Audio Version:

The deeper story of Netflix is not simply that another American corporation drifted left, adopted DEI, and began producing programming that many Americans find politically tiresome. It is the story of how entertainment, politics, government, technology, and the management of information have gradually converged, with Barack Obama and the political network surrounding him appearing repeatedly at the intersections. Once those connections are placed on a timeline, Netflix begins to look less like the story itself and more like one piece of a much larger post-presidential Obama project involving culture, political organization, information management, and now artificial intelligence, all accompanied by an expanding argument for regulation and control of the systems through which Americans communicate and obtain information. The justification offered again and again is the protection of democracy. The uncomfortable question is what happens to democracy when protecting it increasingly requires someone to decide which information the public should be permitted to hear, and even how "democracy " is defined.

The story begins, strangely enough, with Netflix co-founder Marc Bernays Randolph. Randolph was Netflix’s first CEO and came from one of the most remarkable families in the history of modern persuasion. Sigmund Freud was a relative, but the more relevant connection was Edward Bernays, Freud’s nephew and Randolph’s great-uncle, commonly described as the father of modern public relations. Robert and I explored Bernays and the history of modern propaganda in considerable depth in our book PsyWar: Enforcing the New World Order. But the essential point here is that Bernays saw public relations as something far more powerful than simply selling soap or cigarettes.

His famous 1928 book was bluntly titled Propaganda, and his central insight was that human behavior could be influenced by manipulating desires, fears, symbols, and social expectations, often without the people being influenced recognizing what was happening. Randolph himself has acknowledged the peculiar family connection. His own professional career was built around direct marketing, branding, and consumer behavior, and he has remarked upon how closely his work resembled Bernays’s.

Randolph left the company in 2003, long before Netflix became the cultural colossus it is today, and pretending otherwise would turn an interesting historical connection into an unsupported conspiracy theory. But what Netflix subsequently became would probably have fascinated Uncle Edward. A company that began by mailing DVDs eventually built one of the most powerful storytelling and behavioral-data platforms in human history, capable of learning what hundreds of millions of people watch, recommending what they should watch next, and placing cultural narratives directly into their homes. The question is not whether Netflix possesses that power. Obviously it does. The question is what happened when the people running and influencing that institution began explicitly viewing storytelling as a mechanism for changing society.

Netflix formally incorporated “inclusion” into its corporate culture in 2017. This was not confined to the unobjectionable proposition that qualified people should receive equal opportunities to work there. Netflix explicitly connected representation inside the corporation with representation on the screen: who gets hired, who gets to tell stories, and who tells stories with which stories audiences ultimately see.

The company adopted an institutional philosophy that changing the composition and worldview of the people producing entertainment would change the entertainment itself, and influence culture worldwide. What critics later called “woke Netflix” was therefore not simply an accidental accumulation of unusually progressive writers and producers. Netflix itself described inclusion as part of the machinery through which content was created.

Then came 2018. On March 28, Netflix appointed Susan Rice to its board of directors. Rice was not a Hollywood producer or entertainment executive. She had been Barack Obama’s ambassador to the United Nations and his National Security Advisor, one of the most powerful foreign-policy officials in the Obama administration.

Less than two months later, on May 21, Netflix announced an extraordinary production partnership with Barack and Michelle Obama. The Obamas created Higher Ground Productions and entered a multi-year agreement under which Netflix would distribute their films, television programs, documentaries, and other productions to a service that already reached approximately 125 million member households in 190 countries.

This was not a former president retiring to write his memoirs. It was a former president entering the global entertainment business in order to exert cultural and political influence, and the Obamas were remarkably candid about why. Michelle Obama explained that storytelling could inspire people and make them think differently about the world around them. Barack Obama later explained that Higher Ground had been created to “harness the power of storytelling.” Its productions would address race, class, democracy, and civil rights, and would not merely entertain. They would “educate, connect, and inspire.” Netflix later described Higher Ground as producing stories that embodied the values the Obamas had championed throughout their lives. Those are not accusations made by conservative critics. Those are descriptions supplied by the Obamas and Netflix.

