RE: Sub-sidy/scription for ArXiv

["Nat Gustafson-Sundell" <n-gustafson-sundell@northwestern.edu>]

Sorry, I just returned from a conference, so I'm just thinking 
about this now.  I don't disagree with any of your points, 
although I continue to have an issue with the perceived cost of 
peer review. I'm sure that the American Physical Society made a 
good faith estimate of their costs (the link from your article to 
the slideshow didn't work, so I can only guess) and other 
publishers can probably whip up even higher costs, depending on 
the formulae they construct and the dead weight they carry, but 
continue to think is necessary or can't think of operating 
without and maybe really can't operate <as such> without.  As I 
said, I think there are all sorts of ways to make publishing, or 
any enterprise, expensive.

I've seen many numbers fresh out of the opposite end of 
somebody's blackbox, back when I did financial work, and I've 
learned to believe in not a one. Aside from simplifications, 
estimates, etc. that go in, the actual operations are generally 
assumed.  My experience is that you have to knock hard on every 
single number that goes into a financial formula.  Usually, you 
can come up with a list of priorities for saving (just on 
operations), if you don't end up also finding mistakes or 
exaggerations hidden in the formulae themselves.  (Honestly, 
though, I think you need to be inside the walls of an operation 
for some time before you can see what or who, within an 
operation, is a structural support and what or who is an 
expensive decoration).

Many (most, all?) library and university hosted OA journals do 
not pay editors or staff to manage peer review and thus do not 
pay money for peer review.  We can assign an imaginary dollar 
figure for the cost of this peer review, but that would be like 
coming up with an imaginary dollar figure to describe the cost of 
writing the article -- it just isn't meaningful (although we 
could talk meaningfully about the cost of the research). Scholars 
choose to serve as OA editors for journals probably for 2 main 
reasons: 1) it is helpful for their careers, 2) open scholarly 
communication matters to them; regardless of motive, however, 
they are generally providing the value at no dollar cost (how 
many OA journals pay a stipend?).

You know this already.  It is beside your point.  Yes, more and 
more open access journals are appearing.  More and more libraries 
are getting into the business of hosting journals and providing 
the 'publisher' infrastructure and staff to support peer reviewed 
journals in varieties of fields (where the cost conversation has 
more meaning, but many library publishing specialists are simply 
adding this work to what they already do; or, in any case, the 
costs are much lower as shown indirectly by page cost studies for 
NFP and OA journals).  This has been much of the growth of OA 
which, while really quite impressive, you have elsewhere 
described as glacially slow.

I know you already know this, but I wanted to stress it again 
because this is what I was largely thinking about when I posted 
my earlier comments -- as library and university publishing 
programs continue to grow, and there's no reason to believe they 
won't (there are several big university libraries now in the 
business), they can and should think about economies of scale, 
shared standards (efficiencies, as well as improvements in such 
areas as metadata), further sharing infrastructure, and yes, I'll 
say it again, improving the value chain.  Also, as you point out 
(and which I called overhead and profit-taking), there are lots 
of other reasons why commercial publishers are expensive -- and 
generally, these causes of expense do not apply (or apply as 
much) to OA publishers.

Until we see the mandates and the effects you describe, the 
journal-by-journal growth of OA is extremely valuable.  I 
understand OA pursued in this fashion is unlikely to overtake the 
fact of big-name retail journals, except perhaps on a long time 
scale, but change does happen / has happened ( in the past three 
days, I heard interesting stories about library faculty liaisons 
in the *humanities* getting the go ahead from faculty to let 
print go -- now, if that is finally occurring ...).

Coming back to your assertion that there is and will be no need 
to re-build peer review providers: I don't know.  It is just too 
easy to pluck a number from the clouds and say it is real and to 
base fees on it -- you might as well tell me peer review costs 
$1300 per article as $500 or $200.  Given the stark fact of 
commercial journal inflation over the past x years and the 
sometimes ludicrous defenses of that inflation, universities have 
no reason to continue trusting those out-sourced service 
providers, regardless of whether the numbers are really real for 
those particular publishers.  Yes, I understand, we still largely 
have no choice but to buy retail journals in my scenario, given 
the fact that peer review is not the only value being added in 
the chain by publishers (the other big one being reputation or 
career effect ... sorry, I'm retreading over the ground of my 
previous memos), but universities only benefit by building 
internal services (which can also be yardsticks for external 
service) and by seeking to make such internal operations as 
efficient as possible -- the better both to judge external 
service providers and, perhaps too slowly, to replace those 
providers as opportunities arise.

