Lots of Open Access publishers, but how many readers?

When I started graduate school in 2003, PLOS Biology was just about to launch. There was no PLOS One and the fraction of publications in open access (OA) journals was miniscule. It is remarkable what the progress has been – a few days ago, a study by Science-Metrix for the European Commission indicated that over 50% of all 2011 papers are open access, calling this moment a “tipping point” towards open access (though I can’t wait for the moment when 100% of research is open, the progress is truly astounding). However, the real measure of success for the open movement is not just a count of freely available papers from a year ago. The real question is – what fraction of papers that a researcher needs today are openly available? Regulatory requirements force many non-OA journals to make articles freely available after 6 or 12 months. But no scientist is going to wait a year to read a key discovery in her field, so libraries must continue to pay subscriptions. Also, when I mention how successful the OA movement has been in transforming the publishing landscape, I often hear in reply, “PLoS One may publish a lot, but who reads those articles? How many of the important papers are open access?” Valid question, and one that we can actually answer with the data in PubChase. Our users save articles to their libraries, and that lets us ask what fraction of the saved (and by implication important/read) papers are indeed OA. We do not consider delayed OA where articles become freely available months after publication to be actually open. Therefore, in this analysis, we only included as OA journals that make all of the research articles instantly available upon publication ((Specifically, we count as OA journals from the PubMed Central list that have "Immediate" or "0 months" access type, and whose participation level is "Full")). The answer for PubChase users is that 22% of the 2013 articles saved to libraries are open access. ((Articles do not have to be unique; if 100 users save the same article to their libraries, the count for that journal is increased by 100)) We can also plot how this fraction has grown over the last decade ((The PubChase by-year journal table used for this plot is available here)).     Perhaps not a tipping point yet, but a monumental achievement nevertheless. As in our previous post, caveats apply. PubChase users may not be representative of biologists as a whole. For the current year, the number is still in flux because the year is not over. Also, due to hybrid journals that publish a mix of OA and non-OA articles, our approach underestimates instantly-OA publications. There are three conclusions we can draw from this analysis. First, to OA advocates – there is still a lot of work to do. Second, to the critics dismissing the importance and relevance of open access journals – you are wrong – the articles are important and read widely. Third,  if you are a subscription journal clinging on to the subscription model – now is the time to prepare for the imminent 100% open access future. (Post by Lenny Teytelman. Analysis by Matt Davis, Alexei Stoliartchouk, and Lenny Teytelman.) P.S. If you are curious about PLOS One articles specifically, yes they are widely read. In total, a fifth of all the OA articles in the libraries of our users are from PLOS One.