Open Access Archivangelism

---

SUMMARY: One can calculate the price a subscribing institution pays per article, journal by journal and field by field. The number of institutional subscribers per journal may be listed or estimated, but that's still all revenues rather than costs. If all text-generation, access-provision and archiving are offloaded onto the distributed network of institutional repositories, the only service left for a journal publisher to provide is peer review. The only two factors modulating that cost would be the journal's submission and rejection rates (since the referees are unpaid). That gives a more realistic idea of what Gold OA will cost per article once we have 100% OA (rather than the arbitrary asking-prices we have from today's Gold OA and hybrid Gold "open choice" journals). Green OA self-archiving mandates might have the eventual side-effect of inducing this transition to Gold, but the real objective of OA is not to save money on subscriptions: It is to put an end to needless loss of research usage and impact.This can be achieved by Green OA self-archiving mandates, whether or not they lead to an eventual transition to Gold OA.

---

A fellow OA advocate has just asked me whether I know any research or data on the costs of research journal publication, globally and broken down by discipline and/or journal types.

I had to reply that I wish I did, but even 25 years as editor in chief of a very high impact journal did not give me those figures, even for that one journal!

What is easily calculated, journal by journal and field by field, is the price a subscribing institution pays per article. (That's just the annual institutional subscription price divided by the annual number of articles.)

The publisher's revenue per article is a bit harder to determine: Asking the journal publisher for the number of institutional subscribers may provide it in some cases. Using the average ball-park figure of 800-1200 institutional subscriptions for journal sustainability gives a rough estimate.

But that's still all revenues. Costs are another matter, and not only are those data closely guarded by publishers, but in several respects, their reckoning is arbitrary. There is the usual arbitrary figure of "overhead" and "infrastructure." But apart from that it is very hard to tease out how much the print-run, mark-up, distribution, fulfillment, and advertising cost. And then there is the even vaguer task of estimating what expenses would be left if the paper version were scrapped altogether, and the journal were online only.

And last, and in fact most important, no one can say what costs would be left if there were no online edition either: If all text-generation, access-provision and archiving were offloaded onto the distributed network of institutional repositories, what would be left for a journal publisher to do? To implement the peer review (and possibly a certain amount of copy-editing). The only way to find out how much that would cost, per submitted paper, is not to try to infer and extract it from all of the added costs and services with which it is currently (and hopelessly) co-bundled by conventional publishers, but to see what it is costing, per submitted paper, for an OA publisher that is providing that peer-review service, and that service only.

I suspect that if that figure were looked at directly, in actual cases, the only two factors modulating the size of the cost would be the journal's submission and rejection rates (which might require a separate submission fee and, for accepted papers, an acceptance fee), not the journal's discipline or subject matter. This is because on the service-implementational side (which is all that is being paid for) the only variables are the submission and rejection rates. The thoroughness and rigor of the peer review itself, and the effort put in by referees, will no doubt vary from field to field and journal to journal, but that is not what is being paid for (since the referees are unpaid!). Peer review processing costs are just volume-based.

So I am sorry I could not help with the top-down answer. I do think the bottom-up answer can be derived from actual cases of pure OA journals doing nothing but peer review, or almost nothing but that, today. Then that bottom-up answer can be used to estimate how much would be saved by downsizing today's conventional hybrid (paper/online) journals into such peer-review-only OA journals -- and, more important, it could give us a much more realistic idea of what Gold OA is likely to cost per article, once we have 100% OA (rather than the arbitrary asking prices we have from today's Gold OA and hybrid Gold "open choice" journals -- based usually on dividing their current annual revenues from a journal by the annual number of articles published in that journal).

It does not follow, of course, that established journals will willingly downsize to just the peer-review service and its price! But this brings us back to the far more important and urgent matter of Green OA self-archiving, and Green OA self-archiving mandates:

What might possibly have the eventual side-effect of inducing this downsizing by conventional journals is mandated Green OA self-archiving. The competing functional and cancellation pressure from the free Green OA version might force publishers first to cut needless costs, products and services (the paper edition, then the online edition) and to offload all of those instead onto the network of Green OA IRs. Then still further cancellation pressure might not only force a conversion to the Gold OA cost recovery model, but it would then by the very same token release the institutional subscription cancellation funds that would pay for the institutional per-article Gold OA publishing costs.

Estimates like the ones we've just discussed here -- of the ratio between the current per-article revenue of conventional journals and the per-article costs of an OA peer review service alone -- will give an idea of just how much money would be saved by the cancellations and conversion. A conservative estimate might be 3/1 or 4/1, but the ratio could conceivably even turn out to be as much as an order of magnitude.

The real objective of OA, however, is not to save money on subscriptions: it is to put an end to needlessly lost research usage and impact, so as to maximize research productivity and progress.

The Green-to-Gold transition scenario is just speculation, of course; but as there is already so much idle speculation rampant, I would call it counterspeculation. It is not speculation, however, that the real objective of the OA movement, namely, 100% OA, can be reached by mandating Green OA self-archiving, whether or not it leads to an eventual transition to Gold OA.

Stevan Harnad
American Scientist Open Access Forum