Re: Study Identifies Factors That Could Lead to Cancelled Subscriptions

[Sandy Thatcher <sgt3@psu.edu>]

I am coming new to this list, so am not aware of what may have 
preceded this part of the discussion. But as President-Elect of 
the AAUP (Association of American University Presses) charged 
with preparing a white paper on OA for the Association (now in 
its third draft), I may have something useful to contribute from 
our perspective.  I am aware of Dr. Harnad's long record of 
advocacy in this arena and have recently initiated a dialogue 
with another OA apostle, Peter Suber, whose views are certainly 
well thought out and serve as a good touchstone for debate on 
this subject. (Perhaps he is subscribed to this list also?)

Not knowing what may have been discussed previously, I begin by 
asking whether this list has focused any attention on the 
relatively new study from the Publishing Research Consortium 
titled "Self-Archiving and Journal Subscriptions: Co-existence or 
Competition? An International Survey of Librarians' Preferences" 
(November 2006), accessible at the PRC's web site: 
http://www.publishingresearch.org.uk

This is the most sophisticated study I have ever seen of this 
subject, and it speaks directly to the point that Mr. Esposito 
raises, by focusing our attention, not on what has actually 
happened so far, but what may be expected to happen in the future 
given the attitudes of librarians as revealed by this survey. If 
publishers look at this study, they are going to be very worried 
about the future if it moves in the direction of more 
peer-reviewed material being freely accessible through author 
self-archiving or institutional repositories. They should even be 
greatly worried if the FRPAA passes Congress and mandates a 
6-month delay before posting because this study clearly shows a 
marked difference in librarians' willingness to cancel 
subscriptions if the embargo is reduced from 12 to 6 months.

Another very interesting finding for me, as the publisher of 11 
journals in the humanities and social sciences (10 of them 
available through Project Muse, which we were the first 
university press to join in 2000 when it opened up beyond Johns 
Hopkins, which hired away our journals manager in the mid-1990s 
to help get this venture started), is that librarians care a lot 
that the material is peer-reviewed but care very little whether 
they have access to the final published version.

Not long ago we adopted a policy at our Press that allows 
contributors to our journals to post their articles, once 
accepted but before copyediting, to their own sites or their 
institutions' repositories. So as not to undermine Project Muse 
and to provide continued incentive to librarians to subscribe to 
Muse, I thought-naively, as it turns out-that librarians would 
value the final version highly enough not to cancel subscriptions 
if only the peer-reviewed but not final versions were available 
in OA mode. Well, how wrong this study shows that assumption to 
have been! Librarians seem to place little or no value on the 
final processing of manuscripts after acceptance, which should be 
an eye-opener to publishers like us who all along felt that we 
were adding further value to the article after acceptance. I 
certainly will consider changing our Press's policy soon and, one 
other presses get wind of this, I'm sure many of them will as 
well.

All of which underlines Mr. Esposito's point that it is 
perception that matters here, not reality. Once we publishers 
think something is going to happen, we will act on those beliefs 
if they seem to be firmly supported, by such studies as the 
PRC's. What this study really is all about is identifying the 
factors contributing to the "tipping point" i.e., when behaviors 
will start to change based on beliefs, however erroneous they may 
be. (By the way, the PRC study directly confronts the "evidence" 
of the physics preprint archive not affecting cancellations of 
physics journals, by pointing out that the archive combines 
peer-reviewed and not peer-reviewed materials, thus making it 
less than fully reliable as a source of completely authenticated 
work in the field.)

I think the tipping phenomenon, which we know already to have 
shown itself operative in this arena when e-journals came to 
displace print journals as the main product in the marketplace 
(rather more quickly than many people anticipated), is extremely 
important to keep in mind here. This is what I see as a real 
possibility: enough of the major commercial journal publishers in 
an ever more consolidated market (after the purchase of Blackwell 
by Wiley) become convinced that their subscriptions will erode 
seriously (if, say, the FRPAA becomes law) and therefore decide 
to abandon the arena of STM journal publishing because they 
cannot sustain the expected profit margins under the new regime 
(as outlined by Dr. Harnad).

This could all happen very quickly, as "tipping" phenomena 
generally do. Where would that scenario leave the academy? With 
several thousand journals suddenly left to fend for themselves! I 
suggest that the infrastructure of universities today is simply 
not prepared, in any shape or form, to deal with that "crisis" 
and find some way of sustaining those journals. Certainly, the 
existing university presses would not be able to do so. SPARC 
couldn't handle any such burden. Self-publishing would then 
proliferate, and chaos would ensure for some time to come. Are 
librarians prepared to deal with the consequences?

I do not depict this nightmare scenario in order to defend the 
existing system. I have no personal stake in the persistence of 
the existing system (except to the extent that I serve on the 
board of the CCC), and indeed I have written in favor of an 
OA-type system for much longer than that term existed, going back 
to articles I wrote for the Chronicle of Higher Education and the 
Journal of Scholarly Publishing in the early 1990s. But I do 
think university faculty, administrators, and librarians need to 
think through these issues and possible scenarios very carefully 
and "worst-case" planning would probably be appropriate here.

The system as it exists even now, as Peter Suber points out to 
me, has many elements of the "gift economy" involved in it, with 
faculty donating their time to peer review, universities 
supporting editorial offices at no cost to the publishers, etc. 
But the scenario I depicted would increase the burden on 
universities exponentially in a way that could have very serious 
short-term consequences for the whole system of scholarly 
communication.

I long ago predicted that university press journals would migrate 
to the electronic environment much more quickly than monographs 
because what university presses do for journals is much more a 
functional service than a full publishing support system. In the 
print world, presses typically provided the services of design, 
typesetting, printing, marketing, and subscription fulfillment. 
They did not, however, control the peer-review process, and many 
did not even offer copyediting, replying instead of what the 
editorial office provided by way of copyediting (sometimes by the 
journal editors themselves, often by grad students, only 
occasionally by real professional staff). Perhaps this is the 
reason librarians do not place much value on the copyediting that 
goes into journals! It was therefore much more possible, and more 
likely, that journals could spring up online without the support 
of publishers, if they went OA and did not have to bother about 
the complications of outsourcing orinting and handling 
subscription fulfillment. (And a journal only has to be designed 
once, and the template followed thereafter, while marketing takes 
care of itself if the journal is aimed at a niche community 
anyway.) Books, however, cannot so readily migrate to the 
Internet because the infrastructure of book publishing is much 
more controlled by publishers and other vendors in the supply 
chain (literary agents, wholesalers, retailers,warehousing and 
distribution suppliers etc.). And hence it is no surprise to us 
that the migration of monographs to the Net has lagged far behind 
that of journals, But it is a cause for increasing concern, and 
such groups as Ithaka are now trying to find ways of closing this 
new "digital divide." So, looking forward, I have an interest in 
how OA will affect books as well as journals. In the world of 
knowledge, the container should not matter as knowledge itself is 
seamless, but as it is now, there is very little interactivity 
online between book and journal content.  The problem of going OA 
for books, though, is a lot more complicated than it is for 
journals. If it costs $2,500 for an author to publish in a PLOS 
journal today, you can bet it will cost a great deal more 
topublidsh a book in a full OA mode. Taking away the cost of 
printing, binding, and warehousing (which constitute about 30% of 
the overall cost of publishing a monograph), you are left with 
still supporting about $20,000 in costs for the average-length 
uncomplicated monograph. Will universities be willing to pony up 
subsidies on that scale, especially in the humanities? (And it is 
really the humanities we need to worry about. generally speaking, 
scientists don' need to publish books, and publishers encourage 
them to do so mainly when they think there is money to be made. I 
know because I used to be editor-in-chief at Princeton U.P. where 
we did a lot of science book publishing.) If the pressure to move 
in that direction arises-which in principle we presses do not 
oppose because we too believe that it is our mission to 
"disseminate knowledge far and wide" in the most economical way 
possible-it will have to be managed very carefully so that 
inequities that now exit in the toll-access system for journal 
users are not re-created for authors in a full OA system for 
monographs, where only faculty at the richest schools can afford 
the subsidies required to publish.

Sorry to go on so long, but I felt it important to lay some of 
these issues about OA on the table from a university press 
perspective. I would welcome reactions, particularly from Dr. 
Harnad and Mr. Esposito, and hope that everyone will take a look 
at the PRC study if they have not already.

Sandy Thatcher
Director, Penn State Press

P.S. We are embarked on a quasi-OA experiment in humanities 
monograph publishing at Penn State through our joint 
Press/Library Office of Digital Scholarly Publishing, but more 
about that later, if anyone is interested. (The Press is now, 
administratively, part of the Library, so I feel more comfortable 
now contributing to a list initiated by librarians. Ann Okerson 
and I go back a long way....)

_______________________________

Joe Esposito wrote:

> I am inclined to think that Professor Harnad has "the question" 
> wrong. It is not to seek evidence that is irrelevant; it is for 
> the managers to pursue the interests of the ownership of their 
> publications.  The evidence is irrelevant because (a) decisions 
> will be made (have to be made) before the evidence comes in, 
> which is why we associate the word "risk" with investment; and 
> (b) even if the evidence unequivocally demonstrated that OA does 
> not result in a decline of subscriptions, the management of a 
> publication may determine that OA is still not in the interest 
> of their ownership. For example, the publisher may begin to 
> market back issues separately for an incremental fee.  There is 
> in fact no situation that I can think of where a toll-access 
> publication can ever benefit from any form of OA beyond limited 
> product-sampling. Thus for the publisher of such a journal to 
> have some portion of the publication become OA is a breach of 
> fiduciary duty.
>
> There are, however, circumstances that are wholly appropriate 
> for OA. Examples of these are BioMedCentral and the Public 
> Library of Science, which have established revenue models that 
> absolutely require that their publications be OA.  Whether these 
> models will be sustainable long-term remains to be seen, but I 
> for one am rooting for them.  For these models the principal 
> beneficiary of a publication is the author (who thus pays), not 
> the reader (hence OA).  It is my view that the long-term future 
> of academic research publishing will be a sophisticated 
> extension of what BMC is doing today.  (BMC may or may not make 
> it to that future point, but it is showing the way.)
>
> The one form of OA that benefits no one and should not be 
> supported by any responsible individual is so-called 
> self-archiving, which I prefer to call informal publishing. 
> The problem with informal publishing is that it cheats:  it 
> wants the infrastructure of the formal publication without the 
> attendant costs and responsibilities. If the formal publication 
> were to disappear, could the informal publication (that is, an 
> editorially similar, if not identical, version of the formally 
> published article) exist?  I think not. This is parasitic 
> publishing.
>
> Unfortunately, this form of OA adds to costs in the form of
> institutional repositories (an emerging budget item for more and
> more libraries) and in evolving services whose objective is to
> identify the authorized version of an article when a multitude may
> be strewn across the Internet.
>
> So, OA, yes; toll-access, yes; but self-archiving, no.
>
> Joe Esposito