The Knowledge Toll — FSA Enclosure Series · Post 3 of 4

The Knowledge Toll  ·  FSA Enclosure Series Post 3 of 4

The Knowledge Toll

How Academic Publishers Captured the Distribution of Publicly Funded Science — and What It Costs Everyone Else

The Open Counter

arXiv has hosted physics and mathematics preprints since 1991 — free, immediate, no paywall, no embargo, no Article Processing Charge. Its annual budget is approximately $2 million. Elsevier's STM segment generates over $3 billion in annual revenue at a 34.8% margin. The counter-architecture has existed for thirty-three years. This post documents what it is, what the NIH zero-embargo mandate adds to it as of July 2025, and why the impact factor lock remains the single obstacle preventing the counter-architecture from replacing the toll entirely.

Randy Gipe  ·  Claude / Anthropic  ·  2026  ·  Trium Publishing House Limited  ·  Forensic System Architecture

In August 1991, Paul Ginsparg — then a physicist at Los Alamos National Laboratory — set up an email-based system for distributing preprints of physics papers to colleagues. The system allowed researchers to deposit manuscripts, share them with the community before formal publication, and access each other's work without institutional subscription. Within months, thousands of papers were being distributed. Within years, it had become the primary means through which high-energy physicists shared findings — faster than peer review, faster than journal publication, faster than library subscription processing.

The system became arXiv. Today it hosts over two million papers across physics, mathematics, computer science, quantitative biology, and related fields. Its annual operating budget is approximately $2 million. Cornell University and the Simons Foundation provide the majority of its funding. It charges authors nothing to deposit. It charges readers nothing to access. It employs no peer reviewers at commercial salaries, pays no editors a commercial rate, and has no shareholders.

Elsevier's STM segment generates over $3 billion in annual revenue at a 34.8% operating margin. Both systems distribute scientific knowledge. One costs $2 million to operate. The other generates over $1 billion in annual profit. The difference between those two numbers is not the cost of distribution. It is the toll — and the toll has been operating alongside a functional alternative for thirty-three years.

"arXiv operates on $2 million per year. Elsevier's STM segment generates over $3 billion at a 34.8% margin. Both distribute scientific knowledge. The difference is not the cost of distribution. It is the toll — and the toll has operated alongside a functional alternative for thirty-three years without being displaced." FSA Analysis · Post 3

The Counter-Architecture: Four Instruments

Counter-Instrument 1

arXiv — The Original Bypass

Founded 1991 at Los Alamos; now operated by Cornell University. Over 2 million preprints across physics, mathematics, computer science, quantitative biology, economics. Zero cost to deposit; zero cost to access. Annual budget approximately $2 million. Standard dissemination channel in physics and mathematics — papers are read, cited, and built upon from arXiv before and often instead of their journal versions. The preprint server does not replace peer review for credentialing purposes, but it demonstrates unambiguously that distribution of scientific knowledge requires neither a commercial publisher nor a paywall.

Counter-Instrument 2

bioRxiv and medRxiv — Life Sciences and Medicine Join

bioRxiv launched 2013 at Cold Spring Harbor Laboratory; medRxiv in 2019 (Cold Spring Harbor, Yale, BMJ). The COVID-19 pandemic provided the most significant test of preprint infrastructure in the system's history: within weeks of the outbreak, preprints on viral structure, transmission, and therapeutic candidates were distributed globally and informing public health decisions before any peer-reviewed publication completed review. The pandemic demonstrated that preprint distribution at global scale is technically straightforward, institutionally manageable, and scientifically valuable — and that the commercial journal system's timeline is a commercial convention, not a scientific necessity.

Counter-Instrument 3

Plan S and cOAlition S — The Funder Mandate

Launched September 2018 by approximately 28 European research funders. Core requirement: immediate open access with zero embargo and author rights retention (CC BY license by default) for all peer-reviewed articles from funded research. Transformative arrangements — the industry's revenue-preservation-as-reform strategy — lost cOAlition S support at end of 2024 after explicitly failing to produce genuine transition. The 2026–2030 strategy pivots to diamond open access, where neither authors nor readers pay, and to preprint infrastructure. Plan S is the first major coordinated funder refusal to subsidize the commercial publishing toll through grant funding.

Counter-Instrument 4

NIH Zero-Embargo Mandate — The Scale Instrument

Effective July 1, 2025. Every peer-reviewed article resulting from NIH funding — accepted on or after that date — must be deposited in PubMed Central and made publicly available with zero embargo on the official publication date. The Government Use License embedded in NIH funding agreements gives the federal government an irrevocable right to make the Author Accepted Manuscript available, overriding conflicting publisher agreements. NIH funds approximately $47 billion in research annually. No APC is required for compliance — authors deposit accepted manuscripts directly. The publisher's version of record remains behind its paywall; the NIH-deposited manuscript is freely accessible simultaneously on the same day.

