Half of all papers from 2011 are open access

Open access publishing is growing far faster than previously thought, according to a new report prepared for the European commission. Half of all research papers published in 2011 are now freely available online, it is claimed – nearly twice as high as previous estimates.

The research was carried out by Science-Metrix, a Canadian research evaluation firm, as part of a wider study on open access for the commission’s Directorate-General for Research and Innovation.

‘We basically took a sample of 320,000 papers from various databases and developed “harvester” software to look for the availability of articles,’ says Science-Metrix CEO Éric Archambault. ‘If it was able to download a paper it would classify it as a free article.’

The software searched papers across different disciplines, published between 2001 and 2011 in Europe, Brazil, Canada, Japan and the US. The sample included material from online databases, such as Scopus and PubMed, as well as the websites of publishers, institutions and researchers. Over 40% of peer-reviewed articles published worldwide over the whole 10-year period could be downloaded for free, and for 2011 alone this figure rose to around 50%.

The report’s estimates are significantly higher than those generated by similar studies, most of which put the proportion of published papers that are open access in the region of 20–30%.

Archambault says the difference in estimates could be a result of how open access is defined. Science-Metrix counted any articles that were freely accessible online as open access, so their results also take into account papers uploaded onto websites by universities and individuals. Previous studies have focused solely on papers available from open access journals or repositories.

‘I’m delighted to see the increase in the number of open access articles,’ says Peter Murray-Rust, a chemist and keen advocate of open access at Cambridge University, UK. But he adds the murky definition of ‘open access’ can distort the results of studies like this. ‘Some journals do not label their articles so it is impossible to tell whether they’re open or not,’ he says. ‘Simply being on the web is not a guideline.’

Archambault acknowledges that the Science-Metrix study doesn’t recognise copyright breaches, or distinguish between different kinds of open access. But he says it is in line with other factors that suggest research is becoming more available. ‘The open access movement has reached a kind of critical mass,’ he says. ‘It’s only going to accelerate. There are a lot of people behind it: governments, academia, even publishers to an extent. It’s here, and it’s here to stay.’