Monday, December 31, 2012

December 31, 2012 Dramatic Growth of Open Access

Updated Jan. 1, 2013 thanks to very helpful informal peer review from Pablo de Castro (E-LIS figure and explanation for discrepancy in growth rates of OpenDOAR and ROAR).

2012 was another awesome year for open access!. This post highlights and celebrates just how much open access is available already. The Bielefeld Academic Search Engine (BASE) cross-searches over 40 million documents in over 2,400 repositories - nearly double the number in just 3 years, demonstrating yet again strong growth in open access archiving. The Directory of Open Access Journals, in addition to listing fully open access journals, provides an article-level search that is available for a growing percentage of DOAJ journals. The number of articles searchable through DOAJ is just under one million - a third of this growth taking place in 2012 . What a great way to show that not only is the number of open access journals growing rapidly - more importantly, the number of articles published in open access journals is growing even more dramatically!

A heartfelt thanks to everyone around the world who is making open access happen - all of the scholars sharing our work, readers, advocates, repository managers, publishers, librarians, and others. All the best to you and your OA endeavours in 2013!

Notable annual growth by percentage

Directory of Open Access Journals - articles searchable by article level grew by 234,449 articles to a total of 955,720 articles, an increase of 33% on a substantial base. The Registry of Open Access Repositories showed a growth rate of 730 repositories or 28%, a somewhat puzzling contradiction to the relatively slow growth rate of OpenDOAR. The number of journals depositing all articles in PubMedCentral as open access grew by 191 to a total of 893, a 27% increase this year, and the number of paper downloads from the Social Sciences Research Network was more than 2 million higher this year than in 2011 (10 million downloads total), for a 27% increase.

The numbers

Directory of Open Access Journals
8,519 journals
2012 growth: 1,147 journals (3 journals / day)
# articles searchable at article level: 955,720
2012 growth in searchable articles: 234,449 (642 articles / day)

Directory of Open Access Books
1,259 academic peer-reviewed books from 35 publishers
new in 2012

Electronic Journals Library
37,805 journals that can be read free of charge
2012 growth: 5,421 journals (15 journals / day)

Highwire Press Free Online Articles
2,151,420 free articles
2012 growth: 41,640 articles (114 articles / day)

2,253 repositories
2012 growth: 89 repositories (7 repositories / month)

Pablo de Castro (see comments) explains the discrepancy in the growth rates of OpenDOAR and ROAR: the growth rate mismatch between ROAR and OpenDOAR is due to an intensive database cleansing by the latter towards the end of 2012 that led to removal of a number of wrong/outdated entries.

Registry of Open Access Repositories (ROAR)
3,340 repositories
 2012 growth: 730 repositories (2 repositories / day)

Bielefeld Academic Search Engine (BASE)
40,506,905 documents
2012 growth: 6,908,293 documents (18,926 documents / day)

2,600,000 articles (from PMC site)
2012 growth: 300,000 articles (from PMC site - update schedule not known so not sure about accuracy)
1,199 journals deposit all articles in PMC
2012 growth:  220 journals (.6 journals / day

809,849 e-prints
2012 growth: 83,886 e-prints (230 e-prints / day)

14,206 documents as of Jan. 1, 2013 (courtesy Pablo de Castro - see comments)
14,242 documents as of Dec. 11 - cannot find # of documents on new site (E-LIS migrated to a new e-prints server in the past few days - looks great!)

Social Sciences Research Network
372,772 full-text papers
2012 growth: 65,715 full-text papers (180 / day)

Open Access Mandate Policies (from Registry of Open Access Material Archiving Policies)
353 open access policies (total)
2012 growth rate: 44 policies  (4 policies / month)

Internet Archive
1,110,878 movies
110,448 concerts
1,474,756 recordings
3,781,142 texts
(new in 2012)

The full data edition can be downloaded from SUMMIT, the SFU Institutional Repository

A perspective 

Correction Jan. 13, 2013 - the correct number is .4% not .04% as originally posted. Thanks to Bryan Thompson (see comment below).  Informal peer review of this kind is very much appreciated! There's a lot of math and a second pair of eyes always helps, especially since I often do this work quite late in the evening.

Elsevier claims to serve "more than 30 million scientists, students and health and information professionals worldwide" from:

This is .04% of the 7 billion people on this planet open access aims to serve! Note that internet access and literacy is required to take advantage of open access to the scholarly literature - but unlike Elsevier, open access is not placing barriers, we're just doing the best we can to overcome them.

This post is part of the Dramatic Growth of Open Access series. 


