Proceedings of ICSU Press Workshop

on

Economics, real costs and benefits of electronic publishing in science - a technical study.

Keble College, University of Oxford UK
31 March to 2 April 1998

An

Executive Summary Report

Dennis Shaw & Roger Elliott

ICSU Press held a workshop in Oxford (UK) at the end of March 1998 to follow-up one of the main recommendations of the Joint ICSU Press/UNESCO Conference of Experts (February 1996 - ref1) - to conduct a technical study on the economics, real costs and benefits of electronic publishing in science.

Invitations were sent to participants in the 1996 Conference and others representing the scientific library, publishing and information sectors, and fifty-five accepted and registered for the event held at Keble College over three days (31 March to 2 April).

The first day was devoted to the presentation of eleven invited papers on the benefits and real costs of e-publishing (the term now used to define digital publications accessible online). The necessity was accepted that the benefits should first be established before attempting to evaluate the costs. The benefits identified by Anant Parekh (a young research physiologist) were speed of publication and access, rapid retrieval of related papers, accessibility from a wide variety of locations with minimum time constraints compared with print-on-paper research journals, and the facility for referees to conduct a more comprehensive review as a result of wider access to data. This reflects the opinion of the majority of younger research scientists.

Fytton Rowland of Loughborough University summarised the advantages (and disadvantages) to the librarian resulting from the growing dependence on the internet resources. Some disadvantages of presently available systems were discussed, such as the non-availability of a universal browser and the many different software packages in use which are not all mutually compatible. Copyright issues and intellectual property rights of authors are in conflict with the demands of some with-profit publishers and European and North American legislators are not always sympathetic to the scientist.

Bernard Donovan gave a detailed account of the costs of peer review and indicated how these might be affected by electronic publishing. The benefits of giving referees access to data, and links to related publications, as well as the original authored contribution reporting the results of a research project, has been emphasised by many scientists (e.g. see Hall and Heck, ref2). "Peer review is a critical component in the competition between rival journals. . .since good refereeing and editing raises the perceived quality and increases reader appeal" As a consequence a good journal will spend more on the process of refereeing, and this indicates the significance of peer review in costing journal production. Electronic methods should eliminate postage, some administrative and stationery costs and possibly mailing delays, all of which could be perceived as benefits. However, the lack of standardisation of software (often complex), and the variety of electronic formats now available and in use, counteract these advantages. Donovan's paper included some useful data, not previously, available on the present and future likely costs of peer review.

Fred Spilhaus of the American Geophysical Union reported on the experience of learned society publishers in providing online editions of their publications and it was generally agreed that the additional costs of online access was in the range of 20-30%. This estimate did not include the additional cost to the user of training techniques and access provision, but did include a major component, namely, the cost of providing access to and maintaining a digital archive. William Mischo of the University of Illinois reported on the archiving costs evaluated by his research team in the digital library experiment. The rapid growth in the number of conventional print format journals adding electronic editions, and the growing numbers of electronic-only journals, have intensified the concerns about the problems and costs of digital archiving. The term archiving denotes not only the storage of materials but the systematic organisation and exhaustive provision of access to these materials. In the case of electronic publications one of the major problems to be addressed in access-provision has been the wide variety of formats in use. This was illustrated by the statement " I can read a printed book published 300 years ago but it is impossible for me now to read a Microsoft Word II document written in 1988".

