Embargoed articles
61 views
Skip to first unread message
Henrik Spoon
Sep 1, 2026, 12:24:20 PMSep 1
to OpenAlex Community
Hi,
I am one of the authors of “Math and Aftermath: Impacts of Unbundling a Large Journal Package on Researcher Perceptions and Behavior.” Library Resources & Technical Services 69, no. 1 (2025): 1. Our paper covered the first four years after the unbundling (2019-2022). We are eager to extend the analysis to 2025 (and possibly 2026 if we are slow).
We are interested to include the effects of OA publishing and embargoed OA on the interpretation of the Counter compliant turn-away statistics for the 853 journals that we cancelled per January 2019.
Here is my question:
Does OpenAlex update the OA status of an article after the embargo period is over?
Thanks in advance,
Henrik Spoon
Physics, Astronomy and Mathematics Librarian
Cornell University Library
Gabor Schubert
Sep 2, 2026, 6:05:03 AMSep 2
to OpenAlex Community
Hi Henrik,
It seems that OpenAlex/Unpaywall somehow tracks the date when something becomes OA: See under "OA dates" here: https://help.openalex.org/data/works/open-access/
Though it might be quite tricky to observe the effects of emargoed OA on the COUNTER statistics you are using in your study. There are probably many journals which allow the repository of a "Green OA" (accepted manuscript) version of the articles after an embargo period, although it is not obvious that all the articles will use this possibility. Most likely OpenAlex/Unpaywall tries to locate these Green OA versions continously, and they are updated in the database later at some point.
There are also a few journals which publish all of their articles as OA on the journal homepage after an embargo period with a CC license (delayed OA), but this is probably not very common. One such example is the "IJREE – International Journal for Research on Extended Education" (https://budrich-journals.de/index.php/IJREE/openaccess) which publishes all articles after 12 months embargo as OA. Although as far as I see OpenAlex struggles with this: if one checks the latest articles in this journal (https://openalex.org/works?filter=primary_location.source.id:s4210222983&include_xpac=true&sort=publication_year:desc&page=1) they seem to be OA in OpenAlex but they are not OA in reality. Like this article: https://openalex.org/works/w4403980939?tab=locations , which seems to be OA in OpenAlex (and Unpaywall) but not OA yet at the journal homepage (https://budrich-journals.de/index.php/IJREE/article/view/47194).
Best regards,
Gabor Schubert
Stockholm University
Bianca Kramer
Sep 2, 2026, 3:19:37 PMSep 2
to Gabor Schubert, hs...@cornell.edu, OpenAlex Community
Hi Henrik, Gabor, all,
A word of warning on oa_date - this variable used to be super useful to at least approach empirical analyses of observed embargo periods. I always thought the methodology behind it quite clever for repository-based OA, though not without caveats of course.
From August 2020, the variable oa_date was added to oa_locations, indicating when the full-text document first became available at that location. For repository based OA, oa_date represented the OA-PMH timestamp when the full text article was first detected in the repository. Before the 2025 code rewrite, oa_date was available for 28.4% of repository versions.
After the rewrite, oa_date is populated for all OA repository versions (as well as for all publisher-OA versions), but in all cases, it is set to be equal to the variable published date. This means the variable oa_date can no longer be used to calculate empirical embargo periods, which is relevant both for analysis around OA policies and mandates, and comparative analyses of developments around repositories.
Attached see, as example, an aggregated analysis of oa_date for all best_oa_locations, compared to published_date. This is using Unpaywall data from 2026-08-06.
I honestly don't know why the documentation in OpenAlex still gives the 'old' information on what oa_date entails, but as shown, it currently does not provide any useful information on embargoes...
kind regards,
Bianca Kramer
Sesame Open Science
From August 2020, the variable oa_date was added to oa_locations, indicating when the full-text document first became available at that location. For repository based OA, oa_date represented the OA-PMH timestamp when the full text article was first detected in the repository. Before the 2025 code rewrite, oa_date was available for 28.4% of repository versions.
After the rewrite, oa_date is populated for all OA repository versions (as well as for all publisher-OA versions), but in all cases, it is set to be equal to the variable published date. This means the variable oa_date can no longer be used to calculate empirical embargo periods, which is relevant both for analysis around OA policies and mandates, and comparative analyses of developments around repositories.
Attached see, as example, an aggregated analysis of oa_date for all best_oa_locations, compared to published_date. This is using Unpaywall data from 2026-08-06.
I honestly don't know why the documentation in OpenAlex still gives the 'old' information on what oa_date entails, but as shown, it currently does not provide any useful information on embargoes...
kind regards,
Bianca Kramer
Sesame Open Science
Op wo 2 sep 2026 om 12:05 schreef Gabor Schubert <gabor.sch...@gmail.com>:
--
You received this message because you are subscribed to the Google Groups "OpenAlex Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openalex-commun...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/openalex-community/2365676d-a1a3-4d31-9903-ceaf433a899bn%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages