Newsletter on Serials Pricing Issues 2000-253

     ISSN: 1046-3410


                NEWSLETTER ON SERIALS PRICING ISSUES


                    NO 253 -- September 22, 2000
                        Editor: Marcia Tuttle
                      Guest Editor: James Mouw



                             CONTENTS


        253.1 AGREEMENT: UKB (DUTCH ACADEMIC LIBRARIES) AND
              ELSEVIER SCIENCE, Azim Koning
        253.2 PRICING OF CHEMISTRY JOURNALS, Dana Roth
        253.3 GORDON & BREACH SUIT FAILS AGAIN, Rory McGee
        253.4 TACKLING THE SERIALS CRISIS AND UNBUNDLING SUBSCRIPTIONS,
              A.B.A.Schippers

253.1 AGREEMENT: UKB (DUTCH ACADEMIC LIBRARIES) AND ELSEVIER
SCIENCE 27 JUNE 2000 Azim Koning, Beleidsmedewerker Stafbureau Directie Bibliotheek van de Universiteit van Amsterdam, koning@uba.uva.nl

[Received June 27, 2000]

UKB (the consortium of all university libraries in the Netherlands, the Koninklijke Bibliotheek/Royal Library and the Netherlands Institute for Scientific Information Services) and Elsevier Science, have reached an agreement about the price movement of journals for the next five years and about the accessibility of electronic versions of Elsevier journals.

Last year, the academic libraries in the Netherlands associated in UKB took up a position in the debate on developments referring to scientific information. One of its important elements focused on the price increases of scientific journals, another element on the gradual though rapid transition into digital information.

Since August 1999, UKB has arranged a series of consultations with some large publishers on the subjects mentioned above. Meanwhile, UKB came to an agreement with the world's largest publisher of scientific information, Elsevier Science, on all items discussed. The agreement will last for five years (2000-2004). Arrangements have been made on the price development for the forthcoming years, on the availability of the digital versions of journals published by Elsevier Science. Furthermore, the libraries and Elsevier agreed to do further research (in co-operation with other interested parties, if any) on transitions within the information chain as a result of the availability of ICT in the academic world. These transitions will probably induce more explicitly basal 'business models' on scientific information supply.

Consultations with other publishers will continue. As renewals of next year's journal subscriptions have to be made before September 1st, 2000, UKB has asked the most important publishers to provide in good time information citing the 2001 subscription prices. UKB likes the publishers to motivate important price increases, particularly in regard to price developments in the last decade.

The Dutch scientific libraries, associated in UKB, will evaluate the publishers' offers regarding reasonableness and fairness. They will then jointly respond to these offers. UKB expects to improve in this manner the availability of scientific information.

Contact person for UKB is N. Verhagen, chief librarian of the University of Amsterdam. Tel. (++ 31 20) 525.2307

For more information, i.e. the joint UKB-Elsevier Science press release:

http://www.uba.uva.nl/en/projects/journals-pricing-ukb/pressrelease-ukb-elsevier.html

or:

http://www.elsevier.nl/inca/esav/shownews.htt?item=WN/nt0000171

or via: http://www.uba.uva.nl/en/news

253.2 PRICING OF CHEMISTRY JOURNALS
Dana Roth, CalTech, dzrlib@library.caltech.edu

[Received July 18, 2000]

At the recent SLA Annual Conference, several attendees questioned the relatively high cost of Organic Letters, in comparison with other ACS titles. Subsequently, an item appeared in the Newsletter on Serials Pricing Issues -- 249.5 PRICING OF CHEMISTRY JOURNALS -- A COMPARISON OF JOURNALS PUBLISHED BY NOT-FOR-PROFIT ORGANIZATIONS AND COMMERCIAL PUBLISHERS. This article was authored by Kristina Kurz, Managing Editor of Synthesis and SynLett, which are commercially published by Georg Thieme Verlag. She compared thirteen chemistry journals on the basis of price per page for institutional subscriptions (the complete data and graphic representation of the results are available as an Excel file from kkurz@thieme.com). The journals were divided into two groups. Group One consists of titles published by commercial publishers. These journals cost between $US 0.22 and $US 0.91 per page. Group Two covers journals from not-for-profit publishers and, not surprisingly, the average costs are lower, ranging between $US 0.15 and $US 0.54 per page.

