James Sweetland, Library Link Regional Convenor - USA

A recent conference, "Scholarly Communication and Technology" held at Emory University, Atlanta, Georgia, USA, on April 24-25, 1997, provided some very interesting and thought provoking information on this subject.

The first session involved discussion of the economics of electronic publishing, primarily from the publishers' point of view. Other sessions included the user and librarian point of view as well.

In essence, the results of experiences of MIT (Massachusetts Institute of Technology) University Press, the American Economic Association, Johns Hopkins University Press and the University of Chicago Press are similar. To date, the costs of producing an electronic journal are much higher than those for producing a printed journal. These total costs are higher even after factoring in the savings in purchase of paper and ink, and the printing and postage costs for distributing the paper journal.

Some data will be of use here:

For example, the Chicago Journal of Theoretical Computer Science, a wholly electronic publication, issued since June of 1995, costs over $47,000 in production and overhead costs vs. about $16,000 for a similarly-sized print journal also produced by the same press.

The University of California has found that there are no savings to date from the original development costs for their electronic journals, due to changes in technology. Since they began, with a gopher-based system, they have had to develop two completely different Web-based systems. Thus, the total development costs have not be amortized at all, leading to significantly higher costs for the electronic than the printed products. Perhaps of more long-term importance, U. C. calculates that the "first copy" costs of a journal, regardless of format, about over 82% of the total costs.

First copy costs, of course, are all of the overhead, editorial and related costs to produce a given issue of the journal; the rest of the costs are the result of printing, posting, and the like related to the multiple copies actually produced.

Assuming that these results are typical, there are thus little savings to the publisher in producing an electronic journal, the savings in printing, paper and postage more than being absorbed by the additional costs of creating the electronic publication.

To date, then, there are no cost savings in publication of an electronic serial. Of course, a number of the participants at the conference noted that as technology both improves and becomes more stable, it is at least possible that costs will drop, perhaps even below those of a printed journal. Interestingly, all the electronic publishers currently charge substantially more to libraries for a given journal subscription than to individuals, both for the print and the electronic journals.

Several suggestions for dealing with this issue while waiting for the technology to solve the problems were made; unfortunately, many of these were not the sort to improve acquisitions librarians' attitudes. A representative of the American Economic Association suggested that the true difficulty in selling a journal directly to the user (or to the library) was in the use of an agent to absorb the extra costs. The suggestion was that each article would cost about $14.00. There was apparently little discussion of how many users would be willing to pay that amount for each article. In other discussions, the problem of the library's intrinsic reduction of potential subscriptions was noted, but with little that has not already been said on the subject.

In short, libraries will only want to subscribe to a given journal (regardless of format) if they can assume several users of the product. Thus, publishers feel that differential pricing, or some form of site license is appropriate for the library. However, of course, the libraries are reluctant to spend additional money on electronic publications, especially when

(1) these seem to cost as much or more as the printed ones, and
(2) the libraries lack the funds to increase their journal costs in any event.

Karen Hunter's analysis of the library perspective may be the most telling of all the papers as far as the librarian/publisher relationship goes: she suggests that a perceived "fair price" is not, in itself, sufficient reason for any subscriber to use the electronic format-the new format must have additional value, such as hot links to other sites, a job lead system etc.; as well as the ease of use, high quality content, good graphics etc. of the printed journal.

One suggested solution for reducing production costs to deal with the problems noted above may well be the wave of the future, but may also be a strong deterrent to authors' submissions to electronic journals-the suggestion was that, just as authors are now expected to provide machine-readable manuscripts to publishers, they may be expected to provide fully marked up SGML text in the near future, in order to reduce electronic journal first copy costs. At least one presenter at this conference noted the reluctance of authors even now to submit papers to such journals.

Ironically, as the publishing industry confronts electronic publication, it appears to be increasing the burden upon authors who wish to be published. This begs the question (at least to this author) of how much work an uncompensated journal article is worth-or, to be blunt, at what point should the author avoid the publisher entirely, and merely mount the text him or her-self?

The full set of papers from the conference are available as-

"Scholarly Communication and Technology", Papers from the Conference organized by the Andrew W. Mellon Foundation and Held at Emory University (Atlanta, Georgia, USA), April 24-25, 1997. 468 pp. ERIC Document #ED414916.

And are also on the Web at http://www.arl.org/scomm/scat/index.html

Back to Electronic Resources Viewpoints