SURFACED NIH EMAILS REVEAL SUPERFICIAL COMPREHENSION OF HOW OUTBREAKS BECOME EPIDEMICS AND PANDEMICS
And an instinct to hide information to suppress rational criticsm of public health programs and policies
Francis Collins warned Anthony Fauci that the January 21, 2016 World Economic Forum session titled “Vaccine Innovation for Pandemic Preparedness” looked like a “potential land mine,” according to NIH emails surfaced by Maryanne Demasi and Jeffrey Tucker. The warning was not about a virus. It was about a meeting: a Davos session where the Gates Foundation, Wellcome Trust, Norway, pharmaceutical executives, and U.S. health officials were circling the next model of pandemic preparedness before the world had finished absorbing the lessons of Ebola.
Collins wanted Fauci’s advice because the agenda appeared to involve a proposal for a new vaccine-preparedness partnership. The questions Collins asked were not field-epidemiology questions. They were institutional questions. Was this the model being promoted in the medical literature? Had Gates really bought in? How did BARDA actually work? What position should HHS take? Before the first “Germ Games” email appears, the document already shows the mental operating system: pandemic risk translated immediately into governance, finance, agency turf, philanthropic influence, liability, and platform control.
The later Davos report sharpened the picture. Collins told HHS Secretary Sylvia Burwell that pharmaceutical executives had been blunt. They had spent heavily during Ebola. Their boards had resisted. They could not be expected to do it again for the next outbreak unless governments solved the money and liability questions in advance. They would offer platform capabilities, but they would not surrender commercial intellectual property. Jeremy Farrar then outlined a sketchy nonprofit partnership, with the World Economic Forum serving as interim secretariat.
That is the first buried lesson in the emails: the pandemic-planning class did not begin with transmission. It began with institutions. The outbreak existed as a future trigger for financing, platforms, boards, intellectual property, liability shields, and global coordination. The pathogen had not yet been understood, but the apparatus was already being imagined.
Demasi’s article brought that email trail into public view. The Brownstone companion analysis then connected the international planning culture to the domestic machinery of pandemic governance: the federal plans, emergency-management language, stay-at-home orders, closures, travel restrictions, testing, tracing, countermeasure distribution, and whole-of-government command architecture that later came alive during COVID. Those two reports establish the map. They show the Davos circuitry and the U.S. operational chassis. But the map is not the disease process, and the chassis is not the biology.
The deeper failure exposed by the NIH emails is intellectual before it is institutional. The people preparing for future pandemics understood platforms, finance, liability, command structures, model upgrades, messaging, simulations, and compliance far better than they understood how outbreaks actually become epidemics and pandemics.
That distinction matters because from 2020 through 2022 this same style of model-driven rule helped drive the world toward civic, educational, economic, medical, and constitutional ruin. The catastrophe did not happen because no one had planned. It happened because the planners had mistaken abstraction for comprehension. They treated living societies as programmable systems. They treated models as reality. They treated dissent as a threat to be neutralized. They treated uncertainty as a communications problem. They treated the public as an object of management rather than the sovereign audience to whom science owes the truth.
The surfaced NIH emails therefore deserve a reading deeper than scandal. They show the pre-COVID mindset in formation: outbreaks reduced to scenarios, scenarios reduced to simulations, simulations converted into governance, and governance later imposed on populations as though the assumptions had already been proven. The problem was not preparedness. The problem was superficial comprehension armed with institutional power.
Those two articles therefore establish the map. Demasi shows the elite international circuitry: NIH, WEF, World Bank, Gates, Wellcome, BARDA, WHO, pharmaceutical boards, liability, platforms, simulations, and war-game imagination. Tucker shows the operational chassis: the federal planning documents, the emergency-management vocabulary, the all-of-government cascade, and the civil-liberties machinery waiting to activate again when the next pathogen appears.
I saw incompetence, which is equally scary - if not more.
The pandemic response machinery is not the biology, but it should match the biology, or atrocities without Hitlers can occur. The deeper failure exposed by the emails is of course intellectual before it is institutional: the pandemic-planning class understood platforms, finance, command, messaging, simulation, and compliance far better than it understood how outbreaks actually become epidemics and pandemics.
That’s more dangerous. That’s Chernobyl. They never imagined how wrong they could be. Hubris kills.
