From: "Hamaker, Charles" <[log in to unmask]>
 Date: Sat, 17 Dec 2011 07:41:37 +0000

Sandy Thatcher said: there are still choices librarians can make, such
as between buying into a big deal and providing access "by the drink,"
as I believe you pioneered years ago, right?  At some point, I'd
imagine, depending on rates of usage, and the per-item cost, the
latter becomes a more attractive option.

I do believe we now should be moving to "by the drink" options when
that's the best option. I have an example where we are moving 75% of a
publisher's cost basis to "on demand"  And although I explained to
faculty that we were just trying to be more cost effective, the first
and immediate response I got was "but I publish in that journal!"

So we are dealing with more than simple supply issues. There are
issues of prestige, recognition (my library subscribes to "my"
journal). Or there's the classic from Darby Orcutt, who says faculty,
when he went to them about how low usage was of some journals said,
"but our graduate students should be reading those titles". But that
they weren't was clearly a function of what the graduate faculty were
requiring of the students. So we need to make a number of things "go
right" to move to on demand article level purchasing. One of those
conditions, outrageous pricing, is already in place. Recognition of
that pricing on the part of librarians, and then hard numbers to go to
faculty with. Some of them do pay attention when we have our facts or
ducks, in a row.

Another condition may be making sure each faculty article is
represented in the on campus repository, so the whole campus knows of
their achievement. We do it with books all the time, i.e. dear vendor,
send me all books published by authors on my campus. Why not do the
same with journal articles to recognize faculty achievements? Make it
as standard a procedure as prestige book purchases?

The ILL/Document Delivery option for articles  has improved so
markedly in the last two to three years, with a simple click to
forward, meaning its now possible to have much quicker than 24 hour
response. And the per use cost numbers are so dramatic in some
instances, that even nay sayers, I would say even the publisher
associations who act like ILL is anathema, should be convinced.

One of the barriers to having hard cost data is that too many
publishers have opted out of COUNTER's journal report 5, which gives
current year only use data.  I.e. about the only measure we've had
till now is Cost Per Use (CPU)measured across the whole run of a
journal. But what we are actually paying for is thgee current year of
the journal. What few brave publishers who have provided us with
current year use, need to abe commended for providing full disclosure
for their customers in terms of what we are actually paying for. Too
many refuse to provide that, since its an "optional" report. But my
my, it is revealing.

My CPU for one large publisher, across all years content is about
$53.00 (publisher A) an article. But if I take what I know to be a
comparable publisher's current year only use data, and apply that
figure (18%) to publisher A, I get a cost per use of $200.00 an
article across all journals from that publisher.  Aand if I just look
at the top half of that average CPU , i.e. journals whose total CPU
for all lyears is $53.00 an article or higher, and assume 20% current
year only use, I get over a thousand dollars a use. When I use that
estimate, eyes pop. Is it really possible? Yes it is.

I suspect this same logic would open up any number of big deal
publishers to consideration for on demand access as well.

Chuck Hamaker