archiving outside PMC, but they've been slow to do so (more on this below). The NIH exhortation, tied to NIH funding, may move authors to act.
* While the NIH exhorts authors to choose early release, many publishers will demand that authors choose late release or even exercise their option to deny the request and never deposit in PMC at all. Leaving the decision up to grantees invites publishers to make their own preferences known to grantees.
How will authors deal with these conflicting pressures? We just don't know. There are benefits in pleasing one's funder, just as there are benefits in pleasing one's publisher. There are risks in snubbing one's funder, just as there are risks in snubbing one's publisher. For many authors, the dilemma will be painful and career-jeopardizing.
So the outcome could be better than the six-month embargo in the previous policy or it could be worse. On this point, I'll be an empiricist and wait to see what happens. Or I'll be a hopeful, Heisenbergian empiricist trying to affect what I observe. (There are two avenues for changing the outcome: Work hard to encourage authors to take advantage of the early access option, and work hard to get the NIH to revise the policy to mandate early access.)
In short, I'm not criticizing the policy because I predict that the average delay will be longer than six months. That's up in the air. I'm criticizing it because it invites publishers who dislike the policy to voice a preference contrary to the NIH's preference and (to that extent) because it creates an untenable, high-risk dilemma for authors. I'm criticizing it because it makes public access depend on the behavior of conflicted grantees, not the terms of the policy. I'm criticizing it because it shows the NIH failing to live up to its responsibility to taxpayers.
The NIH could have prevented the dilemma from arising by following the instructions of the House and mandating public release through PMC. Last February a JISC/OSI study (pp. 56-57) showed that an overwhelming majority of authors say that they would willingly abide by an OA mandate from their funder or employer.
http://www.jisc.ac.uk/uploaded_documents/JISCOAreport1.pdf
* It's important that we don't know, and the NIH doesn't know, how authors will respond to these conflicting pressures. This fact weakens the NIH's stated justifications for the policy: that there should be public access to all NIH-funded research, that NIH needs a single source for all of its funded research in order to support interactive links with GenBank, ChemBank, and other NIH databases, and that public funding agencies should have a single portfolio of all their funded work for efficient searching, management, and preservation. These justifications only work if all grantees must deposit their work by some firm deadline. Remove the requirement and the deadline, and all that remains is exhortation, hope, and self-deception.
Another way to put this: The recent backward steps violate the NIH's own criteria for the policy.
* Behind all of this, thank goodness, authors still have the option of postprint archiving in repositories other than PMC. NIH-funded authors are as free to self-archive as any other authors. About 70% of surveyed journals already allow authors to self-archive their postprints, at least at personal home pages or institutional repositories, and there is no evidence that biomedical journals are less accommodating in this respect than other journals. If authors take advantage of this option, then they will largely bypass the new deficiencies of the NIH policy. There will still be obstacles on the road that NIH is building for public access, but authors will be able to reach the public through another road.
But there's the usual snag: authors are unaccountably slow to take up self-archiving. The cause is not opposition to open access, but lack of time and lack of knowledge. We've all been working hard to educate authors about self-archiving and create incentives for them to lift a finger. This is another reason to keep at it.
One way we've sought to create incentives is through conditions on research grants. The NIH was very close to being the pioneer here. But now it's clear that authors who need a funder nudge for self-archiving are likely to feel an equal an opposite nudge from publishers.
If authors seize their opportunity for self-archiving, will they feel pinched by the conflicting pressures from their funder and their publisher? No. The timing of the PMC release is inconsequential to any author who has decided to provide immediate open access through an institutional or disciplinary repository. While it's important to remember this --if only as one more reminder that authors should self-archive-- it's also important to remember that this relief applies to self-archiving authors. If most don't self-archive, then most are vulnerable to the funder-publisher dilemma.
