From “Helping scholars and helping libraries,” SPARC Open Access Newsletter.
Scholars and librarians are close allies in the campaign for open access, but they pursue OA for different reasons. For scholars, the primary benefit of OA is wider and easier access for readers and larger audience and impact for authors. For librarians, the primary benefit of OA is saving money in their serials budgets.
(I've argued in the past that OA will help libraries solve both the “pricing crisis” and the “permission crisis.” I haven't changed my mind about that, but here I'll focus on the library interest in solving the pricing crisis.)
We like to say that OA is a goal to which there are many means. But scholars and libraries in effect treat it as a means to their separate but related, professionally-specific goals. It's not surprising that some OA initiatives help scholars more than they help libraries, or that some OA initiatives help libraries indirectly but don't help them save money in their serials budgets. Here are four quick examples.
(1) Self-archiving, at least at its current low levels: Every new article in an OA repository helps its author and all of its readers. But it doesn't justify librarians in canceling journal subscriptions and it doesn't convert subscription journals to OA journals. As self-archiving spreads, there may be a tipping point after which it helps libraries save money. Or there may not, if the experience in physics transfers to other disciplines. But what matters here is that self-archiving helps scholars even before we get to that tipping point.
Similarly, libraries that host institutional repositories have an important new institutional responsibility and they can benefit from taking on that role. But self-archiving helps scholars even at universities that don't yet have institutional repositories.
(2) Hybrid journals offering OA at the author's choice (the “Walker-Prosser” model): When only a fraction of authors take advantage of the OA option, then libraries cannot justify canceling subscriptions. But scholars benefit when even a fraction of a journal's content goes OA.
(3) Non-OA journals offering OA to sufficiently old back issues: OA to back issues may help libraries save the cost of buying access to the back run, but it doesn't help them save the larger expense of subscribing. If the embargo is sufficiently long, then libraries will not be able to justify cancellations. For journals, of course, this is the point. Finding an embargo period that is short enough to serve research needs in the field without subverting subscriptions (i.e., without helping scholars enough and without helping libraries in the primary way) is an art that many journals are trying to master.
(4) The NIH public-access policy: I've often argued that there are good reasons to think that it won't undercut journal subscriptions. That's the flip side of arguing that libraries won't cancel subscriptions on account of this policy, even if they continue to cancel subscriptions for other reasons, such as rising prices. The best reason to think that the NIH policy will not cause cancellations, and would not have done so even in its earlier and stronger form with only a six-month embargo, is that libraries themselves said so. In its comment on the NIH policy, the Association of College and Research Libraries (ACRL) wrote:
We wish to emphasize, above all, that academic libraries will not cancel journal subscriptions as a result of this plan and that it will therefore not produce economic harm to publishers. Since biomedical journals publish research that derives from many sources other than NIH funding, the articles made available in PubMed Central will not substitute for the content of individual journals. Even if libraries wished to consider the availability of NIH-funded articles when making journal cancellation decisions, they would have no reasonable way of determining what articles in specific journals would become openly accessible after the embargo period. The six-month embargo also provides substantial protection of publishers’ interests. Because most biomedical research is time sensitive, libraries will make every effort to maintain the subscriptions they already have as a way of providing needed access to the most current research.
Eight reasons to think the NIH public-access policy won't undercut subscriptions
The ACRL comment on the NIH public-access policy, November 16, 2004
We know what kinds of OA initiative will help scholars—namely, every kind. But what kinds of OA initiative will help libraries save money in their serials budgets?
I want to approach this question from the side. It looks as though helping libraries has to hurt publishers. It looks as though helping libraries save money in their serials budgets means canceling subscriptions. That is not quite true. It's true that canceling subscriptions is one way to help libraries save money, but it's not the only way. Libraries also save money when subscription-based journals convert to OA. Conversions are voluntary and can come about through persuasion, experimentation, or changing market conditions.
