Cites & Insights: Crawford at Large
ISSN 1534-0937
Libraries · Policy · Technology · Media

Selection from Cites & Insights 5, Number 4: March 2005


Library Access to Scholarship

The big story in January was “NIH: Moving Forward,” a “fairly solid step forward for scholarly access.” As of December 2004, it appeared that the National Institutes of Health would move forward with a plan in which investigators doing NIH-funded work would be asked to submit final peer-reviewed copies of accepted articles; the NIH would make them publicly available at PubMed Central after six months. It wasn’t complete open access, but it was a step.

Too big a step, apparently. Once again, NIH leads off Library Access to Scholarship—but with a classic “two steps forward, one step back” situation.

NIH: Less Certainty, Still Progress

Most observers expected NIH to post their new policy in December 2004 or, later, January 11. That date slipped. Meanwhile, AAP’s Professional/Scholarly Publishing Division (PSP) took another whack at the NIH proposal in a November 15, 2004 letter from Pat Schroeder to NIH director Elias A. Zerhouni, reprinted in the Professional Scholarly Publishing Bulletin 5:3 (Winter 2004). Schroeder notes an October 28 meeting between Zerhouni and “our biomedical journal publishers”—and goes on to urge him “to recognize how diverse medical publishing really is.” In boldface, the letter provides this take on NIH’s modest proposal for voluntary delayed archiving:

AAP strongly believes that it is premature for NIH regulations to fix or bias any specific model at this time. More time is needed to see how the many new publishing models being tried evolve in the reader/author marketplace. Government regulation is likely to foster a rigid dissemination system less able to respond to new and enabling technologies.

Schroeder discusses a proposed initiative between PSP and patient advocacy groups “whereby access to original research studies might be provided to patients and their families in an appropriate context”—which raises the question of why hundreds of patient advocacy groups backed the NIH proposal. She claims the proposal raises unanswered questions about “the disruption of useful journal business models, the risk of censorship and the integrity of the scientific record”—pretty much the standard anti-OA claims, wholly lacking in evidentiary support. Worse, she issues this bizarre interpretation of the UK fiasco:

The United Kingdom engaged in such a process and determined that a competitive global publishing marketplace marked by diverse business models and innovation already exists. They concluded that there was no justification to intervene in a way that would support open access over other business models that already disseminate peer-reviewed scientific research.

To put it another way, using reality-based thinking: The findings of the committee that carried out the “deliberate, participatory process” Schroeder calls for were summarily dismissed by the UK government.

It seems important to PSP that NIH “always link to the final, published articles on the individual publisher’s website and not….make articles freely available until after a period of time compatible with the individual publisher’s business model, as determined by that publisher.” The first clause argues specifically against PubMed Central as a repository, for reasons that aren’t apparent to this reader (who would note that articles on publisher’s sites could always be removed from public access if it suits the publisher—or if the publisher goes out of business). The second is, essentially, a plea that NIH do nothing to improve access. The letter also includes the mandatory indirect suggestion that any change in the current system will somehow “adversely impact” the peer review system.


One supposed reason for the delay in NIH’s policy announcement was that a new Secretary of Health and Human Services, Michael Leavitt, was going through confirmation hearings. Open Access News notes a January 21 Washington Fax report that Leavitt assured the Sente Finance Committee that he supports the principles behind NIH’s policy but “knows very little about the specifics.” Senator Wyden (Oregon) commented that NIH was going to “reduce substantially a proposal to make research that the taxpayers have funded available to the country” and urged that the apparent 12-month embargo window in the forthcoming policy be reduced to the original six months—and that it be a requirement, not a request.

Another Washington Fax article that day noted the claim of journal publishers that “some journals, particularly those that publish infrequently, might be put out of business.” Zerhouni still called the new policy a “breakthrough”—“creating for the first time the precedent and the right for a federal agency to have a venue or pathway for its scientists to publish and give access to the public.”

SPARC e-News (December 2004-January 2005) notes a January 18 Washington Post report that the NIH policy “has been scaled back…under pressure from scientific publishers, who argued that the plan would eat into their profits and harm the scientific enterprise.” It also notes a January 11 letter from the Alliance for Taxpayer Access expressing disappointment in the delayed announcement.

February: Policy and early reactions

On February 3, 2005, NIH issued a press release, a policy implementation statement, and a more extended Policy on enhancing public access to archived publications resulting from NIH-funded research, notice NOT-OD-05-022.