* Under the September 3 version of the NIH policy, PMC would not release deposited articles sooner than six months after publication unless it had the publisher's consent. The January 11 version removes this restriction, which is a forward step that apparently neutralizes at least some of the backward steps. But in fact authors only needed publisher consent for early public access through PubMed Central. They never needed it for self-archiving outside PMC.
So we're back at the same fork in the road: the policy change affects self-archiving authors differently from non-archiving authors. For authors who self-archive, the original requirement for publisher consent was an illusory barrier and its removal is an illusory gain. For authors who don't self-archive, the publisher consent requirement was a real barrier and its removal is real progress.
We're also back to another Heisenbergian phenomenon: whether NIH-funded authors are likely to self-archive is not independent of the NIH policy. The NIH policy might function as an archiving incentive where other incentives have failed, even while it opens the door for publisher incentives to pull in the opposite direction.
This makes it very hard for observers to assess the weight of conflicting incentives. But it also makes it very hard for grantees to do so. The policy didn't have to be this complex, messy, and indefinite. Until the last month or so, it wasn't.
* The policy gives journals an opportunity to ask their authors not to comply with the NIH policy at all or to request late release from PMC. But if most journals permit postprint archiving (indeed, with no delay at all), or they provide free online access to their back run after some embargo (often six months), then will they use this opportunity to influence authors or will they let the NIH exhortation stand unopposed?
Again, we just don't know. They answer is very likely "some will and some won't" and the proportion will have to be observed after the fact.
But the fact that many journals already have access policies equivalent to the NIH policy hasn't prevented them from opposing the NIH policy. Their reasons are difficult to discern. (In the November SOAN, I argued that the real issue was control, not access.) But the same bundle of reasons that leads them to object to the NIH policy might lead them to ask their authors not to participate in it or to participate as late as possible.
Just last month Nature adopted the two-edged new policy to encourage author self-archiving six-months after publication (going beyond mere permission) but to wait six months (introducing a self-archiving embargo for the first time). If more journals adopted the first half of Nature's new policy, then no NIH-funded authors would face the career-jeopardizing dilemma I've been describing. But if more adopted the second half, then publishers will oppose not only the author's funder, but the author's self-interest in visibility and impact. (More on the new Nature policy in the Top Stories section, below.)
* Why did the NIH retreat? In the same Janet Coleman piece from the Washington Fax cited above, we read this:
NIH's inability to quantify the potential economic impact on industry ultimately led to the agency's current proposal for a more flexible posting timeframe. Zerhouni noted that comments on the draft proposal asked, "Can you prove that you'll be economically non-damaging to all the entities? Does a one-size-fits-all, date-certain policy work?" What NIH realized, he said, "is that you can't prove it, and by law if you can't have an economic impact analysis, you really can't implement a policy." The choice, Zerhouni explained, was whether to take two or three years to do an economic analysis with the danger that "you can never really get to the answers" or eliminate to the six-month deadline.This suggests that NIH was under some kind of legal requirement, perhaps to write an economic impact statement before creating any risk of economic harm to private-sector publishers. However, when SPARC Director Rick Johnson asked an NIH official to explain the quotation from the Washington Fax, the official said that Dr. Zerhouni was misquoted and never suggested that there was a legal requirement.
* The NIH might wish to avoid or minimize economic harm to private-sector publishers even if it is under no legal requirement to do so. On several occasions, Dr. Zerhouni has said that this was a concern of his. But this concern does not justify the policy retreat --or at least not yet and not in this form. The NIH could have proceeded with the six-month embargo policy and collected evidence of its impact on journal subscriptions. If the impact was sufficiently harmful, then NIH could have revised the policy in light of the evidence.
For example, if the evidence showed that the six-month embargo harmed quarterly journals more than monthly journals, then NIH could have extended the embargo period "surgically" for quarterly journals. If the policy didn't harm any journals significantly at all, then NIH could continue with the superior access model and perhaps even shorten the embargo period. By acting in advance of the evidence, the NIH gave all publishers an out when only some could claim to need it. More importantly, the NIH gave the benefit of the doubt to publishers rather than the public.
