Stevan Harnad

I was expecting to be disappointed by  Swan & Brown's (2004)  report because it had been commissioned by JISC/OSI as a study primarily on Open Access Publishing instead of on Open Access Provision.

What an agreeable surprise, then, that not only did the authors manage to gather some new and valuable evidence despite the narrow confines of their mandate, but they managed to make useful sense of it too, following through its implications in a way that shows a rare grasp of what is really going on in the Open Access (OA) world today -- and what is still needed.

The cynics will say my admiration is merely because I agree with their conclusions (and I do!), but I could not have invented their data! And although most of their questions were obligatorily focussed on how authors liked publishing in OA journals and whether they would like to do it again (yes, they would, and that's peachy, but it's not news!), some of the questions they included in the survey generated extremely useful information on the other road to OA:

Based in part on methodological details and data breakdowns kindly provided for me by Alma Swan that were not explicitly included in the published JISC/OSI report, it turns out that whereas (i) about 2% of the conventional-journal author sample (3/140) had made at least one article OA by publishing it in an OA journal (20/160 weren't sure!) and (ii) about 39% (62/160) had made at least one article OA by publishing it in a conventional journal and also self-archiving it, (iii) 69% of the random sample of 160 (and an even higher percentage of the second, targeted sample of 154 OA Journal authors: 83%) stated that they would willingly self-archive all of their articles if required to do so by their funders or their employers.

Now stop right there and think: Much of the research community has realized that OA would be a good thing, both for authors (in terms of impact) and for users (in terms of access). According to this JISC/OSI sample, only 39% of authors have self-archived (at least one of) their TA (Toll Access) journal articles and only 2% have published (at least one of) their articles in an OA journal (I will return to this). But that current total of 39% + 2% = 41% OA by the two means could immediately be raised -- willingly! -- to at least 69%, so the survey shows, if authors' employers and funders were simply to require it!

If the Swan & Brown survey's take-home message for university provosts and pro-vice-chancellors as well as for research funders is just this -- that the fastest and surest way to generate OA right now is to require your authors to provide it (by whichever of the two means is suitable for each article) -- it will have done a great service, and may just get OA provision up to speed at long last.

Another initiative is ready to be launched to help out: A Call for Institutional Commitment (to implementing an official institutional policy of requiring open access provision for all institutional research article output): not another declaration of support for the principle of OA provision, but a commitment to an institutional policy of requiring OA to be provided.


Now a little more on the Swan & Bown findings: Because of the survey's primary focus on OA publishing, half of the data were not from a random sample of journal authors but from a specifically targeted population of known OA Journal authors (154 of them responded). Then an approximately matching number of authors (160) was collected randomly from among authors in conventional journals.

Let us call the first population OAJ authors and the second population TAJ (Toll Access Journal) authors. It is important to understand that the TAJ authors actually constitute a much larger population (because over 95% of journals are TA Journals today and fewer than 5% are OA Journals). In fact, the TAJ population could have included any author who had ever published an article, i.e., it subsumed the OAJ population too, for there are as yet vanishingly few authors who have published only in OA Journals. This was borne out by the fact that 3 of the TAJ sample of authors did in fact turn out to have published in an OA Journal. (Those 3 were accordingly eliminated from the analysis and report -- because the two questionnaires were for various reasons different -- and the remaining 157 TAJ authors were then dubbed "NOAJ" authors: i.e., not-OA-Journal authors.)

But meanwhile, there was still that (not explicitly reported) finding that about 2% of TAJ authors (3/160, or 3 out of the 140 who were sure!) were OAJ authors. Add that to the finding that 39% (62/160) of the TAJ authors had provided OA through self-archiving), and you not only have a population estimate that about 39% of articles are OA today, but also that about 17 times as many articles are made OA via self-archiving than via OA journal publishing today.

(The true figure may be closer to 10 times as many: Alma Swan informed me that because the TAJ sample kept growing after the study deadline was reached, its size has since risen from 160 to 245. Eliminating 27 of these because the authors had stated that they did not know whether or not they had ever published in an OAJ [!], 8/223 reported that they had published in an OAJ, which raises the OAJ estimate to about 4%, which is close to the approximate OA proportion (5%) among OA + TA journals: about 1000/20,000. In the extended sample, the self-archiving estimate -- excluding OA journal authors, lest they were merely self-archiving their OA journal articles -- was 88/218 or 40%.)

