The Two Billion Dollar Toll

Consider a market with the following structure.

The core product is manufactured entirely by the customer, at their own expense, over months or years. The quality control is also performed by the customer, unpaid, as a professional obligation. The finished product is then handed to the intermediary at no cost. And then the customer pays the intermediary, sometimes thousands of dollars, for the privilege of having it distributed.

No functioning market in the world economy is built this way. Scientific publishing is, and almost nobody inside it stops to examine that, because we are all too busy producing the product.

In the last few years, the price of that arrangement has gone vertical.

$910M → $2.54B

An article processing charge, or APC, is what you pay to make your paper open access. Funders increasingly require it. Institutions increasingly expect it. And the price is set by the publisher, who faces very little competitive pressure, because you cannot substitute one journal for another the way you substitute one supplier for another. The journal you need is the journal your field reads. That is not a market with elastic demand. That is a toll road.

Here is what the toll actually costs in 2026.

Now the detail that should genuinely annoy you as an operator. Hybrid journals, meaning subscription titles that sell open access one article at a time, charge on average roughly one and a half times what fully open access journals charge. Same peer review. Same copy editing. Same distribution infrastructure. Roughly a fifty percent premium, because the pricing reflects what the market will bear rather than what the service costs to deliver.

You are the factory, the quality control department, and the customer. You are the only participant in this market who pays at every stage and captures value at none of them.

That is the economics. Here is the consequence, and it is the part that matters more than the numbers.

A charge of three thousand dollars is invisible to a researcher at an institution with a transformative agreement, where the library has prepaid and the fee never appears. That same charge is prohibitive for a researcher in Athens, Cairo, Lagos, or Manila with an equally important finding and no institutional cover. Both submit. Only one can afford to publish where it belongs.

The evidence base that medicine is built on is therefore shaped not only by who produces the best data, but partly by who can pay the toll to have it seen. That is a structural filter operating quietly on the global scientific record, and it is dressed up as an invoice rather than a policy.

I am not going to pretend an individual researcher can fix that. But you can absolutely stop being naive about it, and most people are, in ways that cost them real money and real placement.

If you are choosing where to send something now, it is worth working through the fit and the cost together rather than sequentially. AskDimitri can help you weigh journal choice against reach and expense before you commit.

Weigh your journal choice →

Every profession eventually has to understand the economics of the system it works inside. Clinicians learned to read a payer contract. Investors learned to read a fee structure. Researchers, for the most part, still have not learned to read a price list, and it is costing us more than money.

Editor's note: in September 2026 the figures in this letter were checked against the sources listed below, and the text was corrected where the letter as sent differed from them.