On 2013-09-13, at 7:20 AM, David Solomon <dsolomon@msu.edu> wrote:

We have made the data available for the paper: "The publishing delay in scholarly peer-reviewed
journals" by Bo-Christer Björk & David Solomon recently accepted for publication in  
Journal of Informetrics.  

Paper http://tinyurl.com/ms5dk2u
Codebook  http://tinyurl.com/m2fwxtk 
Data http://tinyurl.com/mslr3c7

Abstract: Publishing in scholarly peer reviewed journals usually entails long delays from
submission to publication.  In part this is due to the length of the peer review process and
in part because of the dominating tradition of publication in issues, earlier a necessity of
paper-based publishing, which creates backlogs of manuscripts waiting in line. The delays
slow the dissemination of scholarship and can provide a significant burden on the academic
careers of authors.
Using a stratified random sample we studied average publishing delays in 2700 papers
published in 135 journals sampled from the Scopus citation index.  The shortest overall
delays occur in science technology and medical (STM) fields and the longest in social
science, arts/humanities and business/economics. Business/economics with a delay of 18
months took twice as long as chemistry with a 9 month average delay.  Analysis of the
variance indicated that by far the largest amount of variance in the time between submission
and acceptance was among articles within a journal as compared with journals, disciplines
or the size of the journal.  For the time between acceptance and publication most of the variation
in delay can be accounted for by differences between specific journals. 

Now it's time to put two and two together (and this pertains more to the lag between 
acceptance and publication: the timing of peer review and revision is another matter):

1. The research community is clamoring for access, particularly those who are denied
access to articles in journals to which their institutions cannot afford to subscribe.

2. In many fields, the most important growth region for taking up and building upon new
findings, hence research progress, is within the first year of publication.

3. The average delay from acceptance to publication for subscription journals is about
6 months (and especially long for arts & humanities journals)

4. Björk and Solomon point out that for Gold OA journals the delay is much shorter:
under 2 months.

5. The delay for Green OA self-archiving is even shorter: zero if self-archiving
is immediate (and even negative if a pre-refereeing preprint has also been made
OA even earlier).

6. Subscription journals say they are in favor of OA, but they need an embargo in order 
to keep their subscriptions sustainable.

7. Subscription journals already have a built-in "embargo" because of the 6-month
delay between acceptance and publication.

8. So the 6-12-month Green OA embargoes demanded by STEM fields and even 
longer embargoes demanded by arts & humanities journals not only impedes research 
progress by denying access during the embargo, but they compound the publisher-end
delays between acceptance and publication.

This is why the Liege-model immediate-deposit mandate ( together with the 
repository-mediated copy-request Button) -- now recommended by
both HEFCE and BIS -- is so important:

It makes it possible for researchers to request -- and authors to provide -- immediate
access with one click each as soon as the final, refereed, revised draft is accepted for 
publication, irrespective of publication lags or publisher OA embargoes.

Stevan Harnad