Cover Story

Any psychologist who has followed the science media over the past year has likely caught wind of the debate over "free access" or "open access," terms used to describe free, unrestricted public Internet access to scientific information. Fueled by a San Francisco-based group called the Public Library of Science (PLoS), the movement's idealistic aim is to keep taxpayers from what PLoS calls "paying twice" for scientific data: once when they fund the government agencies that sponsor research, and again when they pay online fees to access scientific journal articles. The problem could be stopped, PLoS advocates argue, by changing the financial nature of science publishing from a system based on subscription fees--which they deride as overly profit-based--to one based on fees paid up front by authors.

The plan has a few strong supporters and many critics, both among science publishers and scientists themselves. While advocates praise the plan for its democratizing agenda, critics say it fails to account for the realities of publishing.

"It's a good idea in principle," says APA Publisher Gary VandenBos, PhD, "but it is also incredibly naive."

The money for "free" publishing has to come from somewhere, he and others say; on top of that, many scientists feel unable to afford the hefty fees required by the plan, particularly if they use complex funding streams.

Spearheading the free-access plan are three heavy hitters: former National Institutes of Health director Harold E. Varmus, MD, PhD, now head of Memorial Sloan-Kettering Cancer Center; Patrick O. Brown, MD, PhD, an eminent Stanford University geneticist; and Michael B. Eisen, PhD, an evolutionary biologist at the University of California at Berkeley.

The group voices its mission in radical terms: "We want to do whatever we can to revolutionize the whole system, so that publishers don't have monopoly control over what they publish and the material is freely available in the public domain," says Brown of PLoS. That aim, he and other PLoS advocates contend, could be achieved by the proposed switch to a single up-front author fee, which they argue would remove the opportunity for science publishers to build in big profit margins through overpriced individual or institutional subscription fees. Free access, PLoS advocates hope, will open banks of knowledge to the public for medical and educational reasons, and force scientists to share data--an initiative that has been pushed by federal agencies for the past 15 years, but resisted by many scientists.

Indeed, Brown's inability to find good data links in his field of genetics prompted him to help launch PLoS in the first place: "I found that in principle, a lot of this information was available in digital format," he says, "but in practice, the restrictions on use and access prevented it from being integrated into databases. That got me thinking about the problem, and I saw there was a more important issue that had to do with the rights of the public and scientific community to get the full benefits of the published works."

However, VandenBos notes that the basic data of a research project is rarely included in published articles, so the initiative is mandating additional work by both scientists and publishers--without a clear plan of how the additional work will be paid for. In general, VandenBos adds, researchers do not want their data available to others to analyze until the scientists who put the effort into the data collection have done all of the analyses that they wish to do. In an effort to increase journal access, PLoS launched its first free online journal, PLoS Biology, in October, and it has others in the works. No existing journal, however, has proposed to move into PLoS.

Moreover, PLoS has backed a congressional bill, H.R. 2613, or the Public Access to Science Act, which seeks to ban copyright protection on all scientific works that are significantly funded by the government. The bill, introduced last June by Rep. Martin Sabo (D-Minn.), however, is considered unlikely to pass anytime soon for lack of bipartisan support--it had only three Democratic co-sponsors and no Republican co-sponsors. But with another year in this congressional session, it remains a bill to watch. For all of its good intentions, the PLoS proposal has serious flaws, critics say. It is the plan's radical proposal to make all such information free that's the problem, they maintain, not the organization's quest for greater accessibility. Indeed, many publishers have been considering and implementing open access--meaning available at low cost--for some time, they note. For example, APA currently makes the tables of contents and abstracts of all newly published journals freely available to any user--and, on a use-by-use basis via site licenses, access to APA electronic documents averages less than 10 cents.

"What they really mean is making it free to the user or the consumer of the information," says psychological researcher J. Bruce Overmier, PhD, of the University of Minnesota, an APA Board of Directors liaison to APA's Publications and Communications Board. "The public may have paid for the research, but they haven't paid for its publication and dissemination."

Adds APA's VandenBos: "When you say you're going to give it away for free, you have to think about all of the different ways you're going to have to pay for it that you're not paying for it now, everything from getting the government to build money into grants, to passing that money through states to the universities, to transferring the money to professional associations to handle the many tasks of publishing. The initiative would also force the development of a funding mechanism to cover the costs of publishing raw data, an expense for researchers and publishers that does not currently exist."

