Academic publishing has a peculiar economic structure. Researchers produce the underlying work. Many peer reviewers and academic editors contribute without a direct fee from the journal. Yet the businesses that organise, distribute and analyse that literature can be highly profitable.
It is tempting to conclude that the whole industry is simply charging for research it receives for free. That description catches an important tension, but it misses the real variety of products, customers and costs in today's scholarly publishing market.
To understand where the money comes from, follow one article from submission to discovery—and then look beyond the article itself.
Four businesses often bundled into one label
A scholarly publisher may operate subscription journals, open-access journals supported by article processing charges, books and reference works, and software or data services. Some groups also provide workflow tools for researchers, research offices, libraries and universities.
An article can therefore support several transactions. A library may buy access to a collection of journals. A funder may cover an open-access publishing fee. A research organisation may purchase analytics. A university may license a platform that helps users discover or manage the literature.
Revenue from these activities should not be added together and described as 'journal profit'. Annual reports are structured around corporate divisions and accounting definitions, not necessarily around the cost of publishing an individual paper.
For a concrete illustration, RELX reported £2.714 billion of revenue and £1.035 billion of adjusted operating profit in its Scientific, Technical & Medical segment for 2025. Those are reported segment figures, not the income and profit of journal publishing alone. The division includes databases, tools and electronic reference products as well as primary research publishing.
Informa's Taylor & Francis business, meanwhile, reported continuing subscription activity alongside growth in open-access publishing, and also discussed revenue from agreements involving access to content for AI-related purposes. The market is no longer adequately described by 'subscriptions versus APCs'.
Why subscriptions survived the internet
Moving scholarly literature online did not remove the need for licensing, reliable access, indexing, preservation and institutional workflows. Large publishers typically negotiate digital-access agreements with universities and library consortia. Some arrangements combine reading access with support for open-access publishing.
The bundle is convenient for institutions that need thousands of titles, but it can make the price of a single journal difficult to isolate. A library may value the breadth of access while questioning whether the package represents good value for frequently used and rarely used titles alike.
For authors, subscription income is mostly invisible. Their paper appears online, readers at subscribing institutions have access and the financial negotiation occurs elsewhere. That separation between producers, readers and purchasers is one reason the economics of scholarly publishing often feel unfamiliar even to experienced researchers.
APCs changed the payer, not the need for governance
In an article-processing-charge model, the author, funder or institution generally pays for publication once the manuscript has been accepted under the journal's rules. Readers then receive access according to the publication's licensing policy.
The attraction is obvious: scholarship can become available without a reader-facing subscription barrier. But an APC may be difficult for researchers who lack grant support, institutional agreements or access to waivers.
There is also a conflict that requires careful management. If the publisher earns more when more papers are accepted, the editorial decision must be insulated from commercial targets. Clear rejection criteria, independent editors, audited workflows and transparent appeals matter more than advertising claims about quality.
Not every open-access journal charges APCs. Diamond open access is organised so that neither authors nor readers face publication or access fees, usually through institutional, public or community support. It proves that 'open access' is an access principle, not one universal business model.
Who ultimately carries the publishing bill?
A useful way to understand the market is to follow the payer, not only the published paper. Under a subscription arrangement, institutions and libraries often pay for reading access. Under an APC model, authors, their universities or their funders pay to make the final article openly available. In a society-supported or diamond journal, membership income, university infrastructure or public funding can cover costs that remain invisible to authors.
A single institution can appear in several roles: it employs the researcher, buys database access, funds article charges and provides an editorial volunteer. Those expenditures may be held in different departmental budgets, which makes the overall cost hard to see. Shifting from subscription payments to APCs might move money from a library to a research office without reducing the total amount spent.
This is why universities evaluating a publishing agreement should ask whether the contract changes costs overall or merely relocates them. They should examine author eligibility, access rights, publishing quotas, unused capacity and the treatment of researchers who do not belong to participating institutions. The same scrutiny helps publishers design transparent products instead of advertising one component as if it represented the entire cost.
The costs that readers do not see
A professional journal needs manuscript handling, editorial support, production, DOI registration and metadata maintenance, hosting, preservation, customer service and legal or administrative functions. Research-integrity investigations can involve substantial time and specialist expertise. A journal that publishes fewer than ten papers a month may still need a functioning editorial office throughout the year.
Some costs scale with article volume: typesetting, XML conversion, copyediting and certain production services. Others are relatively fixed: platform subscriptions, governance, staff oversight and infrastructure. A publisher with hundreds of titles can spread those fixed costs; a one-journal society may find them harder to absorb.
When readers ask why a fee is high, the useful answer is not 'peer review costs money' or 'reviewers work for free'. It is an honest explanation of which services are included, how the work is funded and whether the journal discloses its charges and waivers before submission.
Why a profit margin needs context
The arithmetic looks simple: revenue minus costs equals profit. The accounting is not. Companies classify shared technology, central administration, acquisitions, amortisation and restructuring differently. Adjusted operating profit is not identical to reported net profit. A group-level margin should not be presented as a journal-level margin.
Consider a fictional journal receiving 300 submissions, publishing 90 articles and charging an average of $1,000 on 70 paid publications. Its gross APC revenue would be $70,000 before waivers, refunds, taxes and costs. If its editorial and production operation costs $65,000 annually, it has little room for contingencies. If costs are $35,000, the picture is very different. Neither case establishes what a real publisher earns.
The same caution applies when comparing journals that outsource almost everything with those that employ their own staff. What appears to be a low unit cost may omit volunteer labour or work financed by a university.
The power of portfolios and platforms
Large publishers can distribute expensive technology across many journals. They may also have established discovery traffic, institutional relationships and long-term contracts that a new entrant must build from scratch. Those advantages can be commercially defensible while still raising questions about market concentration and institutional bargaining power.
Digital platforms offer another opportunity: services that help researchers search, evaluate, annotate, measure and manage scholarship. Those products can diversify revenue beyond charging to read or publish a paper.
But the commercial opportunity creates a governance challenge. The companies controlling extensive research content and metadata may also supply the tools institutions use to evaluate research. Universities should understand what data is collected, how it is licensed, and whether analytic products create dependencies that are difficult to unwind.
What transparency would look like
The public does not need every publisher to reveal confidential client contracts. It is reasonable, however, to expect clear APC schedules, waiver policies, copyright terms, publication criteria, ownership information and explanations of additional author charges. The publishing transparency principles maintained by COPE, DOAJ, OASPA and WAME make these basic practices explicit.
Institutions buying access can ask for evidence about usage, preservation, accessibility and the services included. Research funders can examine whether their support improves access and quality or simply shifts costs between budgets. Journal owners can evaluate long-term financial health separately from short-term growth in accepted papers.
There is also a question that receives too little attention: what exactly constitutes value to the researcher? Publication speed matters, but so do fair editorial handling, careful corrections, credible archiving and the ability for others to discover and reuse the work years later.
What comes next
Academic publishing is unlikely to stop being a commercial market. The more consequential shift may be in what organisations pay for: not just access to static articles, but dependable editorial processes, better metadata, research-integrity services, discovery systems and increasingly sophisticated tools built around the scholarly record.
That future presents opportunities for both large and independent publishers. Yet it does not settle the central public-interest question. Scholarship is created to advance knowledge; commercial models should help distribute and preserve it without making scientific quality subordinate to sales targets.
The right way to scrutinise publishing profits is to examine audited financial disclosures, distinguish business segments and ask which parts of the value chain are being funded. Anything simpler makes a good slogan, but a poor guide to how the industry actually works.