My position has always been twofold: (1) that there are good reasons to think that the six-month embargo policy would not undermine journal subscriptions and (2) that even if it did undermine journal subscriptions, the policy would be justified. I can understand why the NIH might not accept the second half of this position. But the NIH had a chance to test the first half, and accept or reject it based on evidence. It passed on the chance. Even this might make sense if the NIH were convinced that the six-month policy would harm journals. But that is apparently not the case. Four days after the postponed January 11 announcement, Jane Griffith, NLM Assistant Director for Policy and Legislative Development, made this assertion in a presentation at the SPARC-ACRL Forum at the ALA Midwinter Meeting (January 15, 2005): "We [at the NLM and NIH] are not aware of evidence that indicates that libraries and individual subscribers are likely to cancel subscriptions because of the NIH policy."
http://www.arl.org/sparc/meetings/ala05mw/Jane_Griffith.htm
* Many of the publishers who object to the NIH policy already provide free online access to their back issues six months after publication. If we wait until 12 months after publication, then the percentage of complying publishers rises much higher. Hence, the recent shift moves the policy substantially away from an access-enhancing new practice, and substantially back toward the status quo. Yet Dr. Zerhouni was widely quoted last summer and fall for telling publishers that the status quo was unacceptable.
Here's another quotation. Last month, Dr. Zerhouni told Lila Guterman of the Chronicle of Higher Education that the 6000+ public comments on the September 3 version of the policy (with a six month embargo) were "overwhelmingly supportive".
http://chronicle.com/prm/weekly/v51/i18/18a02801.htm
http://www.earlham.edu/~peters/fos/2005_01_02_fosblogarchive.html#a110478242790846793
The new policy is weaker than what the House or House-Senate Conference Committee requested. And it's weaker than what the public overwhelmingly supported when asked.
* The lesson for other funding agencies --inside the U.S. and out-- is not to follow the lead of the NIH. By all means give taxpayers open access to publicly-funded research. But improve upon the NIH policy in at least two critical respects. (1) Make OA a simple condition of funding, a requirement of the grant contract, not a "request" with unspecified sanctions that creates the risk of non-compliance from grantees and the risk of countervailing pressures on researchers from publishers. (2) Shorten the delay between journal publication and mandated OA, if not to zero then to six months at the most.
There are a few other, less urgent ways to improve upon the NIH policy. (3) Lift usage restrictions to permit full open access, not just free online access limited to "fair use" or "fair dealing". (4) Allow deposit in any repository that meets certain conditions of accessibility, interoperability, and long-term preservation, rather than requiring deposit in a central archive. (5) Extend the policy beyond published journal articles to the data underlying those articles.
Finally, the NIH itself should not regard the current compromise as the final state of the policy, regardless of the political pressures that make it seem necessary or desirable today. Over time, the NIH should aim to strengthen the policy and better serve the public interest in medical research and health care.
* The NIH policy still hasn't been released and may still be in flux. The time created by the January postponement gives both sides time to lobby for further revisions. Moreover, during Mike Leavitt's confirmation hearings, more than one Senator urged him to scrap the 12 month deposit period and restore the six month embargo. It's too early to know the effects of these initiatives. But for the same reason, it's too early to resign ourselves to the version of the policy that would have been announced on January 11.
The final policy may be announced shortly after this issue mails, or it may not be announced until spring. If it's better than I've painted it here, then I'll praise the improvements. I'm criticizing the backward steps in advance of the announcement, however, in order to do all that I can to prevent or reverse the retreat and to persuade other funding agencies who might be watching the NIH not to take the same backward steps.
To lend your weight to this effort, send your thoughts to the following policy-makers in Washington. If you write before the NIH announces the new policy, then explain why the NIH should not make this mistake. If you write afterwards, then emphasize that the NIH policy must be evaluated on percentage of NIH-funded research that finds its way into PMC for public access and on the average delay between journal publication and PMC access.
