Tuesday, January 30, 2007

Value perception

This is not so much about open access per se, but about the perceived value of journals, or, more precisely, the perceived value of articles published in journals.

From comments I come across on email lists and blogs, I detect two – conflicting – trends. One is the growing tendency to put a value on a journal according to the number of article downloads; the other a desire to base journal pricing on the actual production cost, i.e. the actual cost to the publisher of publishing an article.

Why are these trends conflicting? If it’s not immediately clear, let’s look at the logical consequences of these trends. If the value of a journal depends on the download figures (often, but in my view erroneously, called usage figures), it seems to come with the expectation that if downloads are low or decreasing, the value, i.e. the price, should be low or go down. This would be fine, were it to mean that a high or increasing number of downloads makes a high or increasing price the logical and acceptable consequence. I’m afraid I don’t see comments in this vein in the email lists and blogs I follow.

But this is not really the conflict I had in mind. That lies in the fact that production costs and downloads have no relation whatsoever. If they seem to have one, it’s not unlike the relation between the human birth rate and the stork population (currently declining in both cases, in any case in Western Europe). The cost of coaching an article through the peer-review process and of publishing it is independent of the number of downloads or other usage metrics it clocks up once published. It can be argued that one of the woes of the current subscription model is that it already has some characteristics of a cost-based model. Those characteristics are largely responsible for the price spiral of the last decade. There hasn’t been a concomitant income spiral for publishers. Due to unremitting annual cancellation rounds, if publishers wanted or needed to maintain the same income to keep a journal going, they had to secure that income from fewer subscriptions. Year on year. Voilà, the serials crisis. Not to be repeated or exacerbated, I would have thought.

The problem with value based on downloads or usage is different. The beauty of journals is that the decision to publish any given article is purely a scientific one, taken by the editors. Commerce doesn’t come into it. Should value be defined by downloads, then it is inevitable that decisions to publish will be influenced by the perceived ‘download potential’ of articles. Which is not the same as scientific significance. A glimpse of what could happen can already be seen from the effect of impact factors. Only the rather long timeline and slow effect of impact factors keeps unwanted developments in check, but there have been several cases of editors insisting that in their submitted articles, authors must cite other articles from the same journal as a condition of getting accepted for publication. The current journals system is rightly criticised for its built-in conservatism and the fact that unconventional science is difficult to publish. Imagine the consequences of difficult or esoteric concepts and theories experiencing such difficulties, too, just because they may only be understood by a limited number of scientists and are thus expected to have less than average downloads. Not a prospect to relish.

Open access publishing offers solutions. The process of peer-reviewing and formal publishing is valued, rather than usage. Costs are proportional to research activity. Esoteric and difficult to understand science has no problem being published. And, most important of all, anybody, anywhere, who wants or needs the material, has unimpeded access to it.

Jan Velterop