Are open-access fees disenfranchising?

I had an interesting discussion over coffee at the recent SOAP Symposium about the question of whether the article processing fee revenue model for open-access journals disenfranchises authors with fewer financial resources. It prompted me to write up a fuller explanation of why this worry is misplaced.

Opportunity for full participation in research by as wide a range of scholars as possible is, of course, central to our meritocratic notion of the scholarly endeavor. Perhaps the biggest impediment to such full participation — to getting to the point where one has a scholarly result to present to the world — is gaining access to the facilities for carrying out research in the first place, including access to the published literature. It makes little sense to worry about disenfranchisement from publishing research results if the alternative is disenfranchisement from the reading that would allow generating the results in the first place. For that reason, open access to the scholarly literature is inherently an enfranchising program.

It also bears mentioning that it is not only open-access journals that charge author-side fees, the kinds of fees that critics complain are disenfranchising. Many subscription journals charge quite substantial fees as well. For NIH-funded research, the average is $1250 per article, which is plenty big enough to give your average developing-country scientist pause. One would be hard-pressed to impugn open-access journals on these grounds without roping in many subscription journals as well.

That being said, of course we want everyone to have the opportunity to publish in the scholarly literature, even those with lesser means. And there is a simple mechanism to allow for this with open-access journals that charge article processing fees. Journals can, should, and commonly do waive fees for necessitous authors. The details of these waiver policies differ. (See here for the PLoS policy or here for BioMed Central.) But the effect is the same: authors unable to afford the fees can still publish in these journals. More importantly, they can read the articles published in the journals too.

Some worry that authors requiring fee waivers may be discriminated against in the editorial process. Editorial processes must, of course, be kept separate from the financial processes. Different groups separated by a Chinese wall can handle the two issues. Indeed, the question of whether a waiver will be requested needn't even be raised until an editorial decision on a paper is finalized, eliminating any possibility of a conflict of interest. PLoS has an especially simple method for handling waivers. After a paper is accepted for publication, authors can request a waiver of the fee, which is always granted.

Of course, the waiver idea can't possibly be controversial. It is the same approach that subscription journal publishers use to address the reader-side disenfranchisement argument. They point out that the World Health Organization's Hinari program provides subsidized access to journals for scholars in a specified set of countries that have been deemed sufficiently impoverished.  A similar eligibility criterion could be used for processing fee waivers. But an approach based on targeting individuals rather than countries has much to recommend it. It can be much better focused on the real problem. For instance, it can address authors in needy cohorts who happen to live in a country not on the approved list. There are unemployed scholars in first-world countries or faculty at small schools in developing countries, for example, for whom Hinari is no help, whereas a fee waiver allows them to fully participate in the open-access publishing milieu on both the reading and writing side.

[UPDATE 1/21/11: The recent news that publishers have withdrawn Bangladesh's access through the HINARI program (because Bangladesh is "start[ing] to secure active sales") makes regrettably clear the problem with this approach. Just because some researchers in Bangladesh may now fall within the scope of an institutional subscription, all are deprived access.]

The issue of fee waivers is important, and we should actively promote their availability. By way of example, many COPE-compliant open-access funds — including those at Harvard, Cornell, Dartmouth, MIT, and Columbia — will only cover fees for journals that have a waiver policy. Hopefully, this will provide some impetus for OA journals to institute reasonable waiver policies.

Ironically, Nature Publishing Group is entering the OA arena with Scientific Reports,  a PLoS ONE competitor. Phil Davis reports that they are apparently not allowing for fee waivers, and points out that this could lead to a problem of adverse selection, where PLoS ONE ends up handling all of the fee-waived articles to their competitive disadvantage. On the other hand, if this turns out to be true, Scientific Reports will not be eligible for support from the COPE-compliant open-access funds as discussed above. There thus may be ways to mitigate the adverse selection problem.

With open access, we can enfranchise both the readers and the writers of the scholarly literature. We can, and we should.