The press release notes that the policy is “designed to accelerate the public’s access to published articles resulting from NIH-funded research” and says it “calls on scientists to release to the public manuscripts…as soon as possible, and within 12 months of final publication.” A bit later, Zerhouni admits that the new policy is voluntary. The release notes that PubMed Central is “a stable archive of peer-reviewed research publications…to ensure the permanent preservation of these vital research findings” and that it secures “a searchable compendium of these research publications that NIH and its awardees can use to manage more efficiently…”

The three-page implementation memo is a lightweight call: “Beginning May 2, 2005, NIH-funded investigators will be asked to submit voluntarily to PubMed Central (PMC) the author’s final manuscript upon acceptance for publication…” It defines “author’s final manuscript” as “the final version accepted for journal publication…[including] all modifications from the peer review process” and notes that, at the time of voluntary submission, authors “will specify the timing of the posting of their final manuscript for public accessibility… Posting for public accessibility through PMC is strongly encouraged as soon as possible (and within twelve months of the publisher’s official date of final publication.”

Later, the policy explicitly excludes “book chapters, editorials, reviews, or conference proceedings,” and clarifies that it’s only asking for publications resulting from currently funded projects. As to versions, “the publisher may choose to furnish PMC with the publisher’s final version, which will supersede the author’s final version”—and the publisher can agree to a shorter embargo than the author chose. The new submission policy fulfills the existing requirement to provide publications as part of progress reports—but NIH still wants hardcopy of “submitted but not yet accepted” manuscripts (which don’t go into PMC).

The 14-page Final Policy Statement goes into more detail and includes NIH responses to many of the public comments received. You can find that document at; it’s interesting background. Publishers and other “commenters” raised all the objections you’d expect; the NIH has sound answers in every case. One response addresses the OA issue:

Some commenters believed that the NIH Public Access Policy constitutes an open access model of publishing. The NIH Policy is not a form of publishing; rather, it creates a stable archive of peer-reviewed research publications resulting from NIH-funded research.

Interestingly, although we have PSP proclaiming how much it favors eventual access, “some commenters also noted that the vast majority of journals currently offer no free public access at all, thus arguing that a 6-month waiting time is too aggressive.” Six months is certainly much sooner than “never,” and for that matter so is a year. There’s a lot more here, including the fundamental answer to publishers who assert that “their” copyright is being undermined: To wit, although NIH isn’t relying on it, the government-purpose copyright license gives NIH absolute rights to reproduce, publish, or otherwise use copyrighted works resulting from NIH funding “for Federal purposes, as well as to authorize others to do so.” Do Federal purposes include seeing to it that Federally-funded research is disseminated to the widest possible audience? Why not?

Technically, SPARC Open Access Newsletter 82 (February 2, 2005) came out a day before the policy was published, but Peter Suber had a pretty good idea what that policy would include (since it was pretty much the January 11 policy, just delayed). Suber notes that the NIH is retreating and that “the weakening is unjustified and harmful.” He calls the weakening “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” and that, long as it is, the “12 month figure is an illusion,” since deposit is now voluntary. The upside is that the NIH’s request appears to be a strong one—and that it could result in faster access than the original fixed six-month embargo.

But, Suber notes, “[M]any publishers will demand that authors choose late release or even exercise their option to deny the request and never deposit in PMC at all.” Suber criticizes the policy “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.” Suber provides seven pages of comments and two more of links to related articles; as usual, you should look at his full commentary.

With the policy out, the Alliance for Taxpayer Access issued a release declaring, “The just-announced policy falls short of their expectations and long-standing recommendations.” Key concerns: the policy is entirely voluntary; it lacks any definitive time frame; it puts grantees in “the untenable position of trying to meet the contradictory expectations of their funding agency and their publisher.” Rick Johnson of SPARC notes that the policy isn’t what they hoped for—but they’re eager for it to succeed. Others offer similar comments.

Then there’s the DC Principles gang, which issued a release calling the NIH rule “a missed opportunity,” decrying the “waste of research dollars,” and asserting that NIH “should take advantage of the fact that most not-for-profit publishers currently make all their content…available for free to the public within 12 months.” The release claims the public “would be better served if NIH created an enhanced search engine that works like Google to crawl the journals’ full text articles and link to the final published articles residing on the journal websites”—asserting, with no evidence at all, that this would offer “significantly more assistance to those seeking medical research results than a database of NIH-funded manuscripts can provide.” It goes on to claim that the PubMed Central version will be “an unedited version.” Naturally, the enormous cost of expanding PubMed Central is mentioned several times but never enumerated—since, at $2 to $4 million out of $38 billion, it’s an odd 0.01% sort of enormous burden. That’s “costly and duplicative”—and PMC will somehow “harm the scientific societies” and “put authors at risk of inadvertently violating copyright agreements.”