In essence, PLoS proposes a move that would replace a painstakingly developed, time-tested system for one lacking in history and safeguards, Overmier believes. "It's a case that's ripe for the principle of unintended consequences," he says. "The fallout is going to be incredible, and you won't know it until it's too late."

How the models differ

Under the traditional model of scientific publishing, publication costs are covered by individual subscription and library licensing fees. For example, APA members subscribe to individual journals or to PsycINFO, and academic libraries pay subscription and licensing fees for those same goods and services. In turn, funds are funneled back to the association to support publication costs as well as other member activities.

Under the PLoS model, journal revenues would instead come from authors or their backers, who would pay a lump sum to publish their work. PLoS Biology, for example, charges investigators $1,500 to publish an article, which in Brown's view provides a cleaner, less compromised money pool than a subscription-based system.

"We're saying we're just going to redirect the money and pay the cost of publication up front," says Brown. "You get everything you want and the quality control is just as good."

As simple as it sounds, the proposal is rife with problems, opponents maintain. Scientists don't always have the money to pay author fees, they say, and the typically proposed solution to that problem--writing fees into grant proposals--fails to cover all of the bases. In addition, in psychology, only 20 percent of research is federally funded.

On top of that, grants, in particular in psychology, often come from multiple sources, cobbled together from government, charitable and private funds, says APA's former deputy publisher Susan Knapp, who retired in December after 30 years at APA. "This is the thing that confuses me about the [Sabo] copyright bill, frankly," she says. "It says we shouldn't copyright any government research, but more often than not, there's more than one funding source involved."

Others are concerned that the fees would end up degrading the science. Kurt Salzinger, PhD, APA's former executive director for science, believes that adding money into grants for publishing fees could siphon funds away from other aspects of the research process. For one thing, it's hard to estimate how many papers will result from a given project, he says; for another, "If scientists had to include the cost of the papers in grants and put those amounts into their budgets, there would be less money left for actually doing the research," he notes.

And Susan McDaniel, PhD, who chairs APA's Publications and Communications Board, worries that author charges may have the unintended effect of garnering fewer submissions.

"The most likely outcome of adding such a fee is that the best people will submit their work to publications that do not charge," she says. "And junior people with small bank accounts might be dissuaded from the whole endeavor."

And fewer submissions means lower quality, says Overmier: "As long as there is an excess of articles over what they can publish, there is little risk to quality," he notes. "But if the number of submissions shrinks, then a publication needs to accept lower quality articles just to sustain the enterprise."

Psychology would face additional hurdles were the PLoS model widely adopted, VandenBos adds. Unlike some other fields that have used the author-fee model for years, psychology, which often is not funded by large government grants, lacks such a tradition.

"Psychologists would have to pay something like 80 percent of the publication costs out of their own pockets," he says--a big reason APA decided not to charge author fees in the first place.

In addition, the $1,500 fee charged by PLoS does not reflect typical costs, particularly of high-quality journal articles, many say. A typical APA journal article, for instance, costs between $7,000 and $8,000 to produce. PLoS itself was launched with a $9 million startup grant that's projected to last four years; what will happen to its publishing enterprise, critics ask, when the money runs out?

Similar goals, different means

These concerns notwithstanding, many science publishers, including APA, are sympathetic to PLoS's goals, but believe in moving toward them in a more evolutionary fashion. Many are moving toward more open access anyway, to better serve their customers and to take advantage of expanding digital and electronic capabilities.

Literature searches cost libraries very little, VandenBos adds. "APA has worked with several universities to track usage of the APA PsycINFO database at their campuses," he says. "It is heavily used, and when the licensing fee is divided by the number of searches done, the resulting number is always very, very low--less than 10 cents per search under a typical licensing agreement." APA has also created options for the public, who tend not to want annual access to APA products. They can, for example, gain 24-hour access to PsycINFO for $11.95. In 24 hours, some such users do 20 to 25 searches, so the cost per search is low. The association is also developing other low-cost databases with the public in mind, including PsycEXTRA, which will condense popular psychological literature into a user-friendly format.

The American Physiological Society is another example of an organization working to achieve more open access. The society was one of the first to make access to its online journals free and to publish articles online before they came out in print as well as to give authors a choice to pay an initial fee to make their articles open access, according to the society's director for publications, Margaret Reich.

