The Wrong Sort of Open Access

I came across two articles this week on the subject of Open Access and thought I’d share them here with a few comments of my own.

The first article was published in the LSE blog on 11th August with the title Article Processing Charges (APCs) and the new enclosure of research. For those of you not in the know, an Article Process Charge (APC) is a fee that authors are required to pay a publisher to allow Open Access to the paper on publication, i.e. without readers having to pay. The fees for some journals can many thousands of dollars. The lede for the LSE blog post reads:

Drawing on a recent analysis of APC pricing and movements within the commercial publishing sector, Gunnar Sivertsen and Lin Zhang argue that APCs have now firmly established themselves as the predominant business model for academic publishing. Highlighting the inequalities inherent to this model, they posit now is the time to consider alternatives.

In the text the authors reveal that in 2020 alone APCs contributed over $2Bn in revenue to academic journal publishers. I agree with the authors’ conclusion that the APC model is unfair and unsustainable. Indeed I would go further: it’s a complete con. The actual cost of processing an article for publication is a tiny fraction of the APC – the rest is just profit. The academic community is being fleeced. The right “time to consider alternatives” was many years ago, however, when we could have prevented this ridiculous model from being established in the first place. I still believe that the model will collapse under the weight of it’s own contradictions, however, so it’s not too late to change.

The second paper (which was published in January 2022) is entitled Open Science – For Whom? and is published in the Data Science Journal. It was drawn to my attention by the first author, Martin Dominik. Here is an excerpt:

So-called “Open-Access journals” lift the economic barrier to reading scholarly articles, but flipping the paywall from the reader to the author is not a viable solution and inhibits global participation in the scientific process. While article processing charges as well as read-and-publish deals currently on offer appear to be unaffordable to many institutions or individuals (not only in low- and middle-income countries), already the requirement of somebody else having to sign off for getting research published collides with the principles of academic freedom.

and later:

Flipping the paywall is not a solution for scholarly communication in a global Open Science ecosystem. Author-pays-charge models for disseminating research results are not viable in practice and simply absurd.

Simply absurd is right; see the above comments. How on Earth did we let the APC model take hold? I think the answer to that is inertia and lack of imagination within the academic community. It seems many researchers are willing to complain publicly about the absurdity of APCs but far fewer are willing to do something about the situation.

I pointed out the unfairness of APCs in a blog post ten years ago. I ended that post with this paragraph:

I for one have no intention of ever paying an Article Processing Charge. If the journals I publish in insist on levying one, I’ll just forget about the journals altogether and put my papers on the arXiv. I urge my colleagues to do the same.

I’m glad to say that I’ve kept that pledge and have never paid an APC. I recently completed a survey about Open Access which included a question about what level of APC I thought was reasonable; I put zero.

The way forward, I believe, is Diamond Open Access (i.e. free for both authors and readers), such as that offered by the Open Journal of Astrophysics. This is not the only model, of course, but we have at least demonstrated that it is viable (and indeed rather successful). And at least in setting up the Open Journal of Astrophysics I’ve done a bit more than whinge.

8 Responses to “The Wrong Sort of Open Access”

  1. About 10 years ago, Ed Seidel and I met several times with the top people in the APS and IOP publishing divisions, with the aim of persuading them to make a plan for converting fully to OA. That hasn’t happened yet, of course, but during the discussions the APCs came up, and APS opened up to us their budget and costs. At the time their APCs for hybrid OA were around $1500/article, if I recall. I concluded that this represented a fair estimate (at the time) of the cost of doing Phys Rev. It was dominated by the cost of article production, which in turn was dominated by the cost of running the refereeing system. We need to remember that an APC is paid only by an article that gets published, so the APCs in total pay for the cost of handling all the ones rejected as well as for their own refereeing.

    Given all their costs, APS was not making a big profit out of their fee, and of course what profit there was went back to help fund the APS. I think, therefore, that one might use current Phys Rev APC levels as a comparison standard for other comparable physics journals, particularly those published by Elsevier, to get an idea of how much profit other publishers are getting. I would guess something like 50% or more.

    Of course, in the current model the publishers get subscription income as well as APCs on OA articles. That has led to accusations that they are double counting: subscribers pay for the OA articles to be published and so do the authors. Some journals say they have deliberately lowered the subscription charge when the fraction of OA articles got significant. This is one of the complexities of this hybrid system.

    There are lots of ways to change the system, reduce APCs, and still run a commercial operation, but it would be hard for established journal publishers to adapt. In my view, the most important function of the publishing houses is to referee articles, and it is legitimate for them to get paid enough to cover that cost. But today’s publishers add to their costs by re-formatting and editing manuscripts, by printing paper editions, by operating expensive websites that offer many other services to authors and readers, by giving PR to particularly successful articles, and so on. I am not convinced that such activities serve the community and are worth paying for in the APC. Ditching them and simply running a refereeing house, perhaps on top of the arXiv, would reduce the APC dramatically. But I would argue that paying an APC or an article charge for refereeing of submitted articles is fair.

    Diamond OA, like the Open Journal of Astrophysics or like the journal I established in 1998, Living Reviews in Relativity, have no income but they have to deal with costs one way or another. If they do not perform refereeing then the costs can be minimal. Currently LRR (which does do refereeing but only of invited review papers, and which currently has the highest impact factor among all OA journals in all subjects) is published by Springer as a loss-leader, since it adds little to the overhead costs of its sister journal, GRG. So effectively the subscribers to GRG pay for LRR.

    The ecosystem of journal publishing therefore has a lot of variety, but I would love to see someone start an independent commercial refereeing house sitting on top of the arXiv, which effectively gave a * to arXiv papers that its referees approved! What else do we really need?

    • telescoper Says:

      Pardon me for saying so but there seems to be a lot of non sequitur in this comment.

      For one thing, journals do not do refereeing; referees do. And usually they don’t get paid; the added value of their labour is appropriated by the publisher.

      The Open Journal of Astrophysics is an arXiv overlay journal just like you seem to be arguing for in your last paragraph, except that it’s not commercial. It is run as a community service, with editors and referees giving their time for free.

      Check it out here:

      • I can understand that something on the scale of PRL, which I think publishes now more than 50 articles per week and rejects about four times as many, would be hard to run completely as a community service. I think all current arXiv overlay journals are much smaller. But PRL’s APC does seems very high even considering that the accepted articles have to pay the costs associated with rejected papers.

        Even stranger is how PLoS journals, which supposedly are not for profit and have high (~70 %) acceptance rates, have just as high APCs.

      • I also wonder if journals actually spend a lot of time re-formatting and editing manuscripts? In ye olden days, journals had actual typesetters – you submitted a typed manuscript, and when accepted the journal then had to typeset your text (and figures and tables) and print them (just like a newspaper). Nowadays the vast majority of this work is done by the author when preparing the paper using the journal template – yet the page charges for journals did not decrease, and indeed rose year-on-year.

      • telescoper Says:

        Quite. The cost of production has indeed fallen substantially for the reasons you state. This reduction has not been passed on to authors or subscribers, however, but merely resulted in increased profit.

  2. Anton Garrett Says:

    Totally agree, well said.

    • Rejection rates in astronomy are low. I don’t see this reflected in lower fees?

      • telescoper Says:

        There’s also a difference between rejecting a paper on the advice of referees and a “desk reject” which is rather less time-consuming.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: