Re: A word on calculating costs

["Anthony Watkinson" <anthony.watkinson@btopenworld.com>]

I cannot resist rising to Adam's challenge but I also cannot (alas) claim
to be an expert in this area - rather an observer. I do however run an
annual seminar on journal production for STM and so have been trying to
keep up with what is going on.

I think we have to take online editorial systems separate from production
workflows. There has always been this divide in publishing between what
happens before the article is accepted and what happens after it has been
accepted. There is some evidence that there are now hooks being developed
at the end of the EO process which join us with the front end of
production systems but this is fairly new. There are plenty of reasons why
this is new but in the end they boil down to the ways in which publishers
interact with journal editorial structures.

Editors see themselves as being in charge of the processes leading up to
acceptance. In the past publishers made a great point that they never
interfered with the refereeing process. Nothing is now so clear cut. For
example publishers mostly now claim to the right to refuse to publish
articles, which may have been accepted, but which are unacceptable for
legal reasons. They are also now offering online editorial systems, which
mean that in practice they have a window into the refereeing processes
that they did not have before. These systems are not cheap. As I
understand it there is usually a set up cost and a charge per paper
processed but obviously the financial deal depends on the particular
system concerned, the clout of the publisher and the publisher's
commitment. There are thus some economies of scale but not a lot. These
are new costs.

Where there should be serious savings is in the back-up to journal
editors. If the editor works with one of these systems and entirely
electronically a whole area of cost of secretarial back-up, postage etc.
should disappear. In some cases it does and in some cases it does not.
There are still academics who want secretarial back-up to print out the
e-mails and then be told how to answer them. If these people are top
academics they can get away with it.

If we look at the production process large publishers at any rate have
worked hard to create standard work flows and standard modes of
interaction with suppliers which have undoubtedly saved a lot of money. I
was recently in India where I was able to discuss such matters with two of
the biggest suppliers of typesetting to the STM publishing community. The
level of efficiency and the way the publisher and the supplier now work
together with in some cases the same scheduling processes is a real
breakthrough. Publishers would love to have one work flow but in practice
of course there is always a level of customisation for established
journals to take account of the way they want to work. I assume that BMC
can insist upon one way of working (their way) for the new journals they
are working. What has happened to the savings made. They have usually been
invested back into the huge online developments, all the features of the
big electronic platforms and all the extra staff needed for the purposes
of negotiating with consortia etc. I am not sure whether or not production
editors can now handle more journals than they could before. One would
assume so but any reduction in that area of staffing has certainly been
balanced by all the other people who have been taken on.

Why do publishers outsource and not develop the sort of systems mentioned
above in-house as BMC tends to do?  At one time there was a lot of
reliance on desk top systems (to give one example) and there were gurus
who explained that the whole typesetting (composition) industry was going
to disappear. Now outsourcing (other than core processes) is the norm even
for the very largest publisher. As I understand it the argument is that
the constant development of the systems as expectations constantly
increase is very difficult to fund if you want to remain cutting edge. If
you take online platforms Wiley and Elsevier have developed their own but
Springer and T&F have gone to specialists.

Anthony

----- Original Message -----
From: "Joseph Esposito" <espositoj@gmail.com>
To: "adam hodgkin" <adam.hodgkin@gmail.com>
Sent: Tuesday, January 11, 2005 11:51 PM
Subject: Re: A word on calculating costs

> If the cost per article drops by 50 percent, and the number of articles
> increases by 1,000 percent, what does this do to the total cost of
> scholarly communications?  The current user-pays model has the effect of
> suppressing production, which, provided that most editors are reasonably
> competent (as I for one believe they are), is a very, very good thing.
>
> Joe Esposito
>
> On Tue, 11 Jan 2005 20:38:04 +0000, adam hodgkin <adam.hodgkin@gmail.com>
wrote:
>
> > I think that Jo must be right -- the 'true' cost of an OA model is very
> > hard to predict when so many factors have to be considered. He is right
> > the situation is very dynamic!
> >
> > However, it is not necessary to predict the actual exact cost of
> > anything. The point surely is that there are many reasons to expect the
> > cost (per article-published or per researcher-reading) of the
> > increasingly OA system to be significantly (yes at least an Order of
> > Magnitude) lower. We dont have to get a 'correct' answer to know that
> > the ground has a decent slope. Weather forcasting is very complex, but
> > we now have sufficient evidence to be sure (?) about the reality of
> > Global warming.
> >
> > I didnt mention the Google 10$ per book figure because I thought that OA
> > journal technology would be done in the way Google will scan library
> > books. Of course not. The surprising thing about the Google approach is
> > that they are NOT apparently doing most phases of the process
> > dramatically/technologically different from the way Gale or ProQuest
> > have done large scale digitisation projects in the past and they STILL
> > expect to get this massive cost reduction.
> >
> > re Anthony Watkinsons points. Are we interested in seeing BMC's audited
> > costs? Possibly, but even without sight of this audit I do know that the
> > way they are building their Refereeing and Production systems is
> > intended to be scaleable in the sense in which a web-based publishing
> > system should be scaleable. Vitek Tracz in a recent interview said that
> > they expected their throughput to double each year. Will BMC's cost base
> > move at anything like the same rate? My guess is that they are tooling
> > up to handle four times the papers very efficiently (ie with not many
> > more people than now). Perhaps Anthony can tell us if any of the big
> > conventional publishers are yet on top of this?
> >
> > Adam