Secretary Mike Leavitt
Department of Health and Human Services
200 Independence Ave., S.W., Room 615-F
Washington, D.C. 20201
Phone 202-690-7000
Fax 202-690-7203
House of Representatives Subcommittee on Labor, Health and Human Services, and Education (message to the chair or any members)
http://appropriations.house.gov/index.cfm?FuseAction=AboutTheCommittee.Subcommittees&SubcommitteeId=11
Senate Subcommittee on Labor, Health and Human Services, and Education (message to the chair or any members)
http://appropriations.senate.gov/subcommittees/laborhhs.htm
* Here are some links to resources and recent news stories.
NIH page on the public-access policy, including its own FAQ
http://www.nih.gov/about/publicaccess/index.htm
My FAQ on the NIH public-access policy
http://dash.harvard.edu/bitstream/handle/1/4552059/suber_nihfaq.htm
Transcripts of the Mike Leavitt confirmation hearings from the Federal News Service (not free).
http://www.earlham.edu/~peters/fos/2005_01_23_fosblogarchive.html#a110668150900209252
January 18, 2005
http://www.fnsg.com/transcript.htm?id=20050118t5139&nquery=&query=leavitt
January 19, 200
http://www.fnsg.com/transcript.htm?id=20050119t5274&nquery=&query=leavitt
Janet Coleman, "NIH Public Access Policy Gives Authors Posting Discretion Up To 12 Months", Washington Fax, January 21, 2005.
http://www.earlham.edu/~peters/fos/2005_01_16_fosblogarchive.html#a110634007993066795
Jeffrey Young, "HHS Nominee Leavitt Backs NIH Public Access 'Principle' At Senate Hearing", Washington Fax, January 21, 2005.
http://www.earlham.edu/~peters/fos/2005_01_16_fosblogarchive.html#a110633579386526070
Erika Check, All parties on edge as NIH delays open-access briefing, Nature, January 20, 2005.
http://www.nature.com/cgi-taf/DynaPage.taf?file=/nature/journal/v433/n7023/full/433182b_fs.html
http://www.earlham.edu/~peters/fos/2005_01_16_fosblogarchive.html#a110616656235902533
Lila Guterman, NIH Reportedly Is Weakening Its Plan for Free Access to Journal Articles, Chronicle of Higher Education, January 19, 2005.
http://chronicle.com/prm/daily/2005/01/2005011901n.htm
http://www.earlham.edu/~peters/fos/2005_01_16_fosblogarchive.html#a110616112950115170
Paul Revere, Implausible deniability at NIH, Effect Measure, January 18, 2005.
http://effectmeasure.blogspot.com/2005/01/implausible-deniability-at-nih.html
http://www.earlham.edu/~peters/fos/2005_01_16_fosblogarchive.html#a110615242661668983
Jocelyn Kaiser, NIH Revises Public Access Policy, Science Magazine, January 18, 2005.
http://sciencenow.sciencemag.org/cgi/content/full/2005/118/1
http://www.earlham.edu/~peters/fos/2005_01_16_fosblogarchive.html#a110614891848437588
Rick Weiss, NIH Revises Plan for Quick, Free Access to Study Results, Washington Post, January 18, 2005.
http://www.washingtonpost.com/wp-dyn/articles/A16337-2005Jan17.html
http://www.earlham.edu/~peters/fos/2005_01_16_fosblogarchive.html#a110605414454530190
Dee Ann Divis, The push for public access to journals, Los Angeles Times, January 17, 2005.
http://www.latimes.com/features/health/la-he-access17jan17,1,7155087.story?coll=la-headlines-health
http://www.earlham.edu/~peters/fos/2005_01_16_fosblogarchive.html#a110596898955305883
Linda Watson, Enhanced Public Access to NIH Research Information: Implications for Open Access, Charleston Advisor, January 2005.