Coming back to my science fiction based on arXiv:  I do think an 
innovative OA subject repository slash journal platform 
(...depending on the enhancements, as I said) could impact the 
pace and direction of OA growth. Not only would such a platform 
provide further proof of concept, but it could enhance 
collaboration in building better underlying/ shared systems, thus 
perhaps making further advances more likely at an increasing 
pace, a bit like the way factory production lines increased the 
pace with which horseless carriages replaced horsy carriages ... 
but I won't go down this path a third time, since I see your 
point that immediate mandates would get us further faster.

Let me re-frame my position this way:  I think you are and have 
been proposing the way of the hare, while I continue plodding 
along with the tortoise (seeing opportunities for the tortoise to 
move along a little faster).  If either wins, we both win. 
Since I'm not convinced yet that the hare won't stop to take a 
nap, I'll continue walking with the tortoise, but I do hope to 
hear cheering up ahead.

-Nat

-----Original Message-----
From: owner-liblicense-l@lists.yale.edu
[mailto:owner-liblicense-l@lists.yale.edu] On Behalf Of Stevan Harnad
Sent: Saturday, January 30, 2010 8:48 PM
To: liblicense-l@lists.yale.edu
Subject: Re: Sub-sidy/scription for ArXiv

On 28-Jan-10, at 8:27 PM, Nat Gustafson-Sundell wrote:

> I can't conceive of peer review per se costing billions when
> peer review itself is 'free' and can be managed with little
> overhead and relatively few salaries

Managing peer review costs about $500 per article today
http://www.nature.com/nature/debates/e-access/Articles/harnad.html
and about 2.5 million peer-reviewed journal articles are
published per year (though I think we can get the price quite a
bit lower than that)
http://users.ecs.soton.ac.uk/harnad/Temp/peerev.pdf

By the way, that's still only a fraction of the total cost per
article today, but that includes obsolescent products and
services like the paper and online edition with which peer review
is currently co- bundled...

> One monstrous megajournal does indeed sound scary... the big
> commercial publishers publish separate titles that have
> separate identities within the same fields.

There's a trade-off between economies of scale (and journal-fleet
publishing) and the need for journal autonomy, variety, and
competition.

> a distributed network of local IRs actually pushes up the
> price-tag of the whole system

It is not at all clear that that's true, especially since the
institutions' online infrastructure is already used for so many
other things anyway.

> indexers and harvesters... may also, if commercial, exploit
> their dominance at some point).

Once most or all of OA's target content has been made (Green) OA,
I'll bet on the universities' yearly crop of bright grad students
to beat any commercial indexer/harvester any day of the week.
(And notice that the biggest and best of the commercial
harvesters today, Google, does not charge the user...)

> The other issue, where we clearly see different futures, is
> that the cost of subs for current content might actually be
> pressured up in the cases where publishers lose the revenue
> they've made for providing ongoing access to back-content.

My bet is that institutional cancellation pressure (from user
preference for the self-archive Green OA versions) will induce
cost- cutting not price-raising. But either way, it no longer
matters, once we have universal OA (which will make journal
subscriptions no longer the inelastic necessity they are now).

> universities.. will... need to increase their IR support
> budgets (since, presumably, the IRs begin to be used and this
> has some effect on the expense of supporting the IR).

No, once you set up the free IR software on a server, cost does
not go up significantly, either with more depositing authors or
with more users (though most of those will not be local either).
Total annual institutional peer-reviewed journal article output
is not that big, though it might need some more disk space and a
server that can handle a reasonable amount of download pressure.

> I see no reason at all to believe publishers will re-size fees
> under the scenario you provide, precisely because it is not
> just about content - it is about the other services in the
> value chain (specifically quality control and career effect due
> to title reputation, upon which those publishers that do
> profiteer can continue to profiteer).

That's all conferred by the journal's peer-review standards,
whose costs, as noted above, are only a fraction of the cost per
article that subscription revenue pays for.

> I think your point has much more force if all content is
> mandated to become immediately available, but now I think we're
> really talking about Spanish castles in the air ...

Stay tuned...

> if immediate deposit mandates did become a reality and the
> scenario played out as you picture it, my guess is that we'd
> end up needing to build the peer review service providers you
> mention

A journal is just a title -- i.e., an authorship, readership,
refereeship, editorial board and track-record. Titles migrate.
Either they will stay with their original publishers (downsized)
after universally mandated Green OA or they will migrate to other
publishers. But in any case, no need to re-build or re-invent
peer review.

> More universal availability of back content will certainly give
> universities more wiggle room to build the future (since more
> options on a going-forward basis become more practical).

I like the idea of calling embargoed access "back-content"!

And you're quite right that universal immediate-deposit mandates,
even with delayed OA, will leave wiggle room: and that wiggling
will soon work its way the universal immediate OA that is
optimal, inevitable (and already long overdue) for research,
researchers, their institutions and funders, and the tax-paying
public that funds the funders and institutions.

Stevan Harnad