The estimates for adding online access were challenged by two well known contributors to the debate which has been ongoing for at least the last five years. Donald King described an economic cost model for scholarly scientific publishing which he had developed with Carol Tenopir (University of Tennessee) incorporating five publishing components: manuscript processing (editing and compositing), non-article processing (covers, contents tables, letters and reviews), reproduction, distribution and support. The model contains cost parameters and cost elements for the individual items and the parameters were estimated from a sample of scientific journals. Cost elements were derived from published data. The model was applied to estimate the add-on costs of electronic publishing as an additional feature for a print-on-paper journal and also to predict the savings resulting from replacing a traditional journal with a fully electronic edition. The savings predicted by this model were strongly dependent on the number of subscribers, ranging from 8% for 500 subscriptions to 40% for 5000. Although the results were based on a small sample of journals, the model provides a useful guide to the complex range of processes needing to be considered in any attempt to evaluate real costs; this could account for the wide range of estimates given by different publishers. In a complementary paper, John Scott discussed the 'real costs' of electronic journals drawing on his experience as a former editor for the American Institute of Physics (emphasing that the opinions expressed were solely his own). He warned of the perils of oversimplification due to making unfounded assumptions, such as the achievement of major cost-savings at the same time as providing additional features (e.g. means for digital storage and more sophisticated distribution systems). He asserted that the situation has been in a state of flux for many decades and change will continue indefinitely into the future. Electronic publishing is not a well defined term and there is much confusion resulting from cost comparisons of dissimilar products. The cost of subscription to a journal is different for the purchaser and supplier. A definition of a scholarly journal includes the following features: peer review, archiving, establishment of priority, citable and retrievable content, and universal accessibility world-wide. This definition was applied to obtain a listing of the added costs of electronic publishing through value-added features such as: hypertext links, full-text searching and digital archiving, all of which are expensive to maintain with incessant technological change. Other recognised factors are the higher salaries commanded by the more highly-skilled personnel employed and the increasing sophistication of the systems used. These points were illustrated by a detailed presentation for a fictitious journal with defined parameters, which led to derivations of annual income and expenditure statements. It was concluded that the addition of an electronic version to an established print-on-paper journal increases costs. An alternative electronic journal could be produced for about the same cost as a printed version but there is as yet insufficient experience to show this result as established.

Bob O'Shea of Lindsay Ross International presented a report on a study carried out with Owen Hanson of the City University (London) on the hidden costs of e-publishing. The results supported many of the challenges presented by other publishers. There are significant costs and problems to be handled, if the benefits that are readily apparent on first consideration of electronic publishing are to be achieved at an acceptable cost, and without loss of the rigorous academic scrutiny that traditional scientific publishing receives. At the end of the first day David Pullinger (Nature-Macmillan electronic publisher) summarised the presentations and added some observations drawn from his own experience.

The second day was mainly devoted to discussion in working groups of the issues raised on the first day but there was time for six contributed papers Gary VandenBos and Susan Knapp of the American Psychological Association reported on the economic aspects of mounting a full-text database online and selling electronic subscriptions. Knud Thomsen of the European JET project presented the origins, structure and management of a very large international database on plasma confinement data as an example of the way electronic publishing has greatly altered the exchange of information between scientists engaged in a major international research project, and Vitaly Nechitailenko, a Russian Geophysicist, gave an account of the international collaboration of geophysicists in their electronic publishing. Atilio Bustos from Valparaiso and Abel Packer from Sao Paolo reported on their experiences from Latin America and indicated the advantages to scientists in some less well endowed research environments of being able to mount the results of their work on the Internet. Virginia Cano (Queen Margaret College, Edinburgh) presented an outline study indicating the ways in which a market may be created for e-publishing in developing countries.

The third day started with group reports by the chairmen and general discussion. This led to the formulation of conclusions and recommendations on which there was general agreement and the conclusions are summarised below.

A dedicated website was established in advance for posting abstracts and comments were sought from those invited who were unable to attend. This facility was welcomed by many participants and assisted the organisers in distributing the conference material. The members of the programme committee were Ian Butterworth (representing the Academia Europaea), Roger Elliott (Chairman of ICSU Press), Glyn Jones (Director of Portland Press), David Price (Head of Systems at the Bodleian Library, Oxford), John Rodda (President of IAHS), Dennis Shaw (Manager of the ICSU Press website) and Anthony Watkinson (Consultant and former Intellectual Property Director of Thomson Science)

The proceedings of this Workshop have been published electronically on the ICSU Press website - (URL - http://www.bodley.ox.ac.uk/icsu) and are freely available. This Executive Summary is produced in hard-copy for distribution to members of the ICSU family.