This is also why the endless argument about whether individual “woke” Netflix programs receive good ratings misses much of the point. Some have succeeded spectacularly. Others have failed spectacularly. Netflix itself remains enormously successful, and people continue watching thrillers, romances, crime dramas, children’s programming, Korean television, fantasy, documentaries, and action movies by the billions of hours. But cultural influence does not require an audience to deliberately select something called A Progressive Guide to Social Justice. In many ways, that would defeat the purpose.

The most powerful cultural assumptions are often the ones the audience is not consciously debating at all. Ideas about race, sex, gender, marriage, government, capitalism, social democracy, American history, and social justice can be incorporated into entertainment until they simply become part of the fictional world being consumed. The viewer never has to agree to attend the lecture because there appears to be no lecture. The action movie gets the viewer through the door. The cultural assumptions come with the subscription.

Susan Rice’s career provides an unusually clear illustration of how thoroughly the political and cultural worlds had begun to merge. Rice joined Netflix’s board in 2018 and remained there until January 2021, when she left the company and returned to government as President Biden’s Domestic Policy Advisor. There she became centrally involved in an administration that elevated “equity” into an organizing principle across the federal bureaucracy. Rice left the Biden White House in May 2023. Four months later, she was back on Netflix’s board. Obama administration to Netflix. Netflix to Biden administration. Biden administration back to Netflix.

Corporate directors operate at the level of governance, executive oversight, leadership, risk, and institutional strategy, and Susan Rice is no different. The more important point is what the revolving door reveals. Government, political networks, Silicon Valley, and entertainment increasingly draw from the same pool of people carrying many of the same assumptions about race, equity, democracy, information, and social change.

Barack Obama, meanwhile, never really left politics. He changed the mechanisms through which influence could be exercised. The Obama Foundation develops future civic and political leaders. The Obama Presidential Center is intended to carry his political philosophy and institutional legacy into another generation. Higher Ground operates in entertainment and culture. Obama himself has remained an influential Democratic adviser, fundraiser, and political figure. Kamala Harris had a longstanding relationship with Obama, and Obama supported Joe Biden’s selection of Harris as his vice-presidential running mate in 2020. Obama also supported Harris’ selection as Biden’s replacement.

What happened after Harris was installed without a Democrat primary is considerably more revealing. Once Harris became the presumptive nominee, a substantial portion of the Obama political machinery moved into her campaign. David Plouffe, Obama’s 2008 campaign manager and one of the architects of his political operation, became a senior Harris adviser. Stephanie Cutter, Obama’s former communications director and 2012 deputy campaign manager, joined the operation. Other Obama veterans followed, while Harris campaign chair Jen O’Malley Dillon was herself an Obama campaign veteran. Obama’s presidency had ended years earlier, but the network had not. Political power had become institutional power, capable of surviving the individual administration that created it.

Then came Stanford, and this is where the Netflix story becomes considerably more consequential. On April 21, 2022, Obama delivered a major address at an event co-hosted by the Stanford Cyber Policy Center and the Obama Foundation called Challenges to Democracy in the Digital Information Realm. The subject was not primarily foreign affairs, taxation, health care, or any of the conventional issues associated with a former president. It was the architecture of information itself. Obama discussed algorithms, social media, disinformation, platform responsibility, Section 230, and the effect of digital information systems on democracy. His conclusion included an extraordinarily important sentence: “Regulation has to be part of the answer.” The Obama Foundation described the larger project as finding a new “North Star for the internet.”

There is another fact about Obama's Stanford appearance that makes the speech considerably more significant. Stanford was not simply a prestigious university providing a convenient auditorium. The event was co-hosted by the Obama Foundation and Stanford's Cyber Policy Center, and housed within that same Cyber Policy Center was the Stanford Internet Observatory, one of the central academic participants in the emerging government-industry-academic machinery for monitoring what Americans said online. The Observatory helped lead the Election Integrity Partnership, which worked with social-media companies to identify and refer election-related content to the government for potential action.