Given that about 95% of journals are TA and about 5% are OA, this confirms that OA journal publishing is far closer to its current 5% ceiling than OA self-archiving is to its 95% ceiling (even though there is 10 times as much OA self-archiving as OA journal publishing, and even though 83% of journals and 58% of publishers have already given their official green light to self-archiving). http://www.ecs.soton.ac.uk/~harnad/Temp/Romeo/romeo.html

But the study also shows how easily this can be remedied: for 69% of TAJ authors (and 89% of OAJ authors) report that they would willingly self-archive if they were required by their employers to do so. So what is needed now is obvious: an official policy on the part of universities and research funders that Open Access must be provided for all journal article output. OA can be provided in either of the two ways -- by publishing the article in a suitable OA journal, if one exists, or otherwise by publishing it in a suitable TA journal and self-archiving it. But OA must be provided.

Requiring OA provision is not at all shocking -- or no more shocking than requiring publication at all. The "publish or perish" policies of universities and research funders already require that research must be published, rather than just put into a desk-drawer where no one can use it (in which case the research might as well not have been done at all). Mandated OA provision is merely a natural online-age extension of existing publish-or-perish policy, to the effect that toll-gated publication is no longer enough: it is merely a bigger desk-drawer, insofar as all would-be users whose institutions cannot afford the access-tolls are concerned. The growing evidence for the dramatic increase in research impact that results from making our research OA shows just how counterproductive that larger desk-drawer really has been for research progress all along (Lawrence 2001, Kurtz 2004, Brody et al. 2004.

Now that we are in the online age, it is time for the research community to make up its mind to come out of the desk-drawer, and provide open access to all of its peer-reviewed journal article output.

Many thanks to JISC/OSI and to Alma Swan and Sheridan Brown for providing this valuable study that clearly points the way.

Stevan Harnad

"The Need To Re-Activate the Provosts' Initiative"

"On the Need to Take Both Roads to Open Access"

"The Green and Gold Roads to Open Access"

"What Provosts Need to Mandate"

"Open Access Provision Policy"

"University policy mandating self-archiving of research output"

"Meeting: National Policies on Open Access Provision for University Research Output"


Brody, T., Stamerjohanns, H., Harnad, S. Gingras, Y. & Oppenheim, C. (2004) The effect of Open Access on Citation Impact. Presented at: National Policies on Open Access (OA) Provision for University Research Output: an International meeting, Southampton, 19 February 2004. http://opcit.eprints.org/feb19prog.html http://www.ecs.soton.ac.uk/~harnad/Temp/OATAnew.pdf

Cox, J. & Cox, L. (2003) Scholarly Publishing Practice: The ALPSP report on academic publishers' policies and practices in online publishing. Association of Learned and Professional Society Publishers. http://www.alpsp.org/2004pdfs/SFpub210104.pdf

Harnad, S. (1995) Electronic Scholarly Publication: Quo Vadis? Serials Review 21(1) 70-72 (Reprinted in Managing Information 2(3) 1995) http://cogprints.ecs.soton.ac.uk/archive/00001691/00/harnad95.quo.vadis.html

Harnad, S., Carr, L., Brody, T. & Oppenheim, C. (2003) Mandated online RAE CVs Linked to University Eprint Archives: Improving the UK Research Assessment Exercise whilst making it cheaper and easier. Ariadne 35 (April 2003). http://www.ariadne.ac.uk/issue35/harnad/

Hitchcock, S., Woukeu, A., Brody, T., Carr, L., Hall, W., and Harnad, S. (2003) Evaluating Citebase, an open access Web-based citation-ranked search and impact discovery service http://opcit.eprints.org/evaluation/Citebase-evaluation/evaluation-report.html

Kurtz, M.J. (2004) Restrictive access policies cut readership of electronic research journal articles by a factor of two, Michael J. Kurtz, Harvard-Smithsonian Centre for Astrophysics, Cambridge, MA http://opcit.eprints.org/feb19oa/kurtz.pdf

Lawrence, S. (2001) Online or Invisible? Nature 411 (6837): 521. http://www.neci.nec.com/~lawrence/papers/online-nature01/

Swan, A. & Brown, S.N. (2004) JISC/OSI Journal Authors Survey Report. http://www.jisc.ac.uk/uploaded_documents/JISCOAreport1.pdf