What is really surprising is that two commercially published journals (Synthesis $US 0.31 and SynLett $US 0.22) are much less expensive than a corresponding ACS publication (Organic Letters $US 0.54). Furthermore, Organic Letters is also anomalous in that, while its subscription cost is about 1/4 the cost of Tetrahedron Letters, its per page cost is only about 1/2 and it is 3.6 times as expensive (on a per page basis) than the Journal of Organic Chemistry.

This anomaly may be related to the fact that Organic Letters is obviously a start-up journal and the ACS may not be comfortable amortizing the substantial costs involved in producing a first class journal with electronic features and functionality over an extended period of time. Another factor may be an attempt to avoid repeated sharp annual subscription price increases, if the journal grows as one would expect, for an excellent ACS title. It is my hope that, by quickly amortizing the start up costs, the ACS will be able to manage the inevitable increase in page count with very limited subscription price increases.

The significant conclusion from the Thieme study, however, is that it is possible to publish excellent scientific journals (Synthesis $US 0.31 and SynLett $US 0.22 have been among the top ten organic chemistry journals in the ISI impact factor ranking) without charging outrageous prices (e.g. Tetrahedron Letters $US0.91).

Furthermore, I strongly feel that extreme care must be taken by librarians to insure that reasonably priced, high quality journals, such as Synthesis, SynLett, Journal of Organic Chemistry and Organic Letters, are made available to researchers before consideration of higher priced commercially published journals of lower quality.

Addendum:

1996- 97- 98- 99 1996- 97- 98- 99

Title ISI Impact Factor # of articles --------------------------------------------------------------------------- SynLett 2.8 2.65 2.75 2.55 430 482 478 577 Tet. Lett. 2.5 2.5 2.6 2.4 2506 2428 2648 2399 -----------

Title 1996- 97- 98- 99 (1999 print-web)Institutional price --------------------------------------------------------------------------- SynLett $450 $478 $480 $575 ($ 748)

Tet. Lett. $6003 $6845 $7936 $8602 ($9892) -----------

Tetrahedron Letters/SynLett     1996   1997   1998   1999    1999(print-web)
article ratio (TL/Syn)           5.8    5.0    5.5    4.2    (4.2)
price ratio (TL/Syn)            13.3   14.3   16.5   15.0   (13.2)

This is an example of a 'picture that doesn't look right.' Tetrahedron Letters has consistently lower ISI Impact Factors, is trending toward a decrease in the relative number of articles (which are roughly related to the number of published pages) but charges over 13 times the price of SynLett, while averaging only about 5 times as many articles. Does this suggest that Tetrahedron Letters may be overpriced by at least a factor of 2?

253.3 GORDON & BREACH SUIT FAILS AGAIN, Press Release
Rory McGee, American Institute of Physics, rmcgee@aip.org

[Received July 19, 2000]

The following statement is issued jointly by the American Institute of Physics (AIP) and the American Physical Society (APS). For more information please contact AIP Director Marc Brodsky at 301 209- 3131 or APS Treasurer Thomas McIlrath at 301 209-3220.

July 17, 2000, American Center for Physics, College Park, Maryland.- We are pleased that the French Court of Appeals in Paris has rejected a suit filed against us by Gordon & Breach and joined courts in Germany, Switzerland and the United States in upholding our right to publish objective data on the prices of physics journals. The French Court found "no element of the [case] file leads one to doubt the trustworthiness of the results published by the" APS, AIP and Henry H. Barschall. This decision, while still appealable to the French Supreme Court, culminates 13 years of litigation in France on the issue of whether survey data prepared by Barschall and published in Physics Today (December 1986 and July 1988) and the Bulletin of the American Physical Society (July 1988) constituted unfair competition by illegal comparative advertising. Barschall's analysis was based on data on hundreds of journals that showed the cost per character, the impact (as measured by existing citation data), and the cost per impact. Gordon and Breach filed suits in four countries alleging various false and misleading statements as well as illegal advertising and disparagement.