I’m immediately struck that “most” is not “all” (some nonprofit publishers do not make their content available), that current availability is not assured permanent availability, that nonprofit publishers do not make up the whole of biomedical journal publishing—and that NIH explicitly invites publishers to avoid the “dual version” problem by submitting the final published version.

Peter Suber commented on the release the next day, at Open access news. His comment: “To me this shows that the recent concession to publishers—lengthening the permissible delay past six months—did not reduce publisher opposition, and therefore was not worth making.” Suber also notes some other reactions. The editor-in-chief of the Journal of the American Medical Association said, “I think it’s great. This is nothing new for us. If it’s important, we make it free to everybody in the world and everything [in JAMA] is free after six months. A spokeswoman for the New England Journal of Medicine noted, “Any material that’s six months old or older is available on our Web site to the general public free of charge.”

The Public Library of Science issued a release noting that the policy “could, and PLoS believes should, have been stronger in several respects” but that it still sets an important precedent. PLoS “urges all other funding agencies…to adopt the progressive components of the NIH policy, and to accompany them with stronger incentives for compliance and shorter periods of allowable delay.” PLoS also urged scientists to seize the opportunity and noted the virtues of open-access publications. The other big OA publisher, BioMed Central, also welcomed the announcement and anticipated “that many other funding bodies worldwide will now follow the example set by NIH.” The press release naturally included a mild sales pitch for BMC journals, noting that those who choose to publish in them “are assured that the published version of their paper will be placed in PubMed Central for them, immediately and without any need for additional work from them.”

That’s where it stands. As Suber and others have noted, now it’s a waiting game—to see whether there will be a significant increase in publicly available biomedical literature within the next year or two. NIH-funded research accounts for about 10% of the articles in the 5,000-odd journals indexed by PubMed; that would still be a substantial increase.

Editorial Policy and SOAN

Before proceeding to a few noteworthy items and articles, I should note an ongoing deliberate change in Library Access to Scholarship. As a rule, I plan not to repeat coverage in the SPARC Open Access Newsletter unless I feel the need to add my own comments or unless it’s an integral part of some subtopic I’m covering. I’m probably not the first to suggest that SOAN is effectively the medium of record for OA, but it’s true—and in my experience, Peter Suber’s clear advocacy does not cause him to cover OA-related issues in a prejudicial or biased manner.

If you care about Open Access, you should be reading SOAN. If you’re reading SOAN, you don’t need a redundant summary from Cites & Insights

I’m sure there will be accidental repetitions, and my take on events is frequently different than Peter Suber’s. That’s hardly surprising. My primary interest is finding ways for libraries to free up enough money to maintain healthy monograph budgets and retain specialized indexes and other services; Peter’s primary interest—at SOAN at least—is spreading Open Access while fairly and honestly covering the controversies surrounding OA. The two goals may be complementary (more so for gold OA, less for green OA), but they’re not identical. Now, on with the items that I found intriguing or important—and that I don’t remember Peter covering!

Shorter Pieces

In January, I noted the special issue of Serials Review on Open Access—and that the articles in that issue are freely available at the moment. As Steve Hitchcock noted on the SPARC Open Access Forum (SOAF), that does not mean that Serials Review (in full or for this issue) is gold OA. The articles will become unavailable at some point, at least from the publisher. Unless they’re archived by authors, this is “sample access,” not OA.

In an early January posting on SOAF, George Porter (Caltech) noted some indicators that the cost of scholarly journal publishing might not always be as high as $1,500 or more. Apparently, the cost of IEEE Electron Device Letters comes out to $186 per page—and given that the journal prefers brief manuscripts, that averages out to $750 per article. The editors seem proud that they’ve encouraged “elimination of verbose sections from published materials and consequent improvement in overall quality.”

Also in January, Nature Publishing Group changed its self-archiving policy—in a way that might also be two steps forward, one step back. In 2002, NPG went green OA (of a sort), allowing authors to post their papers on their personal web sites immediately. The new policy allows and encourages authors to submit their manuscripts to the relevant funding body’s archive and to their institutional repository—but six months after publication.

HW Wilson showed some explicit support for gold OA by adding 38 OA journals to its Education Full Text database. That’s significant if you believe—as I do—that professional abstracting and indexing of journal articles is essential to effective access. Quite a few topical indexes already include OA journals, to be sure (I know of at least ten in RLG’s Anthropology Plus, for example). It’s a trend to be applauded.

