In a recent meta.physics.se discussion on the use of permanent links form physics.stackexchange, I proposed as policy that

when linking to the arXiv, link to the abstract page (arxiv.org/abs/...) and not directly to the pdf, as a courtesy to users on lower bandwidths. If the paper has been published (which will be noted on the arXiv abstract), include a link to the journal version, even if it's paywalled, if only as a courtesy to the referees' hard work.

One comment on the italicized part was

Doesn't pretty much everybody post the final version on arxiv?

and I would like to know (if that is even possible) to what extent this is true. My gut feeling is that this is field-dependent, but I do not know what evidence may be available to explore this, other than anecdotal evidence.

Thus: Is there a clear trend for authors to upload post-referee-process versions after publication? Does this depend on the specific field? What evidence is available for this?

  • 4
    Until recently, I thought that posting to arXiv the post-refereed version was considered bad form, especially if the author has already transferred copyright to the journal. However, I recently discovered that this is allowed by many publishers, including Elsevier. You can check on what is allowed by specific publishers.
    – Dan C
    Aug 22, 2013 at 13:10
  • 1
    @DanC Thank you for the link. I know that publisher policies on what is allowed vary, but for now I'm mostly interested in whether people actually take advantage of those rights or not, or in which fields they do.
    – E.P.
    Aug 22, 2013 at 13:13
  • In math (specifically math.CO), I rarely (maybe 1 out of 20 or rarer) see comments stating that the paper is post-refereed. I think the Elsevier policy on posting post-refereed versions is probably fairly new, in light of the Elsevier boycott. As a result, I think more people may take advantage of it as they become aware. For the time being, I believe it is uncommon.
    – Dan C
    Aug 22, 2013 at 13:18
  • 1
    @AnonymousMathematician The arXiv submission process has a Comments field. When I submit a revision, I use that field to describe the changes from the previous version. I suspected that was standard, but perhaps it is not.
    – Dan C
    Aug 23, 2013 at 3:21
  • 2
    @AnonymousMathematician: moreover, it has a "journal-ref" field, where you can give the final citation. I'm frequently wondering whether arXiv (or other author archived manuscripts) are the final version, and of what paper exactly, so I find that possibility an important help and try not to forget updating it. Aug 23, 2013 at 8:00

1 Answer 1


It is a lot of work to find out which paper in arXiv has actually the same content as a published paper. However, taking arXiv's journal-ref field as a surrogate indicating which paper has a journal version out, I arrive at these fraction of papers: fraction of arXiv papers with set "journal-ref" field

So it looks as if the gut feeling that this is field-dependent is true: high energy and nuclear physics have consistently high fractions of papers with journal-ref field set. On the other hand, this seems far less common for maths papers. Note that a certain drop at the end is to be expected as the preprints may be uploaded to arXiv considerably before the paper is accepted, printed and the journal-ref is updated in arXiv.

Here's the marginal plot over all fields: fraction of arXiv papers with journal-ref set So overall, about 45% of the papers have a journal-ref. However, it is set for almost 60% of the 2001 - 2007 papers. Whether this is a real decline or just the lag between preprint and paper publishing remains to be seen (in a few years).

Also, neither the overall nor the subject-dependent trends seem to have a convincing increasing trend, even if the decline due to "not yet published and updated" is assumed to be the sole cause of the decreases.

For the complete picture, here's the development of total number of papers: number of papers in arXiv

and number of papers with journal-ref set: number of arXiv papers with journal-ref set

  • 2
    Thanks for making the effort to process the data and generate these graphs.
    – Dan C
    Aug 23, 2013 at 3:24
  • 6
    I guess the drop at the end is in fact mostly the publication delay. This is also field dependent - math journals are known for extraordinary publication delays, and that is very apparent in your plots.
    – silvado
    Aug 23, 2013 at 8:31
  • 2
    @silvado: yes I agree. However, I wasn't aware that a delay of 6 years sounds reasonable e.g. for maths. In my field (chemistry/chemometrics) the delay is much shorter, also because we have to be careful whether pre-prints are considered previous publication, so I put my preprints to arXiv only after submission or even after acceptance/end of the embargo time. That makes much shorter to no delay. Aug 23, 2013 at 9:10
  • 2
    @xioxox: May I ask whether you have a particular reason for not setting journal-ref? I use it because I prefer "proper" citations and do not want to put any obstacles to people who wonder whether and to what extent the paper on arXiv and the journal version are the same. In fact, I put a box into the arXiv version stating that they are equivalent. Sep 12, 2014 at 10:29
  • 1
    @cbeleites: I don't have the reference when I put it on after acceptance. No one in astronomy/astrophysics bothers adding it later. I think this likely because we all search for papers with ADS (adswww.harvard.edu), which automatically associates the journal and astro-ph articles, giving the journal article by default.
    – xioxox
    Sep 13, 2014 at 12:54

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .