In response to Kerry Kresse's question from U. of Wisconsin regarding
serial duplication on a campus, I offer the following:

Historically, duplication of serials on a university campus came to be
because branch libraries were developing "core" collections of
periodicals.  By and large, these titles WERE heavily used and,
therefore, duplication was often justified.  But, as time has gone on
and as knowledge has begun to be dispensed by newer journals, the "core"
literature is not necessarily the most heavily used.  Titles like
_Journal of applied physics_ have continued to rank high in terms of
current use.  Others, including society publications as well as
scholarly commercial journals, have been shown to no longer be
cost-effective.

I, therefore, believe that current use patterns should be the
determinants of duplication.  In addition, the availability of
duplication/transmission technology (like ARIEL) at the various branch
sites must be considered in a working policy on serial duplication.  In
a recently completed study of journal use at the four SUNY center
campuses (Albany, Binghamton, Buffalo, and Stony Brook), I found that
there were many "classic" titles which were duplicated on a given campus
for which current use at a certain branch did NOT justify ongoing
duplication.  I am sure that other readers of this listserv would also
concur with this observation.  The hidden agenda here is, of course,
that you need to gather current journal use data in order to determine
which titles fit this description at your particular campus.  Since the
faculty still believe they need these titles, you need real data to back
up your decisions.  Cost-per-use data is convincing, especially if you
turn some of the savings from cancelled duplicate subscriptions into
subscriptions to newer unique titles which ILL or document delivery
requests have shown to be more critical to current research interests.

Kate Herzog
University at Buffalo
Science & Engineering Library