Björn Brembs

Current
[estimates](https://www.nature.com/news/open-access-the-true-cost-of-science-publishing-1.12676)
for the cost of subscription articles converge around US\$5,000 per
article. This number is reached by dividing the estimated US\$10b spent
on subscriptions annually world-wide by the two million published
articles every year. Current initiatives aiming for a transition from
subscriptions to gold (article processing charges, APC-based) open
access [emphasize](https://oa2020.org/learn_more/)
[that](https://blog.scielo.org/en/2015/05/26/can-monies-spent-globally-on-journal-subscriptions-be-completely-transitioned-to-an-oa-business-model-to-free-the-journals/#.Wh6YoDdrwUE)
the
[transition](https://ec.europa.eu/futurium/en/system/files/ged/oa_report.pdf)
has to
[be](https://pubman.mpdl.mpg.de/pubman/item/escidoc:2148961:7/component/escidoc:2149096/MPDL_OA-Transition_White_Paper.pdf)
[cost](https://www.nature.com/articles/486439a)--[neutral](https://www.mpg.de/11188620/new-momentum-open-access).

How realistic is a cost-neutral transition from subscriptions to open
access?

There is ample material on how much the publication of a scholarly
open-access article
[costs](https://bjoern.brembs.net/2016/01/how-much-should-a-scholarly-article-cost-the-taxpayer/).
SciELO publishes for under US\$100 per article. SciELO, however,
publishes largely in regions of the world where labor and other
associated costs are comparatively low. How much costs do other
organizations cite? The Open Library of Humanities is [on the
record](https://www.martineve.com/2017/02/13/how-much-does-it-cost-to-run-a-small-scholarly-publisher/)
with about US\$500, Ubiquity, Hindawi and PeerJ are also [on
record](https://www.nature.com/news/open-access-the-true-cost-of-science-publishing-1.12676)
within the US\$100-500 range and Scholastica, RIO Journal, Science Open,
F1000Research or arpha mention similar figures. Thus, there are now
about ten different companies or organizations which all agree that open
access publishing costs about US\$100-500 and
[Independent](https://www.dtc.umn.edu/~odlyzko/doc/libpubcomp.pdf)
[analysts](https://research.science.ai/article/on-the-marginal-cost-of-scholarly-communication)
agree: it is established that the actual costs of publishing a scholarly
article are in the low hundreds of US\$.

Thus, the maximum we could theoretically get out of a transition deal
are about 90% savings as an upper bound for potential savings. This
would result in about US\$9b annually which we could, for instance,
invest in modern technology for our publication
[infrastructure](https://bjoern.brembs.net/2015/04/what-should-a-modern-scientific-infrastructure-look-like/).
If there is such a huge potential for savings, why do all these open
access initiatives only aim at a "cost-neutral" decision? Surprisingly,
when planning their transition, these initiatives never looked at costs,
only at current prices. Current open access prices (for those journals
where they are charged) come to lie around US\$2500 and US\$3500.
Because the most expensive journals such as, e.g., *Nature* (which is
[on
record](https://www.publications.parliament.uk/pa/cm200304/cmselect/cmsctech/399/399we163.htm)
to have to charge US\$50,000 per article to maintain current revenue
levels) are not yet open access and prices are expected to rise over the
time of the transition period and beyond, these initiatives calculate
with higher APCs.

There are three main components as to why legacy publishers are charging
so much:

