Costs of open access publishing - the Wellcome Trust report

[Liblicense-L Listowner <liblicen@pantheon.yale.edu>]

This message, posted to the SSP and alpsp lists recently, may be of wider
interest.  Mr. Dryburgh has okayed posting to liblicense-l.  Cheers, Ann
Okerson

---------- Forwarded message ----------
Date: Tue, 18 May 2004 20:42:59 +0100
From: Alastair Dryburgh <acd@alastairdryburgh.co.uk>
To: ssp@lists.sspnet.org, alpsp-discuss@mailbase.org.uk
Subject: Costs of open access publishing - the Wellcome Trust
    report 

As the author of the report which provided many of the figures for the
Wellcome Trust report on Costs and Business Models in Scientific Research
Publishing I would like to comment on the report and highlight a number of
areas which it did not address.

The report suggests that open access publishing can be 30% cheaper than
conventional, largely as the result of lower variable costs. This
presumably means no print. A traditional publisher could do this just as
easily if the market justified it. The elimination of print is only the
most obvious of a range of cost issues, which need much more careful
consideration.

The early proponents of open access were most vociferous in their view
that existing publishers made things much more complicated than they
needed to be, and that editorial processes could be radically simplified
so that open access publishing became viable at a publication fee of $500
per article. This point has been emphasised less recently as some OA
publishers, most notably PLoS, have introduced higher article charges with
an emphasis on quality.

There is a question here of how much of the cost of article processing
represents quality which authors recognise to the extent of being willing
to pay for, or things which the publisher holds to be important (such as
the maintenance of a house style) which authors may not value so highly,
or simply inefficiency in the process. The report's recognition of quality
does not go beyond an assumption that higher quality journals have higher
rejection rates and hence higher costs per paper published. It misses the
point that Open Access requires a whole new dialogue with a different
group of customers (authors, institutions and funding bodies) about what
they value and are prepared to pay for. This could have a very substantial
impact on costs.

There are two other points which worry me. There is no allowance for the
fact that not all articles published by open access will in fact be paid
for. Anecdotally, the percentage of cases where neither author,
institution of funding body can or will pay, but the article is published
anyway, can be as high as 50%. We would expect this percentage to reduce
as open access becomes more accepted, but any substantial level of
non-payment would require an increase in the fee to those who do pay.

There is also no allowance for transition costs. The report compares one
steady state (subscription-based journals) with another (open access
journals) with no allowance for the cost of starting or converting titles
to move from one state to the other. These costs will be very material.

Altogether I would feel very uncomfortable advising a publisher to take
action on the basis of this report; there may be the chance of a genuinely
lower cost publishing model based on simplified processes, or they may
not. If there is, the benefits may be swallowed up by transition costs or
negated by large numbers of authors not paying.

Alastair Dryburgh
23 Oldfield Road London N16 0RR
Tel +44 20 7275 8303
www.alastairdryburgh.co.uk
___