Free access to the scientific literature has the greatest effect on the poor — but not poorest — countries, a new study finds.
A short article, “Open Access and Global Participation in Science,” was published in today’s issue of the journal Science. It was authored by sociologist James Evans and graduate student Jacob Reimer, both from the University of Chicago. The journal also hosts a podcast interview.
You’ll recall that Evans was responsible for another controversial article published last summer claiming that online access to the scientific literature results in a narrowing of citations.
Today’s article measured the effect of commercial and free access to the scientific literature on article citations. It focuses on differences across rich and poor countries based on their per capita gross national income.
Evans and Reimer report that the influence of free access on citations was much smaller than other have proposed, about 8% for recently published research. It was more than twice as large, however, for the developing world, although the effect diminished for the poorest countries. They write:
our work provides clear support for its ability to widen the global circle of those who can participate in science and benefit from it
Advocates for open access will see this article as supporting their cause. But those who spend time reading the methodology will notice that message is not as clear as the article implies.
The researchers are not comparing open access journals with subscription-access journals, as reported in the recent article by Tove Faber Frandsen. Evans and Reimer are comparing the effect of freely available articles to subscription-access articles. But this is still an oversimplification.
Due to the size of the study (26 million articles published between 1998 and 2005 in over 8,000 journals), the researchers were unable to code individual articles as being OA or not, so they coded entire volumes. For example, articles from the journal Science are OA when they are older than one year. Articles from PNAS are all subscription-access in the first six months (in spite of the fact that about one-third are author-pays OA), after which they are all coded OA. Because of the macro-level of the study, no attempt was made to find other sources of free copies. In other words, this study focuses entirely on open access publishing. Some freely available articles will be coded as subscription-access articles, and the result is an overly conservative estimate of the open access effect.
The important detail that may be missed is that the source of the vast majority of OA articles in this study were published by non-profit scientific societies who use the subscription model in tandem with a delayed-access model. If anyone should be claiming victory, it should be them.
(In next week’s post, I’ll explore the debate over Open Access from a framing standpoint and why non-profit scientific publishers have been grouped with commercial publishers as villains and impediments to the dissemination of science. )