What stops the organization from embracing the PLoS effort is problems related to author fees, Reich notes. "It is precisely because of our experience with author fees and our authors' dislike for them that we are not sure whether the open access model will work," Reich noted in the August 2003 issue of the society's publication The Physiologist (Vol. 46, No. 4). Instead, the society uses a subscription-based model that distributes costs among authors, readers and institutions, she says.

APA also had experimented for several years with a free, exclusively online journal called Prevention and Treatment, says VandenBos, but there was extremely high author resistance to publishing in an electronic-only journal. (It stopped publication at the end of 2003.) Other free online journals, such as the British Journal of High Energy Physics and the British Medical Journal, have reported problems attracting high-quality submissions and problems related to funding, an article in the Oct. 23, 2003, issue of Science (Vol. 302) notes.

Part of the reason PLoS advocates argue so passionately on the topic, they say, is their view that some commercial publishers reap huge profits from their activities--40 percent or more for some of the biggest publishers, according to the Science article. Indeed, the issue has been a bone of contention with academic and public libraries for years, which they say can scarcely afford the large fees some publishers charge.

PLoS critics agree. "The real culprit in all of this--and the thing that's really precipitating the crisis--is a few of the for-profit publishers," Overmier maintains. "A few of them have profits that are pretty exorbitant. But," he emphasizes, "that contrasts sharply with non- and not-for-profit academic and professional societies like APA," that funnel that money back into publishing and member activities. "There's a huge disconnect between the two."

A matter of quality

Besides financial concerns, critics also fear quality would suffer if a single entity such as the government subsumed the publication process. Not only would publishers' longtime expertise be lost, they fear, but government funding could evaporate at the whim of a new Congress or administration.

Knapp provides a cautionary tale related to the point. During the 1980s, the National Institute of Mental Health (NIMH), using a similar rationale to that of PLoS, asserted it could do a better job than commercial or association publishers of disseminating a compendium of mental health research. The agency developed a database, Mental Health Abstracts, which attempted to provide this service for free.

"It was around for five or six years and the funding got yanked," Knapp says. In an ironic twist, NIMH eventually transferred ownership of the publication to another publisher, which in turn sold the entity to APA in 2000.

Likewise, the PLoS plan could dampen the creativity provided by a free marketplace, critics worry. "APA risked millions of dollars a year before they ever made a penny on PsycINFO," says Overmier. It is likely that the incentive to develop such products would dry up without backup funds, he notes.

Indeed, science publishers helped make open access more possible in the first place, contends Robert Wells, PhD, president of the Federation of American Societies for Experimental Biology, or FASEB, which represents 22 scientific societies that publish a total of 60 journals. "Nonprofit journals published by scientific societies have pioneered some of the most impressive changes in scientific communication out there," he says. "The current system is a very dynamic one, and the journals I am familiar with have been on the cutting edge of electronic publishing and communications. The diversity of the current system," he adds, "is a major strength that leads to experimentation and innovation."

Others ask whether having all scientific information freely available to the public would really serve the public's best interest.

"To me, the public would be much better served by having better access to synthesizing reviews," VandenBos says. "Whether the public can benefit from access to each individual, empirically based, highly detailed, complex article without having any context for it is an open question. I'm not even sure that professionals who aren't in that specialty area would benefit from such a system," he says.

Let the market decide

In the end, the market itself may determine which system wins out, PLoS advocates and critics agree.

"If there is increasing pressure from paper authors on one hand and subscribers on the other--authors to send their works preferentially to open-access journals, and subscribers to put financial resources behind open-access mechanisms--well, that's just the free market speaking, and it wouldn't be fair to portray it as some kind of coercive mechanism," says Brown.

Likewise, "If there's a demand for a product out there and people are willing to pay for it, publishers will meet that demand," Overmier says. "If nobody's going to pay for it, nobody's going to create it."

Even if PLoS fails in its mission, open and free access is an issue that's not going away, those involved acknowledge. Scientists of all stripes, for example, are clamoring for better access to data so they can avoid duplicating efforts and benefit the public (see Data sharing).

VandenBos predicts it will be an issue that keeps all those involved on their toes. But he doesn't think anything will happen overnight, and there's still time to consider consequences and plan ways of coping, he believes.

"There are many, many shareholders in this, and there will continue to be growth and evolution, fits and starts, and movement in a variety of directions," VandenBos says. "At some point there will probably be more and better access," he adds, "but what shape it will take is still unclear."

Tori DeAngelis is a writer in Syracuse, N.Y.