  1. Anonymous8:57 AM

    Hi Heather, this is a comment from @PubAdvisor in reference to your post: The dramatic growth of Open Access. I am a scientist who supports OA and its growth but who is also surprised by a peculiar narrative. That is: the rise of OA as a means to support exclusively the spreading of scholarly information. I am more concerned by the content of scholarly information rather than its instantaneous dissemination. Working at the base of scientific endeavour (in the lab) I know that a large amount of scholarly information is wrong. This problem will certainly increase with the growth of OA. What are your thoughts on the necessity of filtering reproducible information from the ever-increasing flood of data that has been and will be produced by the "publish or perish" paradigm in science? You seem to have noticed that discussion via blog commenting can be disruptive, useless and cluttering your site. Excluding this possibility, what else can be done to make use of filters? Are we going to move into a world of more information, including scholarly content, without the possibility to verify what is correct and what is not?

    1. Thanks @PubAdvisor. There is a growing flood of information, however this has nothing to do with open access. De Solla Price in Little Science Big Science, for example, pointed out that the number of scientists and scientific journals and articles has been growing exponentially at a fairly steady rate since the first scientific journals were published in the 1600's. Publish or perish certainly contributes to this. I am more concerned about the negative impacts of filtering out, even accidentally, rather than the scholar's task of filtering in. I have been giving some thought to potential solutions. 3 possibilities that I'm considering at the moment:

      - research the correlation between quality of work and real quantity of publishing (I anticipate a negative, based on work with scholarly monograph publishers). That is, establish evidence that pushing scholars for high quantity of publication is not the way to quality scholarship, to help change the academic reward structure (e.g. allow a maximum number of publications for consideration by t&p committees)
      - change the academic reward structure to reward contribution rather than results. Rewarding results gives incentive to find results - much important scholarly work involves checking out and rejecting false paths.
      - work collaboratively rather than individually (fits well with the above). Writers on this topic include De Solla Price, Bruno Latour, and Fleck. Contributions need not be journal articles. If there are fewer journal articles, this would reduce the burden of reading for scholars, as well as reduce the likelihood of missing important work because no one has time to read all those articles.

      This is not a complete answer, but is an early version of an "open research agenda brainstorm" I'm thinking of publishing.

  2. Thanks for this comprehensive account of the level of success of OA in 2012. A couple of tiny adds to your figures: according to a quick calculation based on the 'browse by year' functionality,, the number of E-LIS items as of Jan 1st is 14,206. And: the growth rate mismatch between ROAR and OpenDOAR is due to an intensive database cleansing by the latter towards the end of 2012 that led to removal of a number of wrong/outdated entries.

  3. Thanks, Pablo! Your friendly informal peer review obviously requires some revisions in the post and spreadsheet, so I've made the changes and credited you as the reviewer. This is an example of the vision that I have of a future for open peer review to accompany a new form of scholarly communication. Because your comments are open and signed, it is easy for me to credit you. Now what we need is to figure out how to give people who do this appropriate credit. This is true of other kinds of useful input and feedback. The reason this comment follows your comment in particular is the easy comparison with peer review and revising the work.

  4. Anonymous4:15 AM


    There are now Tables of Contents for the latest issues of over 5,000 Open Access journals included in JournalTOCs


    Roddy MacLeod

  5. Hi Heather,

    Thanks for these really thoughtful and incredible numbers. Have you considered doing data visualizations or maybe a short video to convey them? They're pretty mind-blowing.

    If you're in NYC, you should come to the MediaCamp of JustPublics@365 (, a joint initiative of the Graduate Center CUNY and the Ford Foundation. #JP365 is helping academics build the capacity to learn all sorts of tools to share their work with the media and activists, and your work is especially so exciting.

    Also, you should chat with Jason and Heather at That's some great work you would enjoy.

    Jack @jgieseking / @justpublics365

    1. Thanks for your comments, Jack. Normally I don't post promotional comments like this one, however I will make an exception when it comes to projects like JustPublics@365 which is aiming to further the public good.

      I am familiar with ImpactStory and am actively encouraging Jason, Heather, and everyone who works with them to take a critical stance on altmetrics, particularly those based on social media. For example, long before anyone decides "much-tweeted" means "important" or "quality" in an academic sense, we should consider things like this: what articles will a pharmaceutical company point to / tweet - the ones that show the benefits of their products, or the studies that either show no benefits or found those awful side-effects? I would fully expect such metrics to reflect and amplify social biases - the works of men tweeted more than those of women, the works of the wealthy majority more than the wealthy minority. Not to mention the ease with which such metrics could be deliberately manipulated by those with money. Think about how much money big oil has to hire people to tweet, facebook, etc., their preferred "science", i.e. climate change denial.

      If you hear of this kind of appropriately critical scholarship applied to altmetrics, please let me know as I'd be most interested.

  6. Hi Heather, great article and I always love reading the quarterly posts but I think the math is a little off. If there are 30 million people getting access, it's .4%, not .04%. Sorry to be a stickler, love the site!

  7. Oops - much appreciated, I've double-checked and fixed this, Bryan! To clarify: Elsevier claims to serve .4% of the people open access claims to serve, not .04%.


Thank you for your comment. Comments on IJPE are moderated.