The French Court of Appeals pointed out that "the simple publication of given information displaying the high price of the publications of the Gordon and Breach group, in the absence of any acrimonious or pejorative calculation, cannot be held to constitute disparagement."

Unfortunately, Professor Barschall did not live to see the final decisions in the U.S. and Switzerland, nor this favorable Court of Appeals decision in France in which his estate was a co-defendant. It is gratifying, however, that the French Court took note of the June 12, 1997 testimony of a Gordon and Breach witness in the U.S. Court that "There was absolutely no doubt about the fact that Professor Barschall produced a good, complete work, with good intentions" and that Barschall's results were trustworthy.

The U.S. Court likewise found that "Barschall's methodology has been demonstrated to establish reliably precisely the proposition for which defendants cited it -- that defendants' physics journals, as measured by cost per character and by cost per character divided by impact factor, are substantially more cost-effective than those published by plaintiffs." The U.S. Court also noted evidence that Gordon & Breach had engaged in "an aggressive corporate practice... to suppress all adverse comment upon its journals."

253.4 TACKLING THE SERIALS CRISIS: UNBUNDLING JOURNAL SUBSCRIPTIONS
A.B.A.Schippers, Utrecht University, a.b.a.schippers@phys.uu.nl

[Received September 8, 2000]

At the end of August 1999, the Physics and Astronomy Library of Utrecht University cancelled $ 150,000 worth of its subscriptions -- a substantial part of its $ 350,000 total expenditure on periodicals. This decision was not taken 'out of the blue' but was the logical outcome of a series of attempts that the library had made to cut its costs over the past twenty years or so. These attempts included the following projects:

Various faculty libraries were incorporated into eight departmental libraries in order to reduce the need for duplicate subscriptions.
With a view to streamlining the journal collection, we asked faculty members on several occasions to identify those journals that were really essential for their research group. As well as establishing these interest profiles, we performed a citation analysis.
A financial system was established whereby 150 subscriptions to core journals, which at one time accounted for 70% of the journal budget, are paid by a central departmental budget and the remaining subscriptions to 500 or so journals, accounting for the remaining 30%, are paid by the thirteen research groups. Such a dual financing system takes into account the responsibilities of both research staff and library. In this way the well-known skewed distribution (Lotka's law) of subscription prices, citations and library circulation for science journals (S.J.Bensman LRTS 1996) is easier to handle.
Prices continued to spiral and, as a result, in 1994 it became necessary to cut expenditure on core journals still further. On the basis of their existing interest profiles the thirteen research groups were asked to give priority ratings to the journals on their list. This time these priorities were subject to the following condition: the total cost of journals that were given a high priority rating was not to exceed 40% of the total cost of their journal interest profile. If a research group was willing to pay 50% of the subscription that the library proposed to cancel, the journal was retained. As a result of this measure the number of core journals was reduced to 122 and subscriptions amounting to a total of $62,500 were cancelled.

So that some of the above data can be used to draw comparisons with the situation in other libraries, we give some additional information about the Physics and Astronomy Department at Utrecht University. It is one of the larger departments of its kind in the Netherlands. It has 164 researchers (faculty), 50 technicians, 120 PhD students, 100 guests and 500 students. In the year 2000 the library still subscribes to about 700 scientific journals, of which approximately 100 are also available electronically.