Malcolm Getz asserts that research libraries can save money—perhaps as much as $2.3 million per year—through OA publishing, in “Open-access scholarly publishing in economic perspective,” Journal of Library Administration 42:1 (2005). I haven’t seen the (39-page!) article yet and may not have occasion to, but it should be interesting. Presumably the set of assumptions is much different from that used for Cornell’s study (see below).

In early February 2005, the Berkeley Electronic Press announced that the University of California’s eScholarship Repository has logged its millionth full-text download. The repository includes working papers and monographs as well as peer-reviewed articles. The press release says UC’s repository “is believed to have been the first institutional repository” to reach the million-download mark. 98% of readership comes from outside the University of California.

Also in early February, the University of Nottingham and University of Lund announced the Directory of Open Access Repositories (DOAR), a new service to “categorise and list the wide variety of Open Access research archives that have grown up around the world.” Lund operates the Directory of Open Access Journals (DOAJ). Given the fractious nature of OA support these days, it’s hardly surprising that some OA advocates labeled DOAR redundant, a label denied by DOAJ principals.

By any measure I can think of, open access is making progress: Millions of articles are in harvestable repositories and there are more than a thousand open access journals. It’s not setting the world on fire, which may or may not happen, but there’s steady, significant progress. Not according to Sir Crispin Davis of Reed Elsevier, however—at least according to a February 18, 2005 story at uk. According to this piece by Philip Aldrick, Davis reported the company’s full-year results, noting that “The ‘open access’ threat to the system of researchers subscribing to Reed’s scientific journals also appeared to diminish. For the first time in seven years, the publishing method lost market share.” Peter Suber finds that statement incomprehensible; so do I. I can’t think of any plausible measure by which OA is doing worse now than it was a year ago. (There’s another terrifying statement in the article if you’re an academic library or consortium with Elsevier’s hands deep in your pockets: “This year will be a good one as the US education market comes out of its cyclical trough.” That’s great, of course, if it’s true—but it sounds as though Elsevier is primed to take every advantage of improved library fortunes to improve its own.)

SOAN 81 (January 2, 2005) includes Peter Suber’s quick review of 2004 in OA. The dozen points cited are all worth reading. He notes that 2004 was the year funders started to at least consider mandating OA archiving for the research they fund, that some universities are starting to mandate such archiving, and that a significant number of subscription-based journals “turned green.” He’s probably right in saying 2004 saw OA move from the periphery to the mainstream—and certainly right that we’re starting to see a variety of studies and reports on the economics of OA. Unfortunately, “2004 was also the year in which some publishers chose…to jack up the belligerence.”

Suber also posted two excellent notes on his FOS site, each of which prints on a single (double-sided) sheet of paper, each of which should be printed and saved by anyone concerned with OA or working with an OA repository. The base URL for both is www. The first, “A very brief introduction to Open Access,” is precisely what it says. Some might argue with the last sentence in Suber’s description of OA repositories: “The costs of an archive are negligible: some server space and a fraction of the time of a technician.” I see nothing to argue with in the description of OA journals, which ends with this inspirational sentence: “There’s a lot of room for creativity in finding ways to pay the costs of a peer-reviewed OA journal, and we’re far from having exhausted our cleverness and imagination.” Append “brief.htm” for this one, which will also point you to Suber’s longer overview of OA.

The second, “How to facilitate Google crawling,” offers specific pointers (prepared with Google’s cooperation) to make it easy for Google to crawl all of an OA repository. It’s not a long or complicated list—ten bullets with two sub-bullets in one case—and it should be easy to carry out. I love this one: “Browse interfaces should be built as a bushy tree with links to actual articles as the leaves.” Append “googlecrawling.htm” for this vital one-sheet document.

Longer Articles

Goodman, David, “Open access: what comes next?” Learned Publishing 18:1 (January 2005): 13-23. (A later version, “what comes after 2004?,” may be available on the web.)

This article “examines the effects that present decisions about open access (OA) will have over the next ten years.” It’s similar to Goodman’s presentation at the Charleston Conference, where he attempts to model likely outcomes of various publishing futures. I’m not sure I fully understand the models, or how Goodman arrives at his projections, but the article and charts are decidedly worth reading.

Goodman considers what might happen, both with adoption of OA journals (“gold OA”) and with possible mass cancellations of commercial journals—either because of “green OA” (self-archiving) or simply because even the wealthiest libraries can no longer accommodate the pricing policies of the biggest STM publishers. He anticipates either the NIH decision (in its earlier form) or the UK proposal and asserts that the U.S. and UK would “inevitably” adopt whatever was adopted in the other country—and that required OA would become universal in the top publishing nations if it was first adopted by France or Germany.