I want us to achieve the kinds of OA that help scholars and the kinds of OA that help libraries. I want to help libraries in part because I'm in a symbiotic relationship with them. Healthy libraries are necessary to support healthy research and education. I want to help libraries because they are committed allies in the campaign for OA. More importantly, I want to help libraries because the best source of funds to pay for the long-term sustenance of OA is the savings from library serials budgets. If the rise of OA literature lets us spend less on priced literature, then the best way to spend the savings is on the OA alternative that made the savings possible.
Finally, I want to help libraries because we can't take the last steps toward helping scholars without helping libraries as well. For example, low-volume self-archiving may help scholars without helping libraries, but high-volume self-archiving will help both, at least in fields where the experience in physics does not transfer. I want high-volume self-archiving. I want OA to 100% of research literature, through some combination (it doesn't matter what combination) of OA archiving and OA journals.
The inevitable question is whether I, and all others who want to help libraries, want to harm publishers. The answer is no. That is not the goal. The goal is OA to 100% of the research literature. Achieving that goal is compatible with subscription-based access to some considerable percentage of the same literature. (These percentages can add up to more than 100% because some literature can be both free and priced, such as an article in a subscription journal deposited by its author in an OA repository.)
Progress toward the goal of OA to 100% of research literature will eventually help libraries reduce their serials expenses. It may create the quality and quantity of OA literature that justifies them in canceling some subscription journals. Or it may persuade some subscription journals to convert, e.g., by answering fears about conversion or by changing market conditions so that conversion becomes a survival strategy. Or both.
I don't know how much this progress will help libraries. That depends on how much priced access is compatible with 100% OA. Last month I looked at some of the variables that affect this kind of long-term co-existence and some of the reasons why predictions are difficult.
I'm sorry if all of this is obvious. During the long campaign for the NIH policy, I argued that this particular policy would not harm publishers. I made the argument so often that I began to wonder whether it created a false impression about my larger position or at least drew attention away from two equally important parts of it. First, the NIH policy won't do much to help libraries, and second, other initiatives that I do support will certainly help libraries and could threaten subscriptions.
I will never support an initiative whose direct purpose is to undermine publishers. I've argued before that OA does not require publisher setbacks, and that publisher setbacks do not necessarily advance OA. Hence, to pursue publisher setbacks is to mistake the goal. It's harmful and wrong. I haven't changed my mind about that.
But I definitely support initiatives to enlarge the body of OA literature, rapidly and systematically, even if a foreseeable side-effect is that libraries cancel subscriptions. I want to say this in public. One reason is to refocus on the needs of libraries, overlooked in the NIH campaign. Even if libraries benefit indirectly from every OA initiative that benefits scholars, they need more direct forms of relief as well. Another reason is simply to admit that not all OA initiatives will be as innocuous for publishers as the NIH policy.
For example, I support voluntary self-archiving of journal articles by all authors. I support policies at every funding agency, public and private, to mandate OA to funded research. I support policies at all universities to mandate OA to royalty-free research, like journal articles, produced by faculty. Each of these policies will lead to high-volume OA archiving. I support OA archiving for full-text articles, not just abstracts. I support full-text archiving for postprints, not just for preprints. I support postprint archiving immediately upon publication, not just after some embargo period. I support postprint archiving in OAI-compliant repositories, not just on personal home pages. I support these policies for all disciplines, all countries, and all languages. High-volume, full-text, immediate and interoperable OA postprint archiving will help libraries by leading to some combination (I don't know what combination) of journal conversions to OA and library cancellations of selected journals that do not convert. To supplement high-volume OA archiving, I support more peer-reviewed OA journals in every field, whether they are new launches or conversions from TA journals.
As we approach 100% OA, through archives and journals, I'm sure that the responses of subscription journals will differ from discipline to discipline. So I don't know how much this progress will undercut subscriptions. If most disciplines are like physics, or if some TA journals can survive in a high-OA environment, or if TA journals that lose subscribers can survive by converting to OA, then the answer is not much. But I don't want to be evasive. Even if the answer is that progress toward 100% will be more rather than less harmful to existing publishers, I still believe that 100% OA is a goal worth pursuing. Certain services, like peer review and wide and easy distribution, are indispensable for science and society. But no particular journal or publisher is indispensable.