La version française de cet article est ici.


Since a few years, academic publisers, especially the commercial ones (Elsevier, Wiley, NPG), have been prone to stress the value they bring to the scientific process, and more precisely to scholarly communication. Therefore they define the prices they ask for as « value based » and not « cost based ».

Alicia Wise, from Elsevier, exactly states this point :

And yes, we need to talk, because knowing how the value is assessed is paramount. Alexander Brown, from Springer, described in a 2012 Guardian blog post the value academic publishers add in the process of the dissemination of scientific knowledge. Two key points are stressed : the technical aspect, and the set up of a validation process for articles. Let’s analyse quickly each point.

We have seen that in some cases publishers removed meaning, hence value, to the content they publish. It is obviously wrong, but let’s say that they do their best to produce a good quality type-setting. There are costs attached to this process, no doubt, but there shouldn’t be an over-estimated value for it because it is roughly only a question of applying community defined standards. This article showing a publishing workflow using JATS is a good illustration. Whereas the latest STM report (which is very interesting by the way) reads that an article costs 1261 £ to be produced (it is 40 % of its total cost), the boss of the typesetting company River Valley Technologies who wrote the article cited above thinks that this estimate is almost four times too high (Ok, he is based in India, which must have effects on his production costs). Speaking of type-setting companies, it is interesting to see how often this tasks are performed not in house but by third parties : Elsevier outsources 75 % of its editorial activity, it’s 40 % for Springer, and it’s mostly to Indian companies… In any case, claiming that there is added value here doesn’t seem very relevant.

If the value is not brought by the technical workflow, which is a matter of costs, does it come from peer-reviewing ? Even if this process remains well perceived, the value of peer reviewing is regularly questioned : it is said to slow down the dissemination of research outputs, and in some cases to be only a mock mechanism that cannot tell true from false. I am not sure that Nature Publishing Group recent release about fast track peer review at additional cost will calm the situation. Once again, peer-reviewing process has costs, even if peer reviewers do no get paid. Nevertheless, not only these costs are often artificially high because of the artificially high rejection rates, but also they should not set by their own the value if we want to be consistent with this « value based pricing » approach.

So where does the value comme from ? There is nothing much left, except the prestige of the publisher’s brand. Elsevier, Wiley, Springer, what beautiful sounding words ! Alas, a recent study from Carole Tenopir showed that the parameter « name of the publisher » has no impact on a researcher when he wants to read, cite, or publish, as opposed to the parameter « journal title ». Very important and even crucial for a researcher who wants to publish, the journal title is just an element among others for the researcher who wants to read, the one university libraries pay their subscriptions for. When he wants to evaluate the trust he can give to an article he wants to read, the researcher indeeds refers to traditional criteria (journal title, reputation and impact factor), but also to criteria called navigational by Tenopir (no pay wall, abstract, consistent looking data, clear methods, references) and social (the author is personnaly known or the article was recommended by a colleague). I could not find any study showing that researchers particularly like platforms agregating hundreds of journals. The 2012 Ithaka UK survey of academics mentions that « specific electronic research resources/computer databases » are important for researchers, but it is not known whether it refers to resources like Pumed, Scopus, the Wos, or to publishers websites. Anyway, my feeling is that researchers do not credit value to the importance to the size of a publisher’s catalogue.

If we stick to the primary mission of academic publishers which is production and dissemination of research journals, we realise that value based pricing is all smoke and mirrors and that its goal is to melt production costs in a vague greater whole, hence pulling up the prices and the operative margins. Nothing shows this better than when a publisher asks for additional fees (platform fees, metadata fees – which fortunately barely occurs for current subscriptions) and passes costs while talking about value. Nothing seems to validate this shift from cost based pricing to value based pricing but the fear that publishers could not be able to justify their prices, either for subscriptions or for APCs.