Open access publishing - is it the future for scientific journals?

Ron Fraser

Society for General Microbiology (SGM)

This is an extended version of an article published in the February 2004 issue of Microbiology Today »

In the 'traditional' model of scientific publishing, authors submit a paper to a journal, and if it survives the rigours of peer review, it is copy-edited, typeset, printed in the next available journal issue and distributed to individual and institutional subscribers. The publisher's income from subscription sales covers the costs of promotion, review, production and distribution of the journal. This model translated into the electronic age: access to the online version of the journal is restricted to subscribers. However, there have been two main criticisms of the model, one financial and one technological.

In a study of US periodicals, subscription prices rose by an average of 9.5% per year over the past 16 years, compared with an average rate of inflation of 3.1%. Some of this has been justified by increases in page numbers, and by the additional costs of online publication. However, the budgets of university libraries have not increased by anything like 9.5% annually. This has led to the so-called serials crisis, in which subscriptions have been cancelled and publishers have put up their prices even more to compensate for the lost income: a vicious circle that reduces the overall availability of articles to the average reader.

In considering pricing, it is important to distinguish between journals produced by the not-for-profit sector - learned societies and many university presses - and the for-profit commercial publishers. Many of the latter operate with very high profit margins in their journals businesses, and much of this profit is channelled to shareholders - that is, out of the research and educational sector. In a small survey of microbiological journals, I found that those produced by commercial publishing houses cost between 3 and 5 times as much per printed page as those produced by learned society publishers such as SGM.

Of course many learned society publishers do make a profit (tastefully called a 'surplus' in the not-for-profit sector) on their journals; for many it is a major source of income alongside membership fees, investment income and meetings registration charges. In SGM's case, the journals surplus is used in support of the Society's charitable activities, such as grants to students, support of meetings, educational and public affairs work and so on. Thus the surplus is recycled to the benefit of the academic community.

The technological objection to the traditional subscription model is that it is attempting to perpetuate 19th century methods; surely the advent of the Internet offers new opportunities and demands a new approach to making the scientific literature as freely available as quickly as possible? This thought, together with a growing backlash in the academic and library communities to the rising costs of many journals, has led to the development of a number of 'open access' experiments. These range from online publication of manuscripts on individual or university websites, to open access online journals such as Public Library of Science PLoS Biology, or the BioMed Central journals on PubMed Central, which have no subscription income, although BMC is a for-profit publisher.

These open access journals have costs of production and maintenance that have to be recovered: estimates are that electronic-only high quality publication costs only 15 - 25% less than print. Generally this is approached by the 'author-pays' mechanism, in which the author (or the institution on the author's behalf) is charged a fee for publication. Different open access publishers are trying different models: some charge a flat fee per published article, others charge a submission fee for all articles, including those eventually rejected, as well as a publication fee. At the most complex, one publisher is proposing to charge both of these fees, plus extras per word, figure and table! What no one knows at present is whether these models will be economically sustainable in the long term, and acceptable to authors and readers. PLoS charges $1500 per article, but this appears to be subsidized from a $9M start-up grant from a charitable foundation. BMC charges $525 per article, but this is said to be well below the economic costs of production. Several learned society publishers have calculated the true costs of publication, including a small element of surplus: most come up with an average fee of around $2000 - 3000 per article. Oh, and first thoughts are that authors in the UK, and possibly some in the rest of the European Union, would have to pay VAT, currently at 17.5%, on any such fee.

The subscription and author-pays models are the extremes of a spectrum, and there is actually a lot of overlap in the middle.

  • Many traditional journals have had page charges for years; many charge extra for colour illustrations. These are examples of author-pays within the subscription model. SGM has traditionally been against page charges, and offers free colour where scientifically justified.
  • Many subscription journals make some back or future content freely accessible. Thus articles more than 12 months old are freely accessible online at HighWire Press for SGM's Microbiology and Journal of General Virology. Between them, the 346 journals online at HighWire have made a total of over 676,000 articles freely available. This contrasts with a total of 30 research articles on open access in PLoS Biology at the time of writing, and BMC's approximately 100 journals are thought to contain fewer than 3000.
  • Some publishers are experimenting with a hybrid model in which the basic subscription and access control systems remain, but authors can choose to pay a fee to have their article freely available online from the time of publication. Again, the fees currently seem far below the true costs of publication - is this a valid experimental design, which will yield robust conclusions about the value of this approach for migrating from a subscription to an open access model?

