The time taken from receipt of a submission to publication is today frequently used as a ‘key performance indicator’ by academic journals. It is a (partial) measure of the speed or efficiency of the journal’s editorial and production processes (though also highly dependent on the author’s approach to revisions and proofs). The Royal Society has been recording and reporting this metric since the early 1950s, which allows us to produce the graph below:
For the period 1949 to 1986, we have data separately for the Society’s four journals. Articles in the two series of theTransactions typically took longer than those in the Proceedings, presumably because articles for the Transactions were longer than those for the Proceedings, and usually went through a stricter refereeing process (potentially involving more referees, and more author revisions).
Although the Royal Society’s staff claimed to care about speed of publication, the strong impression from the archive is that they were relatively powerless to do anything about it in the pre-1990 period. Referees were often blamed for being slow. Authors were sometimes blamed for being slow to make revisions. The printers and publishers (Cambridge University Press) were often blamed; in return, the Press pointed out that, if the Society would only commit to a regular publication day (on the same day each month), then the Press could plan the Society’s work into the printing schedule, rather than having to fit it in around other scheduled work. The move to a regular schedule for the Proceedings journals in 1982 did help in the short-term.
The improvement in editorial and production times since 2000 reflects the dramatic change in workflows in academic publishing. These include the use of author-generated electronic text (since the mid-1990s), which could feed into computerised typesetting processes (which had existed since the late 1970s, but had initially required re-keying); the introduction of editorial management software, and a close attention to staff efficiency (a new KPI); and digital printing technologies and online publishing.
Where do the data come from?
From 1951 to 1979, the average ‘time taken from receipt to publication’ for the previous two years was printed in the Society’s Year Books (and in the 1960s, the quickest and slowest times were also recorded, as well as the average times for various parts of the process). From 1979-81, these data appeared instead in the Annual Report. The data continued to be gathered until 1986, but were no longer made public.
The more modern data series begins in 1997. It comes from a spreadsheet maintained by Stuart Taylor, Director of Publishing. This data differs from the historic series because it relates only to the journals that are now defined as ‘research journals’ (and therefore, it excludes Transactions A and Transactions B). The modern data should be treated as a continuation of the data for Proceedings A and Proceedings B, but it does also include the new journals founded after 2003.