How many times do we have to pay for the same piece of research? #

Look at Sweden. Public money keeps the whole academic show on the road: roughly 186.7 billion SEK a year goes into R&D, with the higher-education sector getting a quarter of that (about 23%) and new government bills adding billions more for labs and salaries.¹

Researchers then race to publish in “high-impact” journals² owned by the big five (Taylor & Francis, Elsevier, Wiley, etc.). Most of those titles sit behind pay-walls, so our publicly funded universities have to cut another cheque to read the articles their own staff just wrote. Stockholm University’s situation in 2022 is depressingly typical: 26.6 million SEK for open-access publishing + 13.4 million SEK just to read e-journals.

Yes, Sweden’s Bibsam consortium has negotiated “read & publish” deals that bundle subscriptions with article-processing charges (APCs).³ But that’s still taxpayer money—just moved from one column to another—and APCs hover around a couple of thousand euros per paper. Critics call the whole thing double dipping⁴ because publishers pocket subscription fees and APCs to unlock a single article in that same journal.

And remember: the peer-review system runs almost entirely on unpaid academic labour. So what exactly are we buying here—apart from fancy logos and a prestige badge?

It gets worse #

In 2024 Taylor & Francis quietly sold non-exclusive access to the full text of nearly 3,000 journals to Microsoft for $10 million USD up front and projected AI revenues of > $75 million USD.⁵ Translation: the publisher monetised the research we financed again—this time to train large-language models—so they can spin up new pay-walled “AI discovery” tools that our libraries will be invited to buy back at a premium.

So let’s count the bill:

1. Funding the science. Taxes pay researchers, grants, equipment, infrastructure.
2. Buying the paper. Taxes pay big-deal subscriptions so we can read what we already paid to produce.
3. “Open” access. Taxes (or personal grants) pick up APCs so the same publishers will kindly let us share the PDF.
4. AI remix. Publishers license that corpus to Big Tech, and universities will soon be invoiced for the shiny AI interface built from it.

Four bites at the same apple—and peer reviewers still get zero.

We could stop the madness by insisting on community-owned journals, pre-prints plus sensible quality control, or at the very least by refusing to hand over exclusive rights to taxpayer-funded manuscripts. Until then, every krona we spend just keeps the pay-wall-to-APC-to-AI wheel spinning.