The emails reveal a class of officials and funders who believed they understood boards, secretariats, liability shields, global financing, public messaging, quarantine rules, cell-phone infrastructure, institutional models, modeling. But that’s not enough. They were also an elite convening against “resisters” better than they understood how outbreaks actually become epidemics and pandemics.
Their priorities were miswired.
Pandemic planning, as exposed in these emails, did not fail because the people involved lacked intelligence, credentials, connections, or money. It failed because they mistook institutional sophistication for epidemiological understanding. They mistook scenario rehearsal for knowledge. They mistook model complexity for biological realism. They mistook message control for reality, ultimately at the cost of public trust.
They mistook their duty to serve as authority.
When COVID arrived, that mismatch between reality and model was no longer theoretical. From 2020 through 2022, the same governing style drove the world to the brink of civic, educational, economic, and medical ruin, and it did so while insisting that truths inconvenient to the model had to be hidden, suppressed, delayed, ridiculed, or bureaucratically laundered before the public could be trusted to see them.
That is why the surfaced emails matter now. They are not merely historical curiosities from Davos. They are pre-incident photographs of the mindset that later revealed how broken public health really was.
Francis Collins entered the January 2016 Davos vaccine discussion uneasy. He called the session a “potential land mine.” He asked Anthony Fauci whether the proposal from Wellcome, the Gates Foundation, and the Norwegian government matched the model described in a recent medical-journal essay. He asked whether Gates had really bought in. He admitted that he had never quite understood how BARDA operated.
That admission should stop the reader’s blood cold. The Director of NIH, preparing to participate in a high-level pandemic vaccine session at Davos, framed the problem as a question of institutional mechanics: BARDA, Gates, Wellcome, Norway, models, proposals, governance, and whether HHS should participate without stepping into a trap. That was not an epidemiological question. It was a power-architecture question.
Not a sign of “reduction of morbidity and mortality” thinking in sight.
The vaccine-session report makes the same systemic failure visible. Pharmaceutical companies told the room they had spent heavily on Ebola vaccine development under boardroom resistance, and that they could not be expected to do the same thing for every future outbreak unless governments first resolved who would pay and who would absorb liability. They would offer platform capabilities, Collins reported, but not commercial intellectual property. Jeremy Farrar outlined a two-page concept for a new nonprofit partnership with a Managing Board and Scientific Advisory Group, while the World Economic Forum would serve as interim secretariat. Norway mentioned money. France praised the idea without offering any. Collins warned that the proposal was sketchy, that governance would be critical, and that HHS should avoid buying into a plan that degraded existing U.S. programs.
That caution was real. It should not be erased. Collins saw the institutional danger. But the frame remained sterile. The pathogen had already been converted into a finance, platform and public image problem. The outbreak had become another occasion to defend an institution.
Then came “Germ Games.”
A slow-moving Ebola-like outbreak is not the same as a fast respiratory epidemic like influenza. Those in control of the Ebola response made stupid model assumptions that drove the world to near panic in 2014. Jim Yong Kim seized on the idea of large simulations to get G20 attention and avoid a return from panic to neglect. Collins wrote that Kim called on NIH and Wellcome to join the World Bank and Gates Foundation over the coming weeks, potentially drawing on Department of Defense war-game expertise. Gates’ model-driven preparedness frame included cell-phone towers, delivery systems, quarantine rules, and messaging.
They did not deal with the goodness-of-fit of their response options to the reality of whatever pandemic might come next. They positioned. They flexed.
Their words have a technocratic shine. They also expose the error. Not one of those variables is the epidemic. They are administrative levers around an epidemic. They may matter after the biology is understood, but they cannot substitute for contact ecology, pathogen-specific transmission, ascertainment bias, case definition instability, reporting delay, nosocomial amplification, household clustering, funeral exposure, workplace network structure, age-risk stratification, seasonality, immune history, diagnostic performance, or the trust conditions under which symptomatic people decide whether to seek care, hide, travel, report, comply, resist, or flee.
Most importantly, COVID-19 taught us, they also substituted neither for accurate, differential diagnoses nor therapies and treatments.
The people in the room wanted to war-game the response. The first task was to understand the organism in its human environment.
They knew nothing about war-game room requirements.
