The bottom line on open access
The rapidly evolving debate over free online scholarship drives right to the heart of some of the most fundamental questions about research
It may be entitled Theoretical Economics, but for the University of Toronto's Martin Osborne, the challenge of launching an open-access online journal was much more like taking a crash course in practical economics.
The first issue of Theoretical Economics, published by the California-based Society for Economic Theory, will go live this spring at www.econtheory.org - the culmination of months of planning and slogging by Dr. Osborne and a network of his colleagues from Boston University, Penn State and Northwestern. At the heart of their project lies a task that's nothing if not appropriate, given the subject matter: devising a viable business model that allows the society's new journal to be completely accessible online yet also operationally self-sustaining.
On the revenue side of the ledger, their solution began with a decision to charge $75 submission fees for authors ($35 for academics in developing countries), and offering one free submission a year to those who take out society memberships. On the expenses side, Dr. Osborne says online open-access journals have to be parsimonious about administrative overheads, as well as their fees or stipends to editors and peer reviewers. Printing costs, of course, are minimal, as are editing, typesetting and hosting costs. Lastly, to get it all out into cyberspace, Dr. Osborne spent hundreds of hours adapting an open-source, journal-publishing software package rather than pay the licensing fee for a commercial model.
From previous experience, Dr. Osborne knew that contracting with a large commercial publisher like Blackwell would mean forking over about $35,000 a year - a substantial sum for a small scholarly society. Their lean DIY model is much more, well, economical. "I don't want to exaggerate the ease with which an existing journal could convert to open access," reflects Dr. Osborne. "But it is possible."
Worldwide, there are about 24,000 scholarly journals, but only three to seven percent of them are considered to be "open access" - OA for short - meaning that they make their research papers available for free on the Internet. But the rapidly evolving debate over open-access scholarship extends well beyond academic journals like the one at www.econtheory.org, and drives right to the heart of some of the most fundamental questions about research: Do publicly funded universities and granting bodies have a democratic - indeed a moral - obligation to ensure that academic scholarship is available on the Internet? What kinds of public and institutional policies are needed to make such wide-ranging dissemination both possible and useful? And what are the implications for publishers, research libraries, copyright, and for scholarship itself?
Few self-respecting researchers argue with the idea per se. "It's easy to get people to sign off on a principle," says Stevan Harnad, who holds the Canada Research Chair in Cognitive Science at Université du Québec à Montréal. "It becomes interesting and substantive when you take a practical policy."
Since the mid-1990s, Dr. Harnad has been at the centre of an international campaign to promote open access. But it's only in the last three or four years - since the George Soros Open Society Institute orchestrated a 2002 summit of OA activists in Budapest - that granting bodies and universities have begun to look hard at how to translate open access from a feel-good cyber principle into something entrenched in the way academics do business - either by encouraging them to patronize open-access journals or urging them to routinely upload all their published research papers to a growing network of institutional repositories.
"The right to know is at the forefront [of OA]," says John Willinsky, a language and literacy education professor at the University of British Columbia who heads the Public Knowledge Project, a research initiative that asks whether and how online technologies can improve the quality of academic research. "The critical point we're at now is mandated access. We're seeing a momentum build."
The epistemological benefits are difficult to dispute. Dr. Harnad refers to studies showing that citations can more than double for articles that are freely available on the web. Accessible online papers benefit academics in poor countries where universities have few resources. And research libraries see institutional electronic repositories as one way of ensuring the preservation of digitized online material that is highly vulnerable to the problem of disappearing URLs.
At the same time, open access threatens the highly concentrated, commercial journal industry and could undermine the economic foundation of small, not-for-profit publications that depend on subscription revenues and, in Canada, on operating subsidies calibrated to paid circulation figures. In fact in Britain, the Association of Learned and Professional Society Publishers predicted that open access would lead to the destruction of journals, while a spokesperson for publisher Reed Elsevier told the Financial Times last summer that such changes would lower editorial quality and raise subscription prices without effectively altering access. Late last fall, the U.K. publishing lobby upped the ante by releasing a survey of 5,500 senior academics. The survey found "strong support" for the current publishing model.
In the past few years, large research councils in the U.S. and U.K. have grappled with the mechanics of applying the OA principle to publicly funded research. In Canada, in late 2004, the board of the Social Sciences and Humanities Research Council approved OA in principle; council staff are preparing recommendations based on public consultations. Neither the Canadian Institutes of Health Research nor the Natural Sciences and Engineering Research Council has followed suit, although the International Development Research Centre in late December announced plans to create an open archive for IDRC-sponsored research.
