Elsevier has released a new scholarly article sharing policy which is definitely more disappointing than really any cause for cheer.

Basically the crux is that the only place that authors are allowed to have the final publication version of an article in a non-open access Elsevier publication is on the Elsevier website itself. Of course, after any embargo period has elapse or if the author has paid an author processing charge and published in a hybrid or gold open access journal, they are allowed to post the article on their own webpage or institutional repository.

During the time that the article is most important for scholars to access, it’s Elsevier only. Which is not a surprising policy in many ways for a publisher to have, after all they want to maximize their subscription fees as well as APCs not to mention traffic to their sites.

But an issue that I (and many others) have with this new policy is that it may very well be in direct contravention to what authors are required to do to meet various institution and national open access policies. Canada’s new policy requires open access to the final version within 12 months of publication, much shorter than many journal’s embargo period.

As such, this policy is potentially setting authors against their funders. And will no doubt cause many authors to either ignore the policy or put pressure on the government to water down the requirements.

The requirement for a CC-BY-NC-ND license is also much too restrictive, forcing authors to adopt a licence that isn’t the generally accepted (particularly in STEM fields) open access license of CC-BY.

And I could go on. The policy is very long and very detailed, more than probably most people want to wade through. This length and complexity is an issue too. Pressed for time in a publish or perish world, it’s tempting to skip to the end and just forget about sharing — because it’s just easier to do nothing and leave the article as is on the Elsevier site! The pain and anguish involved in sharing are a disincentive.

There is a way to fix this, and it’s not even hard. The policy does mention the physics/math/CS/etc preprint server arXiv by name (and RePEC for economics): “Preprints may be shared, and on arXiv and RePEC they may be refreshed with accepted manuscripts.” It’s easy. Allow all scholars the courtesy and convenience that those that use arXiv & RePEC have. Allow preprints posted to a disciplinary or institutional repository to be refreshed with accepted versions upon publication. If that isn’t a deal breaker in some fields, why is it a deal breaker in all the rest?

As is my habit, I’ve collected a fair bit of recent commentary on this new Elsevier policy. Many of the authors below go into far more detail than I have here about the various issues.

I’m including a bit on the STM principles for article sharing on scholarly collaboration networks, which were the basis for the new Elsevier policy. STM is a STEM publisher industry group. I’ve also included a couple of recent ones on Elsevier that aren’t specifically about this issue for some wider context.

As usual, if I’ve missed anything significant please add it in the comments. If this issue continues to have legs, I’ll probably update this post at some point.

Update 2015.05.28. This story does seem to have legs, so I’ve added a bunch of items.


  1. […] to do around the issues facing publishers, authors, and libraries with regards to Open Access. (John Dupuis just posted a good summary of where we are with the Elsevier policy.) So create […]

  2. […] than 12 month embargo with ranges from 12 to 48 month embargoes). Assessments such as those made by John Dupuis, that the policy is, “potentially setting authors against their funders,” have prompted […]

eXTReMe Tracker