SGM Council and Publications Committee, and the journal Editorial Boards, will obviously be monitoring the situation as the different business models develop, and considering whether SGM should change its procedures. The strategy will need to balance the different objectives in publishing: attracting authors to submit their best work; keeping the support of scientific editors and referees; securing wide dissemination to readers; maintaining scientific and production quality standards; ensuring archival permanence and accessibility of work published online; and retaining economic viability. The subscription model has achieved all of these objectives short of completely open access; the author-pays models still have to prove themselves in several respects.

In the meantime there are many intriguing questions.

  • Will open access journals attract a significant flow of quality papers and build up respectable impact factors, or will the author payments, from those who can and want to afford the fee, relegate many of them to the level of vanity publishing?
  • Will there be a transfer of budgets from librarians to authors, and a redefinition of the role of librarians?
  • One feature of the subscription model is that authors are often unaware of the differences in costs between different journals. In author-pays models, differences in costs (and the effects of different profit margins) will be glaringly obvious. Will this lead to divergence into low cost journals with lower standards of production, and high cost journals giving full support to their editorial boards, retaining high quality copy-editing and production standards, including technically advanced online versions?
  • How will commercial publishers react?
  • If author self-publishing on personal or university websites becomes commonplace, will there be an erosion of the quality standards, and the benefits of centralization, that the established journals have built up over the years through the efforts of their editors and production staff?

Recent discussions with other publishers at HighWire and Association of Learned and Professional Society Publishers meetings have made it clear that there is no single industry view of how things will develop, but most people expect the landscape to be different or at least more varied in future, perhaps with different models running in parallel. Opinions vary on how rapid change might be, but it has already started, and many cite 5 years as the testing period for whether the author-pays model will make substantial inroads, or whether it will wither and leave the subscription model relatively intact. More reports from the Marlborough House crystal ball will follow.

Next, a word about copyright. Authors have traditionally been asked to assign the copyright in their articles to the publisher. This practice had its origins in the early days of printing in the UK, at the time of King Henry VIII, when the ability to make multiple copies of a work created the potential for widespread distribution of criticism of the king, government and church. So copyright - literally the right to print copies - began as a form of censorship, in that the freedom to operate a printing business was restricted to those who had been granted a licence to produce officially approved versions.

It is a tenet of many open-access publishing ventures that authors or their host institutions should retain the copyright in their published work, and for some protagonists, the issue seems to be approached in almost an emotional manner. The theme is 'why should the publisher, who has not contributed to the creation of the intellectual content of the article, steal the copyright from the author, who surely is the true owner?' The publisher is not paying the author for his content, is making money from the publishing process either from subscriptions or publication charges, and is then demanding the family silver into the bargain - the rights to income from future exploitation of the material.

In fact, there are several advantages to authors in transferring the copyright in their work to a publisher such as SGM. Firstly, the total income from copyright exploitation is comparatively small, and what there is comes mainly from sales of single copies of individual articles (document delivery) and licensed photocopying. The latter income is collected through reproduction rights organizations such as the Publishers Licensing Society, to which SGM belongs. This income to SGM is not always differentiated by article or even by SGM journal, so would be impossible to attribute to individual authors. It is most unlikely that authors of journal articles would be able to recover this income if they retained the copyright themselves.

Secondly, SGM does provide valuable services to authors in relation to copyright.

  • We administer 'permissions', requests to reproduce articles in whole or in part in other works. We ensure that the authors and original source are properly credited, and that the material is reproduced without alteration. As a matter of courtesy, we try to obtain authors' agreement for the reproduction, but this can be difficult when authors have moved on and cannot be traced. This in itself is a powerful argument for centralizing the permissions process.
  • We police violations of copyright. Where, for example, work published in an SGM journal is plagiarized by re-publication in another work, without permission or appropriate acknowledgement, we seek to have the offending article retracted or, in less severe cases, to have an apology published. We may also report the offence to the ethics committee or management of the offending author's institution. It is only through ownership of the copyright that SGM has the legal clout to pursue such action to protect the rights of our authors.
  • SGM places no restriction on the freedom of authors to put PDF files of their accepted articles on their own or their university website, other than that the journal publication details must be given.

About SGM

The Society is the largest microbiology society in Europe, with well over 5000 members. It publishes Journal of General Virology, Journal of Medical Microbiology, International Journal of Systematic and Evolutionary Microbiology and Microbiology.