An article processing charge, or APC, is the fee a journal bills to make a paper open access — free for anyone to read. The amounts vary wildly: fully open-access publishers list article processing charges from a few hundred dollars at society and regional outlets to more than $11,000 at premium Nature titles, while hybrid journals charge APCs on top of subscription revenue libraries already pay. Understanding where that money goes, and who ends up paying it, is now unavoidable for researchers, because funders including the US National Institutes of Health and the cOAlition S consortium require open access publication as a condition of grants.
This is an explainer on publishing economics, not financial advice; pricing and policy details change and should be checked with each publisher.
Where does the money go?
Publishers describe APCs as covering the real costs of running a journal: manuscript submission systems, plagiarism screening, coordinating peer review, copyediting, typesetting, hosting, long-term archiving, and compliance reporting. Industry analyses submitted to the UK's Investigatory Powers of universities — specifically the 2017 Wellcome-supported competition study and subsequent publisher disclosures — put the marginal cost of handling one article in the low hundreds of dollars, while list APCs at large commercial publishers can exceed that many times over. The gap funds what publishers call value-added services: brand, discoverability, quality signaling, and profit. Major publishers including Elsevier, Springer Nature, and Wiley report profit margins in the high teens to high thirties on their publishing arms, margins few industries match.
Who actually pays APCs?
In principle, authors do. In practice, the money flows from research grants, institutional open-access funds, and, increasingly, transformative agreements in which universities pay a lump sum covering both subscriptions and publication fees. This creates the equity problem critics highlight most: a 2023 study in PLOS ONE and earlier analyses found APC prices correlate with journal prestige, and researchers without grant support — early-career scholars, those in lower-income countries, independent researchers — either pay out of pocket or publish in cheaper venues, which then gets read as lower quality. Waiver programs at fully OA publishers blunt but do not eliminate the gap.
Why did the system move to author-pays?
Open access advocacy from the 2000s argued that publicly funded research should be readable by the public, and subscriptions made that impossible. The switch kept the revenue model familiar — publishers simply relocated the charge from libraries to the article itself. The transition created a hybrid distortion: subscription journals collecting both library fees and APCs on a fraction of articles, a period cOAlition S explicitly targeted when its Plan S principles barred funded authors from hybrid venues from 2021. As of 2025, coalition funders still require immediate open access, though a 2024 policy revision relaxed some mandates around rights retention.
Related stories: Transformative Agreements: One Contract for Reading and Publishing · Predatory Journals: Nine Red Flags Before You Submit a Paper.
What is the total bill, and is it falling?
Global spending on APCs is estimated in the low billions of dollars annually. Studies modeling the transition — including work from the Max Planck Digital Library's Open Access 2020 initiative — argue total system cost should fall only when subscription payments are wound down as APC coverage expands. That has not clearly happened yet: library spending analyses through 2024 show both subscription and APC lines rising in many institutions, the double-dip scenario critics predicted.
What does an institution pay per article under the deals?
Transformative agreements publish their arithmetic, which makes them the clearest window into system costs. Contract disclosures analyzed by open-access trackers through 2024 put the effective per-article price under major European and North American deals between roughly $2,500 and $9,000 depending on publisher and negotiated volume, consistent with list APCs and in some cases above them. The number matters because it converts an abstract industry debate into a line item universities can compare against what they paid in subscriptions alone. When library economists run those comparisons, the recurring finding is that total spending has not yet fallen for most institutions; the open-access share of institutional output rose, which is a real gain, but the invoice grew with it rather than shrinking.
Do APCs change what gets published?
The uncomfortable question is whether charging authors at acceptance tilts editorial judgment. Fully open-access mega-journals, which publish on scientific soundness rather than perceived importance, depend entirely on APC revenue, and critics have long noted the structural oddity: revenue rises with acceptance volume. Evidence that this bends individual decisions is limited, since internal acceptance data are rarely public, but industry conduct is suggestive. Investigations through 2024 documented special-issue pipelines where acceptance standards slipped and revenue ran ahead of quality control. The lesson generalizes: any fee tied to publication needs published acceptance data, transparent pricing, and independent audit to stay honest, and the journals that publish such data are the exceptions proving the rule.
What would change the picture?
Alternatives are live experiments: subscribe-to-open models, in which libraries collectively fund a journal and everyone gets open access if targets are met (used by Annual Reviews and some university presses); institutional publishing platforms such as the universities behind the arXiv and Diamond OA movements, which run journals at cost; and Diamond open access venues with no fees at either end, now covering an estimated 8–10% of articles. The test to watch is whether libraries' total spend per published article falls. Until it does, the APC system remains a cost relocation, not a cost reduction.