Some of the economic analysis is particularly interesting, such as Goodman’s explanation for the extreme rise in commercial journal prices:

Publishers’ prices have almost always increased faster than library budgets. This is due to positive feedback: publishers’ costs increase each year; they know that a comparable price increase will cause a certain number of subscribers to cancel, and therefore they increase the price to cover both. The obvious result is accelerating cancellations in all following years.

As any engineer can tell you, positive feedback is inherently unstable: It leads to breakdown, one way or another. This analysis offers yet another reason why the STM journal system is broken—despite the rosy claims of its largest commercial adherents.

There’s a lot to think about in this 11-page paper (including two pages of charts). In every scenario, Goodman believes OA will eventually become nearly ubiquitous—with a 90% rate somewhere between 2008 and 2015 (or later), depending on the scenario.

Davis, Phil, Terry Ehling, Oliver Habicht, Sarah How, John M. Saylor, and Kizer Walker, Report of the CUL task force on open access publishing, Cornell University Library, August 9, 2004. 27 pp.

The last two pages of this report have been used repeatedly as evidence that OA doesn’t make economic sense for libraries. That’s a shame. Those two pages offer an estimate of Cornell costs in a 100% OA journal model (assuming that all “author-pays” costs come out of the library’s budget, with no subventions from research funding agencies), concluding that breakeven is at $1,100 per article. That is, if the average cost per article turns out to be less than that, Cornell would save money in an all-OA environment; if it’s more, Cornell would spend more. At $1,500 per article, Cornell would spend about $1.5 million more than in the current model.

But that’s just the appendix. Change the set of assumptions and the numbers change. The report is worth reading on its own merit—and the report is most certainly not an attack on OA journals. From the executive summary: “Open Access publishing should not be regarded as an ultimate solution to the science serials crisis, but it can no doubt offer a pragmatic solution in specific cases. We should be discussing whether OA publishing is better than the current subscription model, and if so, for whom.”

Maybe research libraries should support OA publishing even if it does cost more: “There may be overriding ethical arguments for removing barriers to access.” It won’t much matter what the library believes if the scholars don’t support that belief: “Where Open Access does not respond to felt needs on the part of scholars and their disciplines, it is unlikely to gain support of authors.”

The report recommends that Cornell University Libraries “Foster and support viable Open Access publishing initiatives that respond to or resonate with real needs of specific scholarly communities”; consider OA strategies and projects based on whether the approach seems likely to be cost effective, meets the needs of user communities, and minimizes detrimental effects; continue environmental scanning regarding OA issues; and establish a standing committee to monitor developments.

Read the report carefully; it’s well-written and full of interesting nuggets. Although one OA evangelist seems to deny the existence or possibility of overlay journals (where a journal consists of a table of contents referencing archived papers), this report notes at least three such journals. The report is inclined to take publishers at their word regarding article costs (I’m tempted to call these prices, a quite different animal), but notes the wide range of “costs.” One comment on BioMedCentral is a bit snarky, but a little snarkiness improves a task force report. (Page 10, third paragraph, fourth line: You really should read the report!)

There isn’t one scholarly publishing community or academic community; there are many. That’s not news, but it sometimes seems to escape observers and participants. This task force understands that basic fact. They get that the so-called “crisis of scholarly communication” is really an STM serials pricing crisis—which indirectly creates a crisis in the humanities because libraries don’t have enough money left to buy specialized scholarly monographs.

Good stuff, carefully done. The appendix is just that: An appendix representing one set of calculations. Pay more attention to the first 21 pages (before a four-page bibliography).



Cites & Insights: Crawford at Large, Volume 5, Number 4, Whole Issue 60, ISSN 1534-0937, a journal of libraries, policy, technology and media, is written and produced by Walt Crawford, a senior analyst at RLG.

Cites & Insights is sponsored by YBP Library Services,

Hosting provided by Boise State University Libraries.

Opinions herein may not represent those of RLG, YBP Library Services, or Boise State University Libraries.

Comments should be sent to Cites & Insights: Crawford at Large is copyright © 2005 by Walt Crawford: Some rights reserved.

All original material in this work is licensed under the Creative Commons Attribution-NonCommercial License. To view a copy of this license, visit or send a letter to Creative Commons, 559 Nathan Abbott Way, Stanford, California 94305, USA.