But open-access advocates contend that universities must now step up to the plate and adopt policies that compel faculty to "self-archive." That would mean they upload all their published papers to completely accessible, electronic, institutional repositories, which are linked with meta-search engines and indexing systems. Academics "won't do it until you make them do it," says Dr. Harnad of UQAM. "This is a natural extension of publish or perish."
When OA activists like UBC's Dr. Willinsky think about these issues, they point to historic analogies: the invention of the printing press and the World Wide Web, of course, but - equally important - the advent of scholarly journals like Philosophical Transactions. Founded in 1665 by the Royal Society of London, that journal signaled the end of the secretive tradition of scholars communicating with one another via private correspondence. Over centuries of academic tradition, however, scholars routinely relinquished the copyright in their papers to journal publishers as the price for being published in a predominantly non-commercial environment - a practice that takes on new implications in the Internet age.
The wider context for the current OA debate should be familiar terrain. In the 1970s, journal subscription rates began to rise steeply. In some disciplines, commercial publishers scooped up large numbers of journals and were able to boost their prices because they calculated that university libraries would continue to pay. But funding and space pressures in even very large institutions led to a triage in subscriptions. At the same time, academics began posting papers on personal websites and demanding electronic interlibrary loans of scholarly articles. University classrooms filled up with students who are completely at home in the digital world and often prefer receiving learning materials in electronic form. Meanwhile scholarly journals migrated online, offering pay-per-view fees. And in some disciplines, especially medicine, consumer use of online scientific information exploded, bringing scholarly writing to broad lay audiences.
But scholarly writing has migrated unevenly and unsystematically onto the web. What's more, the seeming ease of digital access, coupled with the proliferation of online material, merely serves to underscore what Tim Mark, executive director of the Canadian Association of Research Libraries, describes as the "absurd" situation whereby academics working for publicly funded institutions give up their intellectual property rights to commercial journal publishers, who turn around and sell the fruits of their labour right back to those institutions in the form of costly journal subscriptions.
This isn't just a Canadian problem, of course. In the U.S. - where many jurisdictions have passed tough laws ensuring public access to government documents - the National Institutes of Health in 1999 moved to create an electronic archive of NIH-funded research. But it soon became clear that the E-biomed database represented a significant source of competition for large journal publishers, who lobbied Washington to embargo articles for a year after publication before they could be posted to the archive, now known as PubMed.
In the U.K., the government, journal publishers and the eight leading research councils have been batting around various proposals requiring researchers to self-archive research papers. Parliament ducked the issue and handed it off to Research Councils UK, which has yet to arrive at a formal position. With SSHRC on the verge of releasing a policy recommendation later this spring, Dr. Harnad feels "it's not out of the question that Canada is the country that does it first."
From where SSHRC senior policy adviser David Moorman sits, the open access issue looks a lot less straightforward than it may seem from a distance. SSHRC is one of the very few research funding councils that subsidizes journals as well as researchers. The council earmarks $2.1 million a year for a program that gives grants of about $10,000 to $30,000 to Canadian social sciences journals, many of them with small circulations and few readers beyond our borders. "We have to make sure we don't undermine the financial stability of those journals," he says. "If we force on them an OA business model, many think they'll fold."
SSHRC calibrates its journal grants to subscription numbers, and Mr. Moorman says it's by no means obvious how the council's grant officers would gauge "impact" if those journals were required to move to an OA business model, such as Theoretical Economics' submission-fee revenue stream. Website hits and download data aren't considered to be reliable indicators, and Mr. Moorman's view is that SSHRC has to come up with an alternative way of dispensing its journal grants before imposing conditions about OA. Moreover, he's uneasy about any kind of policy framework that potentially restricts SSHRC-supported Canadian scholars to publishing in the relatively small universe of open-access journals, both here and abroad. Dr. Harnad says nothing of the sort has been proposed for SSHRC.
The funding policy dilemmas associated with OA journals have already begun to emerge. One heavily cited, open-access, peer-reviewed journal with more than 5,000 regular readers was rejected for a SSHRC publication grant because it didn't meet the rule of a minimum 200 paid subscribers. (The journal - Athabasca University's International Review of Research in Open and Distance Learning - tried to get around the rule by asking online readers to send a one-time $10 subscription fee, but even though 350 readers sent $10 within three days, according to the editor, the application was still turned down by SSHRC.)
