This post is part of our Open Access Week blog series to highlight great work in Open Access communities around the world. It is written by Miguel Said from Open Knowledge Brazil and is a translated version of the original that can be found the Brazilian Open Science Working Group's blog.
Nature Publishing Group reported recently that in October, its Nature Communications journal will become open access only: all articles published after this date will be available for reading and re-using, free of charge (by default they will be published under a Creative Commons Attribution license, allowing virtually every type of use). Nature Communications was a hybrid journal, publishing articles with the conventional, proprietary model, or as open access if the author paid a fee; but now it will be exclusively open access. The publishing group that owns Science recently also revealed an open access only journal, Science Advances – but with a default CC-NC license, which prevents commercial usages.
So we made it: the greatest bastions of traditional scientific publishing are clearly signaling support for open access. Can we pop the champagne already?
This announcement obviously has positive aspects: for example, lives can be saved in poor countries where doctors may have access to the most up-to-date scientific information – information that was previously behind a paywall, unaffordable for most of the Global South. Papers published under open access also tend to achieve more visibility, and that can benefit the research in countries like Brazil, where I live.
The overall picture, however, is more complex than it seems at first sight. In both cases, Nature and Science adopt a specific model of open access: the so-called "gold model", where publication in journals is usually subject to a fee paid by authors of approved manuscripts (the article processing charge, or APC). In this model, access to articles is thus open to readers and users, but access to the publication space is closed, in a sense, being only available to the authors who can afford the fee. In the case of Nature Communications, the APC is $5000, certainly among the highest in any journal (in 2010, the largest recorded APC was US $ 3900 – according to the abstract of this article… which I cannot read, as it is behind a paywall).
This amounts to two months of the net salary of a professor in state universities in Brazil (those in private universities would have to work even longer, as their pay is generally lower). Who is up for spending 15%+ of their annual income to publish a single article? Nature reported that it will waive the fee for researchers from a list of countries (which does not include Brazil, China, India, Pakistan and Libya, among others), and for researchers from elsewhere on a "case by case" basis – but they did not provide any further objective information about this policy. (I suspect it is better not to count on the generosity of a publisher that charges us $32 to read a single article, or $18 for a single piece of correspondence [!] from its journals.)
On the other hand, the global trend seems to be that the institutions with which researchers are affiliated (the universities where they work, or the scientific foundations that fund their research) bear part of these charges, partly because of the value these institutions attach to publishing in high-impact journals. In Brazil, for example, FAPESP (one of the largest research foundations in Latin America) provides a specific line of funding to cover these fees, and also considers them as eligible expenses for project grants and scholarships. As it happens, however, the funds available for this kind of support are limited, and in general they are not awarded automatically; in the example of FAPESP, researchers compete heavily for funding, and one of the main evaluation criteria is – as in so many situations in academic bureaucracy today – the researcher's past publication record:
Analysis criteria [...] a) Applicant's Academic Record a.1) Quality and regularity of scientific and / or technological production. Important elements for this analysis are: list of publications in journals with selective editorial policy; books or book chapters [...]
Because of this reason, the payment of APCs by institutions has a good chance of feeding the so called "cumulative advantage" feedback loop in which researchers that are already publishing in major journals get more money and more chances to publish, while the underfunded remain that way.
The advancement of open access via the gold model also involves another risk: the proliferation of predatory publishers. They are the ones that make open access publishing (with payment by authors or institutions) a business where profit is maximized through the drastic reduction of quality standards in peer review – or even the virtual elimination of any review: if you pay, you are published. The risk is that on the one hand, predatory publishing can thrive because it satisfies the productivist demands imposed on researchers (whose careers are continually judged under the light of the publish or perish motto); and on the other hand, that with the gold model the act of publishing is turned into a commodity (to be sold to researchers), marketable under high profit rates - even without the intellectual property-based monopoly that was key to the economic power mustered by traditional scientific publishing houses. In this case, the use of a logic that treats scientific articles strictly as commodities results in pollution and degradation of humankind's body of scientific knowledge, as predatory publishers are fundamentally interested in maximizing profits: the quality of articles is irrelevant, or only a secondary factor.
Naturally, I do not mean to imply that Nature has become a predatory publisher; but one should not ignore that there is a risk of a slow corruption of the review process (in order to make publishing more profitable), particularly among those publishing houses that are "serious" but do not have as much market power as Nature. And, as we mentioned, on top of that is the risk of proliferation of bogus journals, in which peer review is a mere facade. In the latter case, unfortunately this is not a hypothetical risk: the shady "business model" of predatory publishing has already been put in place in hundreds of journals.
Are there no alternatives to this commodified, market-oriented logic currently in play in scientific publishing? Will this logic (and its serious disadvantages) be always dominant, regardless if the journal is "proprietary" or open access? Well, not necessarily: even within the gold model, there are promising initiatives that do not adhere strictly to this logic – that is the case of the Public Library of Science (PLOS), an open access publishing house that charges for publication, but works as a nonprofit organization; because of that, it has no reason to eliminate quality criteria in the selection of articles in order to obtain more profits from APCs. Perhaps this helps explain the fact that PLOS has a broader and more transparent fee waiver policy for poor researchers (or poor countries) than the one offered by Nature. And finally, it is worth noting that the gold model is not the only open access model: the main alternative is the "green model", based on institutional repositories. This model involves a number of challenges regarding coordination and funding, but it also tends not to follow a strictly market-oriented logic, and to be more responsive to the interests of the academic community. The green model is hardly a substitute for the gold one (even because it is not designed to cover the costs of peer review), but it is important that we join efforts to strengthen it and avoid a situation where the gold model becomes the only way for scientists and scholars in general to release their work under open access.
(My comments here are directly related to my PhD thesis on commons and commodification, where these issues are explored in a bit more detail – especially in the Introduction and in Chapter 4, pp. 17-20 and 272-88; unfortunately, it's only available in Portuguese as of now. This post was born out of discussions in the Brazilian Open Science Working Group's mailing list; thanks to Ewout ter Haar for his help with the text.)