9 like 0 dislike
in Open Science by (330 points)

Many scientists are under the impression that the majority of interested readers for their work have subscription journal access. I find this unlikely as a very small percentage of the global population has subscription journal access. To help illustrate the suicidal impact subscription publishing has on one's research, it would be helpful to have some estimates of the percent of individuals denied access to subscription works.

While such an estimate comes with many caveats, does anyone know of statistics or resources that could help answer this question? For example, how many people are affiliated with organizations that purchase institutional access? Even if the metrics are for a single journal, that could be a helpful starting point.

Note: this question was originally posted to the OpenCon Community Discussion Listserv.

This post has been migrated from the Open Science private beta at StackExchange (A51.SE)
by (130 points)
0 0
This a great question and one I want to get hard data on too. I've often wondered why librarians don't openly publish lists of journals (and the age-range of accessibility e.g. 1990<->2015) they do / don't have subscriptions for at each higher education institution. It would be really interesting and I don't think it would be difficult to do. This seems like something librarians could definitely be in a position to answer.

This post has been migrated from the Open Science private beta at StackExchange (A51.SE)
by (190 points)
0 0
This could be a research paper in itself! I believe that most journals either publish or are willing to provide subscription statistics. After all, how do you prove that you are relevant if you do not demonstrate the reach of your publication? The hard part is collecting all of that information...

This post has been migrated from the Open Science private beta at StackExchange (A51.SE)

1 Answer

1 like 0 dislike
by (125 points)
edited by

The answer to this question is extremely complex, because it should have to reckon with:

  1. Difference between ongoing access and backfile/archive access to previous volumes
  2. Whether or not to include so called "walk in access", unaffilated civilians can get in  librarians
  3. Whether or not to do this at an article level (to include OA access in hybrid journals)
  4. Whether or not to include access to journals in packages (e.g. those from EBSCO), that may have a volatile composition)
  5. Access through personal subscriptions
  6. Access through company subscriptions
  7. Which (pay-walled) journals to include. All, all languages for everyone?


Some data you could work with a come up with some alternative proxies as answers to your question:

  1. Data provided by SciHub
  2. Data in Worldcat on library holdings
  3. Data on the total number of university affiliated researchers and students (from UNESCO, but incomplete
  4. Data held by ResearchGate on number of papers requested from authors
  5. Data held by Open Access Button on papers people quest access for
  6. Usage of the #icanhazpdf hastag
  7. Journal subscription lists of libraries (insofar as openly available)

For a very rough estimate you could use the recent UNESCO policy paper published by UNESCO (policy paper 30, april 2017: "Six ways to ensure higher education leaves no one behind" available at http://unesdoc.unesco.org/images/0024/002478/247862E.pdf) that estimates total enrollment in tertiary education worldwide at 207 million in 2014. Add to that the estimated 8 million researchers (head count) worldwide (again UNESCO data: UNESCO science report - towards 2030. Paris: UNESCO, 2015) and you arrive at 215 million people that might have access to some journals via institutional subscriptions. Divided by the 7.35 billion total population for 2014 (UN figures) that amounts to 2.9%. Using the above considerations you could judge whether that is an overestimation or an underestimation.



Ask Open Science used to be called Open Science Q&A but we changed the name when we registered the domain ask-open-science.org. Everything else stays the same: We are still hosted by Bielefeld University.

If you participated in the Open Science beta at StackExchange, please reclaim your user account now – it's already here!

E-mail the webmaster

Legal notice

Privacy statement