European funders have been leading a charge under ‘Plan S’ to make more of the scientific literature free to read. Yet the nations that publish the highest proportion of their research papers open access (OA) aren’t in Europe, according to a preliminary analysis shared with Nature. Instead, countries in southeast Asia, Africa and South America are leading the way — thanks to a flourishing network of local open-access journals and publishing portals.
Indonesia may be the world’s OA leader: the study found that 81% of 20,000 journal articles published in 2017 with an Indonesia-affiliated author are available to read for free somewhere online, and 74% are published with open-access licences, meaning they could be legally redistributed (see ‘Open-access leaders’). More than 60% of research articles with authors in Colombia, Bangladesh and Brazil are free to read.
These estimates might be surprising, but they fit the expectations of those who follow publishing trends, says Ginny Hendricks, community director at Crossref, a non-profit organization in Oxford, UK, with whom more than 12,600 publishers deposit information about their articles. “People may be so obsessed with Plan S and Europe that they’ve taken their eye off the ball on progress in other countries,” Hendricks says.
The data, which have not yet been published, were collected by Heather Piwowar, co-founder of non-profit organization Impactstory in Vancouver, Canada. (The study covered papers published between 2015 and 2018; Nature is presenting the 2017 analysis because many journal articles are not made open until a year or more after publication, meaning the 2018 data may not be as stable.)
Piwowar used the firm’s Unpaywall service to automatically search online for journal articles recorded in Crossref — more than 2.4 million in 2017 alone. The tool looks for free papers at publisher websites and checks whether they are OA-licensed, and searches repositories for other legally free-to-read versions of the papers, including preprints. Piwowar then linked the author affiliations of these articles to their respective research institutions and countries.
Nations in southeast Asia, South America and Africa that have smaller publishing outputs — fewer than 3,000 papers each in the analysis — also score highly on OA proportion. Authors from Nepal, Peru and Uganda all published more than 70% of their papers open access, and Sri Lanka and Kenya came in at above 60%.
They surpass the leading European nations — Croatia and the United Kingdom, both at 60% — and the 2017 world average of 41%.
Broadly speaking, the countries that score highly do well because they have low-cost OA journals and websites, often supported financially by governments that encourage open publishing. For instance, the SciELO online library, which publishes articles mainly from Latin American countries is subsidized by government funders in Brazil and other countries, and a South African non-profit organization called African Journals Online hosts African-origin titles.
Indonesia, meanwhile, has seen a surge in local open-access journals registered with Crossref, Hendricks says. These titles are published mostly by universities and have no or very low fees, says Walt Crawford, a retired library-systems analyst in Livermore, California, who has analysed the Directory of Open Access Journals, a large community-curated listing of more than 13,000 publications. “I think it’s about time that these efforts were recognized, and the scholars taken seriously,” he says.
A nation’s disciplinary focus could also influence the trends, notes Thed van Leeuwen, a bibliometrician at Leiden University in the Netherlands who is project leader of the European Commission’s Open Science Monitor, a site that tracks open-access trends. Some African countries, for instance, might have a strong focus on health and disease research, which is often funded by large, international health organizations that mandate (and pay for) open publishing.
But the openness of a country’s research is tough to estimate, because different databases cover different subsets of journals — and no database includes all articles worldwide. Open Science Monitor, for example, searches Scopus, which covers fewer local journals than Crossref. The countries that lead Piwowar’s analysis in OA uptake don’t score as highly in that database, van Leeuwen says — although they still do well and, in some cases, outperform European countries.