Fauci’s reply illustrates the conceptual slippage and mismatch. He said HHS, DoD, and DHS had already conducted extensive tabletop exercises simulating aerosolized anthrax, multi-release smallpox, and pandemic influenza. He said this might not be exactly what Gates, Farrar, and Kim were talking about, but it would be “pretty close.”
Pretty close is not close enough.
Aerosolized anthrax is not pandemic influenza. Multi-release smallpox is not Ebola. Ebola is not Zika. Zika is not MERS. MERS is not H7N9. An intentional bioterror release, a respiratory pandemic, a hemorrhagic fever, and a mosquito-borne outbreak do not occupy the same transmission geometry. They do not share the same exposure routes, infectious windows, symptomatic profile, reporting incentives, feasible interventions, or ethical tradeoffs. They do not share the same immunology.
Treating them as variations on one emergency-management template creates the exact error that later swallowed COVID policy: when uncertainty rises, officials reach for the plan they rehearsed, even when the plan answers a different biological question.
Nicole Lurie’s response made the assumption explicit. The best modeling, she wrote, was around pandemic influenza, “but it is a great model for just about anything else.”
No sentence in the email packet better captures the failure. The absence of any dissent or corrective directions to that mindset is equally telling.
Pandemic influenza is not “a great model for just about anything else.” In fact, it’s a horrible model for other pathogens: disease severity, transmission dynamics, mechanisms of pathogenesis, evolutionary potential… I could go on and on. It may provide one preparedness template for one class of threat, but using flu as the governing metaphor for “just about anything else” imports flu assumptions into pathogens whose transmission dynamics reject them. It is the epistemology of bureaucratic convenience. It permits officials to say they have a model before they know whether the model belongs to the pathogen.
And since treatments and therapies for influenza do not factor in for pandemic responsiveness, using it as a model thrusts vaccines to the fore and everything else is cast aside.
Not an effective war-game strategy.
This IS a good example of how simplification becomes policy. First, the organism disappears into a scenario. Then the scenario disappears into a model. Then the model disappears into a playbook. Then the playbook becomes authority. By the time citizens meet the policy, the scientific assumptions have been buried several layers below law enforcement, agency guidance, school closure, hospital protocol, censorship pressure, employer mandate, and media panic.
The surfaced emails reveal at least eight simplifications that later matured into pandemic failure.
They replaced pathogen-specific transmission with generic pandemic imagination. They replaced field epidemiology with elite convening. They replaced human behavior with compliance management. They replaced uncertainty with model outputs. They replaced scientific dissent with narrative threat assessment. They replaced consent with resistance management. They replaced biological heterogeneity with population-wide control. They replaced public trust with messaging strategy.
The 2020–2022 disaster followed from those replacements.
When SARS-CoV-2 appeared, officials treated models as steering wheels instead of hypotheses. They encouraged the publication of simulations as data. They then tried to impose their simulations on reality.
They treated schoolchildren as transmission abstractions. They treated businesses as nonessential categories on a planner’s spreadsheet. They treated nursing-home protection as a slogan while failing the very populations whose risk was already obvious. They treated natural immunity, early outpatient treatment, stratified risk, vaccine durability, adverse events, mandates, masking, six-foot distancing, lab origin, and the social cost of lockdown as topics to be managed before they were topics to be understood. That is how a scientific emergency became a governance emergency.
The downfall of the official class did not occur because the public suddenly became irrational. It occurred because the public watched officials hide uncertainty while demanding compliance and obedience. The public watched institutions reverse positions without confession or explanation. The public watched experts ridicule questions that later became permissible. The public watched the same figures who claimed to represent science call for the public destruction of qualified scientific dissenters rather than open debate. The public watched the architects of model-driven policy avoid the consequences of model-driven damage.
The Great Barrington Declaration episode is not a side story. It is the Rosetta Stone. Collins did not respond to a rival pandemic-policy proposal by convening public debate among qualified scientists. He called for a “quick and devastating published takedown.” That was not science defending itself. That was institutional authority protecting a model of control. One model (non-stratified) over another (stratified).
Collins lost that argument based on data. Bhattacharya won.