1.  Inefficiency -- with a profit margin of 40% and a market that
    carries essentially any price increase, there is little pressure to
    cut down on costs
2.  Paywalls -- legacy publishers have to maintain a huge infrastructure
    related to preventing access. The internet was designed to enable
    access, not to prevent it, so the technical and administrative
    challenges are huge. It is thus credible when publishers routinely
    justify their scheme of big deals, which enclose all of their
    journals, to be cheaper than any smaller selection of their
    journals, by the increased costs of making sure they can accurately
    distinguish whether a user from university X has access to journal Y
    and not journal Z. The
    [frequent](https://rossmounce.co.uk/2017/02/14/elsevier-selling-access-to-open-access-again/)
    [mishaps](https://rossmounce.co.uk/2017/02/20/hybrid-open-access-is-unreliable/)
    with [paywalls](https://rossmounce.co.uk/category/paywall-watch/)
    for nominally open access articles are a testament to how difficult
    and hence expensive something like this is to maintain. Of course,
    therefore, it [makes great business
    sense](https://scholarlykitchen.sspnet.org/2017/11/20/creating-safety-net-double-dipping-wrong-term-right-approach/)
    for a company that still has paywalls, to have all revenue
    contribute to these costs, even open access revenue.
    "Double-dipping" is a normal, sound business practice which
    contributes a fair amount to the high open access prices (APCs) we
    pay today.
3.  Profits -- legacy publishers are accustomed to profit margins of
    about 40%, i.e., about US\$2,000 of the US\$5,000 they are
    collecting per subscription article, on average. So these kinds of
    sums are added to any cost-based APC of legacy publishers.

Even for publishers without paywalls it may be tempting to just ride on
the wave of high APCs just because the customer-base provides for that
sort of revenue. For instance, Emerald recently increased the APCs by
\~70% merely for such [competition-based
reasons](https://poynder.blogspot.de/2015/07/emerald-group-publishing-tests-zen.html).

From this list one can estimate a lower bound of what a transition to
open access should at least yield, if the transition was totally botched
and these initiatives didn't get any advantage at all into their scheme.
Let's assume legacy publishers are half as efficient than their modern
competitors, so their cost would be US\$1000 per article and there
would, as now, still be no competition that would force increased
efficiency. Let's hypothetically assume the public purse is generous
enough to leave current profit margins at outrageous 40%, then the APC
*price* would amount to US\$1,400 per article. There are no more
paywalls to cross-finance so these costs fall by the wayside.

Thus, from these calculations, the worst case scenario for a transition
to open access is savings of about 72% and the best we could do is about
90%. Anything worse than 72% is a free, tax-payer funded gift to an
international oligopoly. Why would the scholarly community support such
a give-away?

Hence, what everybody should be asking of these initiatives is why they
are campaigning for a cost-neutral transition to US\$5,000 per article
with legacy publishers, when there are many competitors that would
perform the same service for 500€? How can they justify a tenfold cost
to the taxpayer in favor of big, international corporations (does
anybody know if any of their names have cropped up in the Panama or
Paradise Papers, btw?), to the disadvantage of smaller, modern
publishers? What are their reasons to prop up a legacy industry with an
obscene waste of tax funds? What service could these legacy publishers
possibly perform that one may use to explain to a non-scientist
tax-payer why they should pay ten times as much?

A different question is, of course, how to achieve this reduction. For
now, only one component stands out as crucial for the transition: the
exchangeability of service providers. Only this exchangeability allows
for actual competition to put pressure on costs. One of the many
afflictions of our current system (and one shared with [APC-based gold
open
access](https://bjoern.brembs.net/2016/04/how-gold-open-access-may-make-things-worse/))
is that publishers cannot easily be switched, as every article resides
only with one company. Merely switching subscription funds to APCs would
not make the service providers exchangeable, because journal rank would
still force authors to publish where their field dictates them to,
preventing competition. One of (hopefully) several solutions would be a
shared publication
[infrastructure](https://bjoern.brembs.net/2015/04/what-should-a-modern-scientific-infrastructure-look-like/)
where service providers are chosen in a [bidding
process](https://bjoern.brembs.net/2017/10/academic-publishers-profiteering-servicing/)
and can be replaced if their services become too expensive or their
cost-effectiveness drops.


Current estimates for the cost of subscription articles converge around US$5,000 per article. This number is reached by dividing the estimated US$10b spent on subscriptions annually world-wide by the two million published articles every year. Current initiatives aiming for a transition from subscriptions to gold (article processing charges, APC-based) open access emphasize that the transition has to be cost–neutral.

Messaggi di Rogue Scholar

Is a cost-neutral transition to open access realistic?