Academic publications are expected to rise by more than 500,000 this year, to 6 million – creating headaches for knowledge custodians (as well as readers).
China overtook the US in scientific research output from around 2018 – which coincidentally is also the year that Australia gave up on a national system to measure our research quality against the rest of the world (the Excellence in Research Australia scheme).
Last year, China become the first nation to surpass 1 million academic publications in a year and has surpassed the output of the entire OECD in some disciplines.
The global rise in publication output is attributed to the rapid growth in publications from China and India, increased access and an increased use of AI to assist publication development and production.
The feast of knowledge is not without consequence, with peer review straining to keep up and concerns about quality – including AI-written peer review and the fact that while Chinese publication rates are much higher than other nations, so are retraction rates.
This presents a significant problem for ANZ university libraries, which have already negotiated open access deals with giant for-profit publishers. The growth in publications makes them out of date.
Over five years, the Council of Australasian University Librarians negotiated better terms with journal publishers large and small – but most importantly large. The new deal is that free-to-read rights are built into the cost of journal subscription bundles.
Getting it done took an heroic effort, given the complexity of separate negotiations with the five big commercial publishers who traditionally made their money through subscription costs, now supplemented by free-to-read-but-pay-to-publish options.
CAUL reports that across 2020-25, members saved $473m on list-prices for article processing charges, with nearly 94,000 articles by ANZ university authors appearing free to read, under capped open access agreements.
But it was too good to last and the number of articles published is now exceeding the agreed caps, “in several cases outpacing 2026 estimates set only months early.”
“When volume assumptions break down mid-term, pricing models built on last year’s data are out of date before the next agreement is signed,” the Council warns.
And it is libraries that will cop costs they cannot control, with publishers producing as much content as they want and way more available due to AI. CAUL reports there will be 6m research articles this year, up 500,000 on ’25.
And so the Council asked four (un-named) of the Big Five what they are going to about three issues:
- More content on offer: publishers are accepting fewer submissions, down from a third to a quarter as a median. The problem is “selectivity is rising, but volume is rising faster”
- More articles accepted is not their fault: “growth was consistently attributed to the research community rather than to publisher strategy, supported by figures on the expanding global researcher population”
- Growth occurs to meet author demand”: “serving researcher choice and editorial merit.”
CAUL’s take-out: “no publisher offered a concrete mechanism to cap cost growth in step with volume growth … falling acceptance rates are evidence of effort on quality; they are not evidence of cost control.”
While publication output is rising and open access makes them more easily accessible, readership does not appear to be rising at the same rate. More than half of Australian research from 2020 was cited five times or fewer in the following five years according to UoM research Matthew Pinnuck.