The NIH Mandate: What Changed on July 1, 2025

The NIH's 2008 public access policy required deposit in PubMed Central within twelve months of publication — a twelve-month embargo that gave publishers a year of exclusive subscription revenue before the paper became freely accessible. The 2025 policy eliminated that embargo entirely.

The mechanism is the Government Use License: a provision in NIH grant funding agreements giving the federal government an irrevocable, royalty-free right to make the Author Accepted Manuscript publicly available. Because this right is embedded in the grant agreement — not negotiated with the publisher after acceptance — it supersedes conflicting publisher copyright terms. An NIH-funded author who transfers copyright to Elsevier cannot waive the government's pre-existing right. The publisher's version of record remains behind its paywall; the accepted manuscript is freely accessible simultaneously.

PubMed Central is not a preprint server. It hosts accepted, peer-reviewed manuscripts — the same scientific content as the paywalled journal version — made available at no cost on the same day the journal version publishes. A researcher at a community college, an independent practitioner, a scientist in a developing country institution without full journal subscriptions, can now obtain every NIH-funded paper through PMC on its publication day. No interlibrary loan. No 30-minute library terminal time limit. No sticky note directing them to another route.

"The NIH zero-embargo mandate makes every NIH-funded paper freely available in PubMed Central on its publication date. The same peer-reviewed content behind Elsevier's paywall is simultaneously available, at no cost, to anyone in the world with an internet connection. The toll remains. The bypass now has a federal mandate behind it." FSA Analysis · Post 3

The Impact Factor Lock: Why the Counter-Architecture Hasn't Won

If arXiv has existed since 1991, if preprint servers host millions of papers, if the NIH now mandates free access to its funded research — why does Elsevier's operating margin remain 34.8%? Why do universities continue paying billions in subscription fees and APCs?

The answer is the impact factor lock. The career incentive that keeps scientists submitting to commercial journals is not irrational. It is structurally correct given the evaluation systems their employers and funders use. A physicist at a research university needs publications in high-impact journals to be hired, to receive tenure, to win grants, to be promoted. The journals with the highest impact factors in most fields are owned by commercial publishers. Depositing on arXiv is standard practice and professionally unremarkable. Publishing in an arXiv-only preprint is professionally damaging. The preprint demonstrates the science. The journal publication certifies the career. Those are two different things, and the commercial publishers own the certification system.

The impact factor is produced by Clarivate, a private company, using a proprietary methodology widely criticized by researchers and professional societies. It was designed as a tool for libraries to evaluate journal subscriptions. Through decades of adoption by hiring committees, tenure review bodies, and grant panels — adoption that publishers encouraged at every opportunity — it became the de facto standard for evaluating individual scientific careers.

Reforming the impact factor lock requires changing the behavior of thousands of hiring committees, tenure review bodies, and grant panels simultaneously. Every individual actor is rational to continue using impact factor as a criterion as long as every other actor does. The coordination problem is the architecture's deepest defense — deeper than the Big Deal, deeper than the NDA, deeper than the APC. It is the reason thirty-three years of preprint infrastructure have not displaced the toll.

"The preprint demonstrates the science. The journal publication certifies the career. Those are two different things, and the commercial publishers own the certification system. That single distinction — not the distribution channel, not the paywall, not the bundle — is what has sustained the toll against thirty-three years of open counter-infrastructure." FSA Analysis · Post 3

Diamond Open Access: The Iowa Moment

The counter-architecture has an Iowa moment. Diamond open access is a publishing model in which both authors and readers pay nothing. Publication costs are borne by funders, institutions, or disciplinary societies. The journal charges no APCs. It charges no subscription fees. It publishes peer-reviewed research freely, immediately, and permanently.

Diamond OA journals operate across virtually every scientific discipline. The Public Library of Science — PLOS — has published peer-reviewed research since 2003 with APCs subsidized or waived for researchers without funding. European Open Science Cloud and national open science initiatives in Germany, the Netherlands, and Scandinavia fund diamond OA infrastructure at the institutional level. The cOAlition S 2026–2030 strategy explicitly prioritizes diamond OA as its primary model going forward — explicitly because it eliminates the APC double-dipping problem and cannot be captured as a revenue stream by commercial publishers.

Diamond OA is not a theoretical model. It operates. It publishes rigorous peer-reviewed science. It does so without a 34.8% operating margin flowing to shareholders. The obstacle is identical to Iowa's: a functioning, lower-cost, publicly beneficial alternative exists and has existed for years. The career incentive lock — the hiring committee that still weights a commercial journal over PLOS, the grant panel that still treats impact factor as quality proxy — is the single mechanism through which the toll maintains position against an alternative that has already proved itself.