The same impulse appears, in a different form, in the Davos notes on health-data access. Collins recorded a discussion with Carlos Moedas about Horizon 2020, health-data access with opt-out, and concern that a New England Journal of Medicine editorial on data sharing would “empower resisters.” Collins followed up with Jeff Drazen, and a new editorial was written.
Those events should infuriate physicians everywhere.
The Germ Games packet Collins, in that email thread, supported keeping data secret because the public might react badly. From a dynamics analysis/process autopsy position, public resistance to expansive data access was treated as a strategic obstacle to be overcome through elite editorial influence. The problem, in that frame, was not inadequate consent architecture, insufficient public trust, weak privacy guarantees, or the moral hazard of centralized data systems.
The problem was “resisters.”
That word belongs in the record.
It reveals how the managerial class sees the citizen when the citizen declines to become raw material for institutional design. A resister is not a rights-bearing person asking a legitimate question. A resister is friction. A resister must be bypassed, educated, messaged, neutralized, or discredited.
The H5N1 dual-use controversy supplies the adjacent secrecy reflex. In that case, the federal biosecurity apparatus initially recommended redaction of experimental details from papers involving mammalian transmissibility of H5N1, before the decision later shifted toward full publication. The issue was real. Dual-use research creates genuine risk. But the episode demonstrates the instinctive move: dangerous knowledge is to be controlled by expert committees, while the public is asked to trust the same apparatus that funds, reviews, classifies, declassifies, redacts, publishes, and later invokes the research as justification for more authority.
A bankrupt war-game option.
A free people cannot outsource the boundary between knowledge and secrecy to the same caste that benefits from both.
Vietnam and Ebola modeling show why this matters.
During the West African Ebola crisis, modelers produced projections that frightened governments and the public. The CDC’s conditional modeling warned that, without scaled intervention, Liberia and Sierra Leone could reach staggering case counts, with underreporting-adjusted estimates widely reported as reaching 1.4 million infections by January 2015. The qualifiers mattered: if trends continued, if interventions did not scale, if underreporting assumptions held. But public meaning stripped away the conditionals. The model became a prophecy. Ebola was going to take down Africa and then go global. There would be no stopping it.
So they invited the symptomatic to Ebola Treatment Units (ETUs) with the promise of a cure from Western medicine. They locked them in. 50% of the people who went into ETUs did not have Ebola. But 100% contracted it, and 50% of those people died. None received any care.
Had they modelers had it right, they would have realized Ebola was going to burn out and that hydration would have saved >90% of patients. Hydration.
That is the managerial class’s plan if it comes down to it. People can count on those like Fauci and Collins to abandon their humanity when their assumptions are called out by reality.
That was never serious field epidemiology. That was Brownian-motion epidemiology: the implicit treatment of human beings as randomly colliding particles in a social gas.
The formal language is homogeneous mixing and uniform transmissibility. The practical meaning is that the model blurs the structure through which infection actually moves. Ebola does not drift through a population like dye in water. Ebola moves through symptomatic bodies, caregiving, hospitals, unsafe burial, household exposure, funeral rites, fear, poverty, weak clinical infrastructure, reporting delays, and the fragile trust between affected communities and outside responders.
A Yale group led by Alison Galvani exposed the crucial flaw by modeling Ebola transmission within and between the community, hospitals, and funerals. That choice changed the policy meaning of the epidemic. The outbreak was not an unstoppable expanding cloud. It had choke points. It had settings. It had rituals. It had identifiable transmission pathways. Sanitary funeral practices, case isolation, contact tracing, quarantine, and protection of health care workers were not theatrical add-ons to a giant curve; they were the levers capable of breaking the chains of transmission.
That is what competent modeling does. It does not produce the scariest curve and call the curve reality. It maps the causal structure well enough to identify where reality can be changed.
Later work tested the assumptions that had made Ebola models brittle. Researchers found that Ebola did not homogeneously mix, that transmission followed predictable patterns, and that assumptions of uniform transmissibility and homogeneous mixing required direct testing rather than ritual acceptance. That is the scientific correction. It is not anti-modeling. It is anti-idolatry.
