For administrators in gleeful expectation of the "library windfall," I note that the percent of the total US research university library budget devoted to serials costs in 2002 was only 26%. http://www.arl.org/stats/arlstat/graphs/2002/2002t4.html This covers print journals, electronic journals, databases, newspapers, etc. ; it includes all fields of study. If 3/4 of it were science journals, that comes to less than 20% of the total library expenditure.
> And yet the library serials budget *is* relevant, for, if open-access > should prevail, it is the library that will enjoy the annual windfall > savings on its erstwhile serials toll expenditures .... > For by then the institutional > library windfall savings will be more than enough to pay the peer-review > costs for all institutional research output several times over. If half the science research journals were converted immediately, which is extremely optimistic, it would provide a potential source of $1.8 million for the typical university. If half the 2000 or so annual papers from a university were so published, that's $1,800 per paper for all costs. The total costs will inevitably equal the money spent--the only way to make the system more affordable is to reduce the costs, not merely redistribute them. The reason why librarians must be concerned about this, is the unfortunate probability that the money saved from library acquisitions will not be used to finance the publication system, but for general university administration. This is not an improvement over the present, where a considerable part of the expense is used for general administrative purposes by the publishers. This explains why many libraries are willing to pay subscriptions to alternative publishers: the basic rule of library budgeting is that if you do not spend all the money, you will lose it forever. -- Dr. David Goodman Princeton University Library and Palmer School of Library and Information Science, Long Island University e-mail: dgood...@princeton.edu