Between 1994 and 1998 the cost of core subscriptions rose from $178,500 to $275,000. It would have been out of the question to counter this inflationary price rise by applying the approach adopted in 1994. A new technique was needed, particularly since the readership of a substantial part of the collection is so widely scattered. Instead of opting for the cost-per-use method (Milne & Tiffany), we decided to adapt the 'voting' method. The method requires each user to distribute 100 points -- to be compared with the 100+ readings per scientist reported in the surveys by King Research -- over the subscriptions to the 122 core journals. An essential adjustment made was the unbundling of journal packages and sections. A survey based on this adapted 'voting' method and the existing interest profiles was conducted in the summer of 1998. There was a high response to the investigation: 73% of the researchers (staff and PhD students, 218 persons in all) completed our questionnaires. Since we know from the replies given in the questionnaire that the median respondent read 50-74 articles in 1997, the voting method turned out to be a reasonable way of assessing the cost-effectiveness of the core subscriptions. After we had corrected and checked the results of the questionnaires we calculated the price per point assigned to each of the journals -- i.e. 'the price per article read.' In view of the number of subscriptions that had already been cancelled, the cost- effectiveness of the journal collection was far less than we had expected: 85% of the 'reading' could be covered by 40% of the cost of the core journals.

Since the cost of document retrieval, which now ranges from $7 to $10, might rise to as much as $25, we proposed to cancel all subscriptions to core journals where 'the price per article read' was higher than $25. In this way 49 subscriptions -- costing $125,000 -- were earmarked for cancellation. Of these journals 70% are produced by commercial publishers. After we had consulted the various research groups the faculty board decided to cancel $125,000 worth of core subscriptions.

The above cancellations are specific to Utrecht's Physics and Astronomy department. A rough comparison of our list of cancelled journal titles with the journal collections of other Dutch physics libraries suggests that at least some of their journals are not cost-effective either.

It is common knowledge that 'citing' or 'faculty's opinion' as an indicator of use does not coincide with the 'reading' criterion. This project leads one to suspect that the overlap between the two groups of journals which can be associated with 'citing' and 'reading' is diminishing.

Lessons learned. These 'pruning activities' were successful because of the close cooperation between researchers and the library. To obtain this kind of cooperation a library needs to understand the 'perverse incentives for authors and editors in scholarly publishing' - to use Odlyzko's phraseology.

We believe that the unbundling of journal packages and sections will lead to a more selective subscription policy. This notion is supported by the results obtained recently in the University of Michigan's PEAK-project.

Of course one gets tired of cancelling subscriptions, but our experience has shown that the amount of information to be delivered by the library is constantly changing and it is important to remain alert to developments in scientific communication.

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Statements of fact and opinion appearing in the Newsletter on Serials Pricing Issues are made on the responsibility of the authors alone, and do not imply the endorsement of the editor, the editorial board, or the University of North Carolina at Chapel Hill.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Readers of the Newsletter on Serials Pricing Issues are encouraged to share the information in the newsletter by electronic or paper methods. We would appreciate credit if you quote from the newsletter.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
The Newsletter on Serials Pricing Issues (ISSN: 1046-3410) is published by the editor through Academic Technology and Networks at the University of North Carolina at Chapel Hill, as news is available. Editor: Marcia Tuttle, Email: marcia_tuttle@unc.edu; Telephone: 919 929-3513. Editorial Board: Keith Courtney (Taylor and Francis), Fred Friend (University College London), Birdie MacLennan (University of Vermont), Michael Markwith (Swets Subscription Services), James Mouw (University of Chicago), Heather Steele (Blackwell's Periodicals Division), David Stern (Yale University), and Scott Wicks (Cornell University).

To subscribe to the newsletter send a message to JOIN- prices@listserv.unc.edu Be sure to send that message to the listserver and not to Prices. To unsubscribe, send a message saying "unsubscribe prices" to lyris@listserv.unc.edu You must send the message from the e-mail address by which you are subscribed. If you have problems, please contact the editor.

Back issues of the Newsletter are archived on two World Wide Web sites. At UNC-Chapel Hill the url is: http://www.lib.unc.edu/prices/ At Grenoble the url is: http://www-mathdoc.ujf-grenoble.fr/NSPI/NSPI.html
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++