A later Republican-led House Judiciary investigation produced internal correspondence stating that the Election Integrity Partnership had been created “at the request of DHS/CISA,” along with the underlying Jira ticket system used to track supposedly problematic speech. The Stanford operation subsequently participated in the Virality Project, which extended this machinery from election-related claims into COVID-19 and vaccine narratives. Robert himself appears in those records. His supposed offense was his June 2021 Tucker Carlson interview, in which he argued that the risks of COVID vaccination outweighed the benefits for teenagers and young adults. But the Jira record did not merely describe the claim being monitored. It also categorized Robert’s affiliations as “right wing” and labeled him “general anti-vaccination.” The House Judiciary Committee subsequently concluded that the EIP’s system disproportionately targeted conservative speakers.

This information-management machinery was not merely academic. The consequences for the people caught in it were real. Robert was permanently removed from Twitter on December 29, 2021, wiping out an audience of more than half a million followers at the height of the national debate over COVID vaccine policy. His professional LinkedIn presence was also ultimately permanently banned, while his reach on other major platforms was restricted or diminished. The documentary record shows that, during this same period, Robert was being tracked through the EIP/Virality Project's Jira system, where his political affiliation was explicitly categorized as “right wing” alongside the supposedly offending vaccine claims. The practical result is difficult to dismiss: a credentialed scientist publicly challenging government COVID policy was simultaneously being monitored within the government-linked misinformation apparatus and losing access to the dominant digital platforms through which he communicated with the public. Twitter eventually restored his account after the company changed ownership. LinkedIn did not. The audience, professional network, and reach destroyed by those decisions were not restored simply because one account was later returned.

That makes what happened at Stanford in April 2022 difficult to dismiss as an unrelated academic exercise. The very first panel at Obama’s event was titled “The Trust Problem” and asked, “What is the role of the U.S. government in facilitating consensus and reducing polarization at home?” Its moderator was Renée DiResta, then research manager at the Stanford Internet Observatory. Obama's keynote is worth reading in that institutional context.

At an event co-hosted by his own foundation and the Cyber Policy Center that housed the Stanford Internet Observatory, Obama argued that online disinformation threatened democracy, called for reconsideration of Section 230, and said explicitly that “regulation has to be part of the answer.” Even more revealingly, he described the internet and social media as tools that “we can remake,” adding that society must decide what it values and then use those tools to advance those values.

What this all establishes is a striking convergence: Obama was publicly calling for greater intervention in the digital information environment from within an institutional ecosystem already deeply involved in building mechanisms through which government, academia, and technology companies could identify, refer, moderate, suppress, or otherwise reduce the reach of information deemed false or harmful.

Obama’s language deserves considerably more scrutiny than it received. The former president who had entered the global storytelling business was now making the regulation of the information environment itself a major part of his post-presidential agenda. The argument was presented in the reassuring vocabulary that has become familiar throughout the Western political establishment: democracy, safety, truth, responsible platforms, healthier discourse, misinformation, and protection from manipulation.

But underneath every one of those abstractions sits a very old and very dangerous political question. Who decides what constitutes misinformation? Who decides which speech is dangerous? Who determines what gets amplified, suppressed, demonetized, labeled, or removed? And what happens when the government officials, foundations, academics, corporations, and technology companies answering those questions overwhelmingly share the same political and cultural worldview? What happens when that viewpoint is skewed by globalism, leftist propaganda, and woke ideologies?