What the Counter-Architecture Has Already Won

The NIH twelve-month embargo is gone. Every NIH-funded paper published after July 1, 2025 is freely available in PubMed Central on publication day. This is the largest single expansion of free public access to scientific research in American history. The transformative agreement is effectively dead as a reform instrument — explicitly named and rejected by the funders who were sustaining it. The preprint infrastructure is now standard in physics, mathematics, and computer science, and expanding rapidly in life sciences following the COVID-19 cultural shift.

The toll has not fallen. But the siege infrastructure is more developed than it has ever been. Post 4 examines the lobby that has defended the toll through every reform cycle — and what approximately $3 million in annual federal lobbying buys the industry that produces no knowledge but owns its distribution.

FSA Layer Certification · Post 3 of 4

L1

arXiv — Verified Founded 1991 Los Alamos; Cornell University operation. 2M+ papers; ~$2M annual budget (Cornell, Simons Foundation). Zero deposit/access cost. Standard dissemination in physics and mathematics. Documented in arXiv institutional reports.

L2

bioRxiv / medRxiv — Verified bioRxiv 2013 (Cold Spring Harbor); medRxiv 2019 (Cold Spring Harbor/Yale/BMJ). COVID-19 preprint role in accelerating research dissemination extensively documented. Zero cost to deposit or access.

L3

NIH Zero-Embargo Mandate — Verified NOT-OD-25-101; effective July 1, 2025. AAM deposited in PMC; zero embargo on publication date. Government Use License (2 CFR 200.315) overrides publisher copyright. No APC required for compliance. Replaces 2008 12-month embargo policy.

L4

Plan S / cOAlition S — Verified Launched September 2018; ~28 funders. Transformative arrangements ended December 31, 2024. 2026–2030 strategy: diamond OA and preprint infrastructure (coalition-s.org, November 2025).

L5

Impact Factor Lock — Verified Clarivate / Journal Citation Reports: proprietary metric used as hiring, tenure, promotion, and grant criterion. Publisher dominance of high-impact journals documented. Coordination problem preventing reform identified in science policy literature. DORA commitments exist; institutional behavior change not yet at displacement scale.

Live Nodes · The Knowledge Toll · Post 3

• arXiv: 1991; 2M+ papers; ~$2M/year; Cornell/Simons — arxiv.org
• bioRxiv: 2013 (Cold Spring Harbor); medRxiv: 2019 (Cold Spring Harbor/Yale/BMJ)
• COVID-19 preprint role: cultural shift in life sciences documented
• NIH zero-embargo: effective July 1, 2025; NOT-OD-25-101; PMC same-day deposit
• Government Use License: 2 CFR 200.315; overrides publisher copyright transfers
• NIH annual budget: ~$47B; full output subject to zero-embargo mandate post-July 2025
• Plan S: September 2018; transformative arrangements ended December 2024
• cOAlition S 2026–2030: diamond OA and preprints — November 2025
• Impact factor lock: Clarivate/JCR; career evaluation instrument; coordination problem
• DORA: San Francisco Declaration on Research Assessment — impact factor reform commitments

FSA Wall · Post 3

NIH zero-embargo compliance rates for the post-July 2025 period are not yet published in systematic monitoring data. The NIH enforces through eRA Commons and RPPR reporting; aggregate compliance statistics have not been released as of this writing.

The long-term impact of the zero-embargo mandate on commercial publisher subscription revenue — how much, over what timeline, the PMC bypass will reduce the effective value of library subscriptions — is not calculable from current data. The behavioral responses of libraries, publishers, and researchers have not stabilized.

Whether DORA commitments and similar impact factor reform initiatives will translate into changed hiring and tenure decisions at the scale required to shift the architecture is genuinely uncertain. Individual institutions have adopted DORA; system-wide behavior change is not yet documented at displacement scale. The wall runs at that coordination threshold.

Primary Sources · Post 3

1. arXiv institutional reports — Cornell University; annual statistics, budget (arxiv.org)
2. NIH NOT-OD-25-101 — accelerated effective date July 1, 2025; PMC deposit requirement
3. NIH Public Access Policy Overview — 2024 policy; Government Use License (2 CFR 200.315)
4. cOAlition S — Plan S launch September 2018; transformative arrangement end December 2024; Strategy 2026–2030 (November 2025) — coalition-s.org
5. bioRxiv founding — Cold Spring Harbor Laboratory, 2013; medRxiv 2019
6. PLOS — Public Library of Science; diamond OA model — plos.org
7. San Francisco Declaration on Research Assessment (DORA) — sfdora.org
8. Clarivate / Journal Citation Reports — impact factor methodology
9. Hicks et al., "The Leiden Manifesto for Research Metrics," Nature (2015)
10. Suber, Peter, "Open Access" (MIT Press, 2012)

← Post 2: The Bundle Sub Verbis · Vera Post 4: The Lobby →