Experimenting with Open Access publishing
In a previous Nature web debate1 we called on the scientific community, publishers and librarians to work more closely together to explore new ways of recovering the costs of publishing that would encourage widespread, cost-effective dissemination of online journals, immediately after publication. There has been considerable debate of access issues in recent months, including an inquiry by the UK government's Science and Technology Committee, to which OUP has contributed both written and oral evidence2. There remains a dearth of factual information, however, to support the arguments about the potential advantages and disadvantages of Open Access publishing. During the course of 2004 we therefore intend to run a series of experiments to test a range of business models. We hope that preliminary results presented here might encourage others to share their experiences and data.
Ten years ago the journal Nucleic Acids Research (NAR) first began publishing an annual special issue devoted to factual databases; our experiments began with a survey of 331 past contributors to this issue, in order to gauge the level of support from authors to publishing the 2004 issue on open access. In our survey we outlined two potential publication models-the traditional subscription-based model and an Open Access model partially funded by author charges. We then asked authors whether they would be willing and able to pay publication charges (at heavily subsidized levels of £300/$500 per article) to fund partially the Open Access publication of the issue. Our survey achieved a 48% response rate and, of those who responded, a two-thirds majority encouraged us to implement an Open Access model.
The NAR Database issue was published in January 2004. Out of a total of 142 papers, 128 (90%) authors agreed to pay the publication charge and the remaining 14 corresponding authors were granted a waiver. Of these 14 authors, four were from developing countries; the remainder were from developed countries but lacked the necessary funding to pay the publication charge.
We have compared the usage data for the 2003 and 2004 Database issues, both of which were published without access restrictions, with usage data for the rest of the journal and the results of this analysis are shown in Figure. 1. During the six months following initial publication, the full-text of Database issue articles were, on average, downloaded 52% more frequently than the average number of full-text downloads of other articles published in NAR. It is difficult to be sure whether this difference is solely due to the absence of access restrictions or whether the Database issue would have generated higher than average usage had access been restricted to subscribers. Once the data becomes available, it will also be interesting to see whether this apparently higher usage results in higher citations; the research by Lawrence3 is sometimes misinterpreted as evidence that Open Access leads to higher citations whereas it actually compares offline and online publication.
In view of the positive response from authors and the apparent increased usage of articles published in the Database issue we now plan to implement the same model for the 2004 NAR Web Server issue, which publishes reports of servers that perform useful computations on DNA, RNA and protein sequences and structures; this follows a further survey of contributors to the 2003 Web Server issue which resulted in a small majority (54%) expressing preference for author-funded Open Access publication.
In both surveys, some researchers expressed concern about the ability of some authors to pay the full Open Access publication charges. We believe that authors should not be prevented from publishing in the journal of their choice, so any author-funded Open Access model would need to be supported by a waiver system thereby increasing the costs for well-funded authors.
Feedback has also highlighted the essential requirement that author charging and editorial decisions be completely separated and seen to be separated. Authors should be reassured that their ability to pay publication charges will not influence editorial decisions in any way. Once manuscripts have been accepted for publication following the usual editorial review process, in all of our experiments the author charging system is therefore administered directly by OUP rather than by the independent academic editors of each journal.
It also became clear that there is still some uncertainty within the academic community about what Open Access actually means-some believe that an unrestricted right to re-use data published in Open Access articles is just as important as unrestricted access.
During 2004, two other journals published by OUP will be experimenting with Open Access. From its July 2004 issue, the Journal of Experimental Botany will give authors of ordinary research papers the option to pay £250 to allow all readers online access to their paper without charge. Review papers and Special Issues, which are particularly highly cited, and papers of those who choose not to pay the £250, will continue to be available only to subscribers in the usual way.
The intention of this experiment is to allow us to gauge the extent of interest in Open Access publishing from the plant science community. The Journal has also received a grant from the Joint Information Systems Committee of the UK4. In return, all papers with a UK author published in the first year of the trial will be made freely available online without an author charge.
The first issue of a brand new Open Access publication entitled eCAM (Evidence-based Complementary and Alternative Medicine) will appear in June 2004. This interdisciplinary journal naturally bridges the Western and Asian worlds and should, therefore, attract a high proportion of authors and readers from developing countries. An 'author-pays' model might prove to be too expensive to attract such authors, and subscription barriers might limit beneficial widespread access.
We are fortunate to have secured sponsorship to cover the online publication costs of eCAM; this allows us to provide all users with free online access to the research material. We will also be launching an 'enhanced' online version of eCAM containing reviews, commentaries and other information for which there will be a subscription charge. In this way we hope to develop a revenue stream that will allow us to continue to improve and extend eCAM in coming years.
In addition to our experiments with Open Access publishing, we are also participating in the SHERPA project, which is investigating the potential for "Institutional Repositories", a somewhat controversial publication system that has yet to become widely established5. Our collaboration with Oxford University Library services as part of this project is designed to investigate how online journals and other material published in Institutional Repositories might be integrated.
Further information on these Open Access experiments:
Nucleic Acids Research
Journal of Experimental Botany
Evidence-based Complementary and Alternative Medicine
OUP participation in the SHERPA project
Martin Richardson and Claire Saxby
Oxford University Press
- Richardson, M. Impacts of free access. Nature (Web Debates) (2001)
- Memorandum from OUP to the Science & Technology Committee Inquiry into Scientific Publications (2004)
- Lawrence, S. Free online availability substantially increases a paper's impact. Nature (Web Debates) (2001)
- Press Release: JISC and Publishers Work Together to Open Up Access to Journals (2004)
- Ware, M. Universities' own electronic repositories yet to impact on Open Access. Nature (Web focus. Access to the literature: the debate continues) (2004)