The Publisher's Association's impact assessment on OA is pretty much as you'd expect

Originally published on eve.gd on 17 February 2021, last updated 6 September 2026.

Read this post on the web (with images, PDF and citation details)

The Publisher’s Association has commissioned a report that seems to be their latest attempt at painting open access to research as economically damaging to the publishing sector. It shamelessly plays on the rhetoric of “UK PLC” in an attempt to appeal to government politicians but doesn’t mention the benefits to society of having research openly available or the savings to the public purse of publicly funded institutions of higher education not paying billions of £s in subscriptions. However, the conclusions therein are, sadly, if predictably, extremely poor distortions of the real situation in my view.

commissioned a report

Let’s have a look at this document:

• Point 3.15: “We estimate that the loss of subscription revenue to UK journals resulting directly from the UKRI Policy would be in the order of GBP 151 million per annum or GBP 1.0 billion in the period from 2022 to 2027”. Section 7 shows how they have come to this figure. In 7.43 they estimate that 14-17% of articles by UK authors are available for free online. I’d say it’s actually more like 99% given Sci-Hub (as discussed in point 7.55), but another point here is that they use the figure from 2016, early in the REF cycle and before the policy kicked in. Deposit rates will be much higher than this now. Anyway, they use this to estimate that just 8% of UK research is available in green OA with an embargo at the moment (7.45); a huge under-estimate that is then used to justify why there haven’t been cancellations so far. The scenario they end up using as an estimate is “that 50% of all research articles published by UK authors in subscription-based journals are posted in online repositories as AMs or VORs” (7.65).
• Point 7.50: “If all articles produced by UK researchers were published in UK journals, UK researchers would account for just under 30% of content. We therefore assume that 2/3 of UK authored articles are published in UK journals while the remaining 1/3 are published abroad”. How can you even justify this as a hypothetical assumption? What is that “therefore” doing? They write that “[t]his implies that, as of 2019, approximately GBP 315 million of subscription revenue for UK journals was associated with content produced by UK authors”. Baffling.
• Points 7.57 + 7.58: these basically say: there’s no evidence of cancellations to date due to green OA. However, there might be in future. This argument has been going for a decade. It’s hardly a minor point, though. The basic supposition beneath this entire report is that libraries cancel due to green OA. Yet the evidence to show that this happens is not actually there.
• Point 7.62: “the UKRI Policy will also give greater legitimacy to certain postings of research articles that would have been in breach of journal policies previously”. Hmmm.
• Point 7.63: “There is also a risk that the policy will set a new expectation among researchers that all research articles should be posted in online repositories upon publication”. Yes. This is not a risk. This is the expected behaviour and intent.
• But wait. So how do they calculate, from that, the portion of revenue that is lost (for UK publications)? They take a figure of £1,576m for UK journal revenue association with subscriptions. They then estimate 20% of that comes from UK authors = £315m. They then estimate that 50% or so of UK authors deposit. They come to a figure for a loss of £151m per year. But then, in bullets 7.67 and 7.68 come up with a figure of £173m per year or £1bn over 2022-2027. They then, again, increase this to £292m in bullet 7.72. In the summary (2.2) they use the figure of £3.2bn for “associated loss” from other industries based on a figure of a 1.6x multiplier in 3.17. This feels like an exercise in seeing how far you can stretch these numbers.
• Why is this wrong in any case in my view? Because libraries subscribe to whole journals, not to articles. It’s not as though libraries can cancel journals that have articles from authors outside the UK that are not deposited. So when they write in point 7.85 that “UK libraries would no longer expect to pay subscription fees for those research articles that had been published through Gold OA” there’s a problem. They can’t choose not to pay for articles, only titles, and sometimes, even, big deal bundles of journals.
• Around points 7.83 there are some mentions of the uneven economic effects of a scaling of gold OA via APCs: “Research intensive universities will need to cover a disproportionately large share of this cost”. This is used to argue that it would be unfair to pursue this route. However, this is an incorrect premise. The answer is not: “APCs are unaffordable when we scale because the distribution is bad”. The answer is: “we need business models that distribute the costs of gold OA better”. There is no mention of any model for gold OA except APCs.
• There’s a whole scary section already warning of the dangers to monographs around section 8.4. It really is “sky is falling” kind of stuff and they attempt to play on researchers’ fears by saying how it will limit author’s opportunities to write academic books. In case the Publisher’s Association hadn’t noticed, though, monographs are screwed already, as it stands (hinted at in bullet 8.24). Their economics have been in decline for decades. Yet, again, they assume that BPCs are the only way that this could be funded and so say it’s not possible. Yet I am showing how university presses can transition to an OA model without BPC fees. Further, there are many scholarled publishers as examples that do not use a BPC model. There’s not even a mention of Knowledge Unlatched. This report gives just “author pays” or “green OA” as the routes to OA monographs. It’s really myopic. The PA here seems to be arguing for stasis and are not willing even to experiment with new models. They then present it as “impossible” because they won’t change. The basic truth of the matter in my view is that many publishers have not even tried to change their models.
• 8.8: “Monographs are particularly valued by researchers for their accuracy and quality, which is ensured through an extensive peer-review and editorial process”. Actually, there has been extensive discussion at policy meetings about how bad the peer-review processes are for many monographs – including at “top” university presses.

I am showing how university presses can transition to an OA model without BPC fees

All in all, these figures do not, to me, seem credible – it just feels like the PA, yet again, trying to slow down OA progress. Doing so in a year when the importance of open research has been shown to be one of the most important things on the planet as we all benefit from, say, coronavirus science, is hardly helpful. We’re not going back and the better course would be for publishers to begin to adopt new, feasible business models for OA, rather than pretending that things will return to the “old normal”.

References

This reference list has been generated, in many cases, through an automated process and may therefore contain errors or incomplete metadata. It is provided purely as assistance for finding the items mentioned within this post.

1. ‘Economic Impact Assessment’, Publishers Association, 17 February 2021 <https://www.publishers.org.uk/publications/economic-impact-assessment/>

2. Opening the Future <https://www.openingthefuture.net/>

https://www.publishers.org.uk/publications/economic-impact-assessment/
https://www.openingthefuture.net/