Summary of the main conclusions and recommendations from the workshop

1. It is neither practical nor proper for ICSU to attempt to force the pace of development of electronic journals. The current fluid situation should be allowed to evolve at its own pace.
   
   
2. The risk of breakdown in the current system of financing the publication of the results of scientific research will not be removed naturally by the development of electronic publishing, in fact, the risk may well be increased by the current need to finance both online and traditional print-on-paper versions.
   
   
3. Technological progress in electronic journals and the different pricing models as presently envisaged will not per se reduce prices nor allow the potential for increased accessibility to be realized. However, the possible simultaneous access by many readers, for a single online subscription, is already a reality in many institutions. An electronic journal would only become cheaper if users clearly identified costly features that they did not require.
   
   
4. There was complete agreement that peer review is absolutely essential for electronic publications.
   
   
5. The most significant advantages of the electronic journal are its speed of publication, ease of access, and the ability to search and link across many other journals. ICSU should consider means of informing scientists regarding all the important elements in the publication process so that they were better able to assess the value of each element and the way in which it affected their own contribution.
   
   
6. It is recommended that experiments on means of financing electronic publications by page charges and by access fees are actively pursued.
   
   
7. It is recommended that ICSU work towards persuading those in authority to accept electronic journals on a par with printed publications as indicators of research contributions.
   
   
8. It is recommended that small learned societies should band together, or operate their electronic publishing programmes under the umbrella of bigger societies or associations in order to better serve their communities. (note: some are already doing so).
   
   
9. It is recommended that ICSU seeks to persuade funding bodies to ensure there is an allocation in all research grants for the acquisition of information services, and for the publication of results.
   
   
10. It is strongly recommended that ICSU continue to put pressure on Governments everywhere to ensure that academic networks remain centrally funded and that they are, indeed, improved. (A specific issue in Europe would be the continued funding of the TEN-34 and its successor by the European Commission).
    
    
11. It is recommended that ICSU approves and recommends a system for ensuring and maintaining the integrity and authenticity of electronic publications as well as devising means for promoting uniform standards.
    
    
12. ICSU should alert national academies to address the problems regarding archiving of scientific research papers and persuade them that government funding is essential for creating and maintaining an archive of electronic publications.
    
    
13. Parallel publication of 'print + electronic' is more expensive than 'print only'. Estimated amounts for the supplier varied but were generally felt to be in the region of 20/30%. This does not include the extra cost to the user.
    
    
14. A full 'electronic only' journal with added features, e.g. with links to databases, related journals, etc. was likely to be at least as expensive as print.
    
    
15. Although there was no agreement on a study of costs, it was agreed that it would be useful if those who were willing to disclose or present cost data were to adopt a uniform approach. In this context it might be useful if ICSU were to produce a model which would include all the cost factors which should be taken into account and how these were ultimately expressed e.g. cost per page, per copy, per article, per read, etc.
    
    
16. Although technological advance may be the ultimate determinant market forces, both financial and social, will continue to be the main factors determining the evolution of electronic publishing for the foreseeable future.
    
    
17. Journal editors should endeavour to maintain quality of content through peer review and the outcome will result from natural selection with relevance, quality and price being the major factors.
    
    
18. ICSU should facilitate, wherever possible, the visibility of electronic journal sites in developing countries e.g. pointing to websites from the ICSU server.

References

1 Dennis Shaw and Howard Moore (eds) Electronic Publishing in Science : Proceedings of the Joint ICSU Press/UNESCO Expert Conference- February 1996, UNESCO, Paris 1996.

ISBN: 0-930357-37-X; URL - http://associnst.ox.ac.uk/~icsuinfo/confproc.htm.

2 Sydney R Hall, A scientist's view of the issues and challenges. Ref1, pp89-94, and Andre Heck, From an early electronic-publishing concept towards advanced electronic information handling. Ref1, pp95-101.