http://charlestonco.com/features.cfm?id=172&type=fr
http://www.earlham.edu/~peters/fos/2005_01_09_fosblogarchive.html#a110567539114265529
The Alliance for Taxpayer Access has publicly released its January 11 letter to Dr. Elias Zerhouni expressing disappointment at the delay in the announcement of the NIH public-access plan.
http://www.eurekalert.org/pub_releases/2005-01/wc-aft011305.php
http://www.earlham.edu/~peters/fos/2005_01_09_fosblogarchive.html#a110566608131947379
Ted Agres, 'Open access' announcement scuttled, The Scientist, January 13, 2005.
http://www.biomedcentral.com/news/20050113/02/
http://www.earlham.edu/~peters/fos/2005_01_09_fosblogarchive.html#a110564067556285904
Bradie Metheny, NIH Public Access Publishing Policy Release Postponed, Washington Fax, January 12, 2005 (not online).
http://www.earlham.edu/~peters/fos/2005_01_09_fosblogarchive.html#a110555692263463210
Lila Guterman, Critics and Proponents Debate NIH's Plan to Free Access to Scientific Materials, Chronicle of Higher Education, January 7, 2005.
http://chronicle.com/prm/weekly/v51/i18/18a02801.htm
http://www.earlham.edu/~peters/fos/2005_01_02_fosblogarchive.html#a110478242790846793
* Postscript. I don't want to be misunderstood. I'm arguing that the January 11 version of the NIH policy is worse than the September 3 version. I'm not arguing that the January 11 policy is worse than nothing. Even the watered down version of the policy will be an advance over the status quo, though a smaller advance than we had been led to expect. We'll still see free online access to some NIH-funded research after some delay. Reducing the requirement to a request means that some will be missed. Allowing delays longer than six months, even with an exhortation for earlier release, means that some will be delayed longer than six months. Since the body of NIH-funded research is very large and very high in quality, even delayed free access to a subset is better than toll access to the totality. Moreover, the policy can always be strengthened, even if that takes another year of working with Congress.
----------
Read this issue online
http://dash.harvard.edu/bitstream/handle/1/3997176/suber_news82.htmlSOAN is published and sponsored by the Scholarly Publishing and Academic Resources Coalition (SPARC).
http://www.arl.org/sparc/Additional support is provided by Data Conversion Laboratory (DCL), experts in converting research documents to XML.
http://www.dclab.com/public_access.asp
==========
This is the SPARC Open Access Newsletter (ISSN 1546-7821), written by Peter Suber and published by SPARC. The views I express in this newsletter are my own and do not necessarily reflect those of SPARC or other sponsors.
To unsubscribe, send any message (from the subscribed address) to <SPARC-OANews-off@arl.org>.
Please feel free to forward any issue of the newsletter to interested colleagues. If you are reading a forwarded copy, see the instructions for subscribing at either of the next two sites below.
SPARC home page for the Open Access Newsletter and Open Access Forum
http://www.arl.org/sparc/publications/soan
Peter Suber's page of related information, including the newsletter editorial position
http://www.earlham.edu/~peters/fos/index.htm
Newsletter, archived back issues
http://www.earlham.edu/~peters/fos/newsletter/archive.htm
Forum, archived postings
https://mx2.arl.org/Lists/SOA-Forum/List.html
Conferences Related to the Open Access Movement
http://www.earlham.edu/~peters/fos/conf.htm
Timeline of the Open Access Movement
http://www.earlham.edu/~peters/fos/timeline.htm
Open Access Overview
http://www.earlham.edu/~peters/fos/overview.htm
Open Access News blog
http://www.earlham.edu/~peters/fos/fosblog.html
Peter Suber
http://www.earlham.edu/~peters
peter.suber@earlham.edu
SOAN is licensed under a Creative Commons Attribution 3.0 United States License.
http://creativecommons.org/licenses/by/3.0/us/
Return to the Newsletter archive