Those questions became considerably less theoretical during the Biden administration. Subsequent litigation, later known as Murthy v. Missouri, exposed extensive communications between federal officials and social-media companies about COVID-19, vaccines, elections, and alleged misinformation. The Fifth Circuit reviewed that record and concluded that the White House and Surgeon General’s office likely violated the First Amendment, finding that officials coerced platforms and significantly encouraged their content-moderation decisions. The Supreme Court refused to rule that this conduct was constitutional. Instead, the six-justice majority explicitly said, “We begin—and end—with standing,” concluding that the plaintiffs had not established the standing necessary to obtain a forward-looking injunction and that the Court therefore lacked jurisdiction to reach the merits. In other words, the Supreme Court left the central First Amendment question unresolved.

But the underlying communications were not hypothetical, and they did not disappear because the Supreme Court disposed of the case on standing. White House officials flagged content to platforms, demanded explanations for material that remained online, pressed for faster and more aggressive enforcement, threatened social media companies if they didn’t comply with their censorship demands, and questioned the platforms' moderation policies. Facebook personnel discussed “Specific Asks from the White House.” Press Secretary Jen Psaki publicly acknowledged that the administration was flagging “problematic posts” for Facebook. The Supreme Court never resolved the legal question of when those interactions crossed the constitutional line, though the Fifth Circuit did. What remains is the documentary record itself: the federal government had inserted itself into the machinery controlling the flow of information, pressuring private platforms over which Americans would be heard, which opinions would be suppressed, and which version of events would be allowed to circulate.

That history changes the meaning of pleasant-sounding discussions about creating a healthier information environment. For Americans whose speech was restricted, labeled, suppressed, or algorithmically buried during the COVID era, the issue is not theoretical. The danger exists whenever government officials, politically aligned institutions, and enormously powerful information companies begin sharing assumptions about what constitutes dangerous information. No smoky room is required. No secret memorandum has to be signed. Institutional alignment can accomplish a remarkable amount when the people staffing government, foundations, universities, media companies, and technology platforms have already reached similar conclusions about what constitutes truth, misinformation, equity, democracy, social justice, and acceptable speech.

Seen this way, Obama’s post-presidential career resolves into three parallel channels of influence. The first is political power: the Obama Foundation, Democratic Party relationships, fundraising, advising, and a political network whose personnel continue appearing in presidential campaigns long after Obama left office. The second is information power: Stanford, disinformation, algorithms, Section 230, platform responsibility, and increasingly explicit arguments that government regulation should help manage the digital information environment. The third is cultural power: Higher Ground and Netflix, using what Obama himself calls the “power of storytelling” to address race, class, democracy, and civil rights while reaching an audience no political campaign could approach.

“Politics is downstream from culture” is the most famous quote attributed to Andrew Breitbart, the American media publisher and conservative commentator who passed away in 2012.

The phrase, often referred to as the “Breitbart Doctrine,” posits that political outcomes are a result of prevailing cultural values, beliefs, and norms. Consequently, the doctrine argues that to effect meaningful political change, one must first shift the broader culture. An axiom of this is that scientism has replaced religion, and this is associated with a priesthood of anointed physicians and scientists. And, as any theology student will attest, religion is upstream of culture.

These channels do not have to be centrally coordinated to reinforce one another. Politics determines rules. Information systems determine which arguments people encounter. Culture shapes many of the assumptions through which people interpret those arguments. One operates through elections and government. Another operates through algorithms, search, moderation, and information distribution. The third operates through stories, heroes, villains, historical narratives, children’s programming, documentaries, and entertainment. Taken together, they represent a far more sophisticated mechanism of cultural influence than anything Edward Bernays could have imagined when he wrote Propaganda nearly a century ago.

And now the same argument has arrived at artificial intelligence. In September 2026, at a Democratic fundraiser, House Minority Leader Hakeem Jeffries asked Obama what Democrats should do about AI. Obama urged Democrats to make artificial intelligence a central part of their agenda and argued that, if Democrats regain control of the House after the midterm elections, they should establish a framework for a major public conversation about AI. He also urged future presidential candidates to develop clear AI agendas. Days later, Jeffries publicly identified AI as a high priority for House Democrats.