"The concerns of journal editors are completely legitimate," says Mr. Mark of CARL. "If the SSHRC consultation can devise a successful business plan to move [Canadian journals] to an electronic business model, that would be a real accomplishment."
But designing a workable journal model is only half the equation. "SSHRC's problem," asserts Dr. Harnad, "is that it's got its lines crossed." The potential of the OA movement, he argues, doesn't begin with policy conditions aimed at altering the operating conditions for a small subset of journal publishers. Rather, it needs a much broader-based effort to make institutional self-archiving a routine and unquestioned part of the work of scholarship - as basic as including bibliographies and reference lists at the end of any paper.
OA advocates say the pieces for such a cultural change are beginning to fall into place. There are now numerous open-source software and "harvesting" systems that allow institutions to create searchable, indexed and networked electronic repositories, among them Dr. Harnad's 13 E-print archives and MIT's DSpace, which has been adapted by nine Canadian universities.
But, as Dr. Harnad observes, the availability of user-friendly archiving software is necessary but not sufficient. That's why, in 2005, OA activists approved the Berlin 3 institutional policy commitment. It calls on universities and research institutions to establish policies requiring academics to self-archive, as well as encouraging them to publish in open-access journals. OA proponents note that the former is a far higher priority if the ultimate goal is disseminating scholarship. So far 17 universities and research institutions - including the University of Zurich, Portugal's University of Minho, and the University of Southampton, where Dr. Harnad taught before joining UQAM - have signed the 2005 Berlin commitment. No Canadian universities are signatories.
How do academics feel about self-archiving? "Authors haven't picked it up," says Dr. Willinsky at UBC. "It has a lot to do with the fact that the focus of [academics'] work is getting published, not getting circulated." Indeed, a U.K. survey of scholars showed that about half of the respondents had self-archived at one point, mainly on personal websites, but many didn't do it routinely. Yet 95 percent said they'd be prepared to self-archive if their university required it as a condition of tenure or employment.
What's become increasingly apparent is that copyright issues aren't a roadblock for the OA movement. "Open access is a friend of copyright," says Dr. Willinsky, who notes that copyright exists to protect and enhance the reputation of authors. He says a vast majority of journal publishers now allow their contributors to self-archive on personal websites or institutional repositories. Some journals don't want their authors to post electronic replicas of published articles. So instead, many academics simply create a pre-publication version and post that instead.
Mr. Moorman and other intellectual property rights experts say the real concern is that many scholars simply don't understand copyright; the situation is exacerbated by the fact that scholarly publishing was built on the longstanding practice of academics forfeiting all their rights to journals. But with the advent of "creative commons" licences (which allow authors to reserve certain electronic rights), younger and more web-savvy academics are moving to ensure that their work can legally circulate on the web.
The American lesson suggests that when a large funding body like the NIH creates a fully accessible online database, the academic publishing giants see red because it functions like a one-stop-shopping mega-journal. Yet it appears that the journal industry and such institutional repositories as PubMed have gradually worked out what Dr. Willinsky calls "a peaceful co-existence."
While journal publishers, from giants like Elsevier to upstarts like Econtheory.org, will continue to work out a sustainable online business model, the OA policy ball has now landed squarely in the university sector's court. "The handwriting is on the wall," says Dr. Harnad. "But it's not going to happen spontaneously."
For further reading:
Study comparing citations of open-access vs. non-OA articles in the same journals http://www.dlib.org/dlib/june04/harnad/06harnad.html
The Scholarly Publishing and Academic Resources Coalition: www.arl.org/sparc
Open access of a sort
Within the Canadian environment, a limited open-access repository solution has emerged in Quebec with Erudit.org. Founded in 1998 as a co-venture of Université de Montréal, Université Laval and UQAM, Erudit is an archive of articles from about 40 journals published in Quebec and elsewhere. It allows free downloading of all articles except those published within the last two years - a modus operandi that puts it in league with some leading U.S. journal publishers but falls short of the full open access terms adopted by a handful of international research institutions, including the University of Southampton and CERN in Zurich. Institutions can acquire licences to access the more recent papers, meaning that Erudit functions like a hybrid between an OA repository and a toll-based database.