The distinction matters because model idolatry produces coercive politics. If the model says infection spreads as an undifferentiated cloud, the official reaches for undifferentiated control. Lock down everyone. Close schools. Close businesses. Restrict movement. Police gatherings. Command speech. Await vaccines. Treat human life as an input into a scenario. If the model respects structure, the official must do harder work: identify the actual routes of transmission, protect the actually vulnerable, fix the actual amplification sites, admit uncertainty, update priors, abandon failed assumptions, and preserve normal life wherever normal life does not drive the outbreak.
The Davos emails show the wrong lesson being learned from Ebola. Instead of concluding that outbreak control depends on local transmission structure, trust, field knowledge, and targeted interruption, the planning class concluded that the world needed bigger simulations, stronger funders, more platforms, better messaging, military-style exercises, and higher-level governance.
That was abstraction with a budget.
The 2020–2022 pandemic response then inflicted that abstraction on civilization. Reality kept objecting. Children were not interchangeable risk units. Nursing-home residents were not protected by slogans. Masking toddlers did not become science because officials needed visible compliance. Six feet did not become evidence-based because it appeared in guidance. Vaccine mandates did not become ethical because a model needed uptake. Censorship did not become public health because dissent complicated the message. The lab-leak hypothesis did not become false because powerful men preferred it buried. Natural immunity did not disappear because policy designers found it inconvenient. Economic life did not become expendable because a spreadsheet classified it as nonessential.
Reality eventually defeated the model, but only after the model had defeated millions of lives.
The emails therefore demand more than outrage. Outrage burns hot and then diffuses. The correct response is institutional redesign.
Pandemic planning must be rebuilt from the ground up around falsifiable pathogen-specific assumptions, not generalized emergency templates. Every model used to justify public restrictions must disclose its assumptions, data sources, uncertainty intervals, sensitivity analyses, validation status, failure history, and policy tradeoffs. Models fitted to cumulative incidence must be treated as high-risk instruments for false confidence unless they are cross-checked against raw incidence, ascertainment changes, reporting delays, and independent data streams. Homogeneous mixing assumptions must be prohibited as default policy inputs unless justified for the pathogen, setting, and time window at issue. Any model that ignores age-stratified risk, institutional amplification, household clustering, nosocomial transmission, and behavioral feedback should be barred from supporting population-wide mandates.
The public must have access to the plan before the emergency. Emergency protocols that alter civil liberties, school operations, medical access, business activity, speech conditions, or bodily autonomy must be published, debated, revised, and constrained in advance. No leaked PanCap-style architecture should govern a free country from the shadows. No official should be permitted to invoke a secret or semi-secret playbook after fear has made opposition politically costly.
Data systems must begin with consent, minimization, independence, and enforceable limits. A citizen who resists broad health-data access is not a public-health enemy. He is the reason ethical science exists. If the public refuses to trust the data architecture, the architecture has failed. The answer is not editorial pressure, linguistic laundering, or emergency exceptions. The answer is governance that deserves consent.
Scientific dissent must be protected during emergencies precisely because emergencies create the greatest temptation to suppress it. The more consequential the policy, the more aggressively debate must be preserved. The Collins “takedown” instinct must become a permanent warning label on federal science: no agency head should use public office to orchestrate reputational attacks on qualified scientists whose policy conclusions challenge the government’s preferred model.
Finally, pandemic planning must restore the difference between public health and command. Public health earns cooperation by telling the truth, showing its work, admitting error, targeting interventions, preserving normal life, protecting the vulnerable, and respecting the public as sovereign. Command demands compliance because the plan requires it. The surfaced NIH emails show a planning culture drifting toward command long before COVID gave it the opportunity to rule.
We know where that road leads because we lived there.
It led to shuttered schools, ruptured families, delayed medical care, censored physicians, coerced workers, damaged children, institutional distrust, economic wreckage, and a generation of public-health authority burned down by its own arrogance. The people who drove the world toward that ruin did not do so because they lacked models. They did so because they had models detached from reality and power detached from accountability.
The lesson is not that pandemics should not be planned for. The lesson is that pandemic planning must never again be left to people who convert living societies into tabletop exercises and then mistake the exercise for science.
The surfaced NIH emails are therefore not the end of a scandal. They are the beginning of a reckoning.
Demasi surfaced the documents. Tucker traced the machinery. The next step is to name the deeper failure: the pandemic-planning class did not understand outbreaks well enough to justify the authority it later claimed over the world.
We must make sure it never happens again.