There are legitimate reasons to worry about artificial intelligence. Fraud, cybersecurity, autonomous weapons, labor displacement, deepfakes, privacy, and genuinely dangerous advanced systems deserve serious consideration. The existence of those problems does not establish that AI regulation is secretly a censorship program, and it would be foolish to make that assertion without evidence. But Obama’s history makes a different question unavoidable: what exactly will be regulated? Artificial intelligence is rapidly becoming the machinery through which Americans search for information, summarize history, compare arguments, evaluate evidence, write, create media, and increasingly understand the world. Regulation of AI therefore has implications extending far beyond whether a chatbot produces a dangerous recipe or a convincing deepfake. The rules governing AI can potentially influence the boundaries within which machines are permitted to answer questions at all. It will be and already is the next frontier in government censorship.

That is why Obama’s sudden emphasis on AI cannot reasonably be viewed in isolation from his broader post-presidential project. At Stanford he argued that the information environment itself required intervention and that government regulation had to be part of the answer. Through Higher Ground, he explicitly embraced storytelling as a means of educating and changing how people understand the world. Through his Foundation he has made democracy, information, and leadership development continuing institutional projects. Through the Democratic Party his political network has remained remarkably durable.

Now he is urging congressional Democratic leadership to make regulation of the most powerful information technology ever created a central agenda after the midterms. This provides ample reason to scrutinize any regulatory architecture that gives government, politically aligned institutions, or dominant technology companies greater authority over what artificial intelligence may say, particularly when the justification once again arrives wrapped in the language of protecting democracy from dangerous information.

The timing deserves scrutiny. An almost unknown 27-year-old Anthropic researcher resigned and warned that AI could kill humanity. He announced his departure internally on Anthropic’s Slack, and roughly thirty minutes later the Wall Street Journal had the story. Only afterward did he publish his public resignation statement on X, where a network of AI-safety advocates immediately amplified it. Within days, the CEOs of some of the world’s largest AI companies were publicly calling for a slowdown, government standards and outside control of frontier models. But the regulatory agenda had not suddenly appeared with Coxon’s resignation.

Anthropic’s Dario Amodei, OpenAI’s Sam Altman and Google DeepMind’s Demis Hassabis had already spent months converging on federal oversight of the technology they themselves dominate. Obama, meanwhile, had been talking privately with both Altman and Amodei about AI safety. Then, at a closed-door Democratic fundraiser, Hakeem Jeffries specifically raised Coxon’s resignation with Obama. Obama urged Jeffries to make AI a central Democratic agenda if his party retakes the House. Days later Jeffries was publicly demanding urgent action and invoking the CEOs’ call to slow development. This shows how quickly an industry-generated doomsday narrative traveled from a virtually unknown researcher, through the national media and America’s dominant AI companies, into a private conversation with a former president, and from there into the stated legislative agenda of the man who hopes to become the next Speaker of the House.

So who benefits from all of this regulation? The largest AI companies certainly can. OpenAI, Anthropic, Google and the other industry giants can afford armies of lawyers, compliance officers, safety researchers and government-relations specialists. A startup, university laboratory or independent developer cannot. Regulations requiring expensive testing, licensing and government approval can therefore protect the companies already dominating AI while making it far harder for new competitors to challenge them.

Government gains something too. An Obama political circle that has spent years arguing that the “information environment” must be managed and regulated now has an opportunity to help write the rules governing the technology that may soon become the world's primary gateway to information. The extraordinary concentration of corporate and government power produced by their proposed solutions should scare us all.

Nor is this confined to the United States. Europe has already moved much further toward regulating online speech and artificial intelligence through the Digital Services Act and AI Act. But the internet does not respect national borders. Global technology companies that want access to Europe's 450 million consumers must comply with European rules, and it can be easier to build those restrictions into their global platforms than maintain entirely separate systems. The result is the so-called “Brussels Effect”: European regulators can influence what Americans see, what platforms permit, and how AI behaves without Americans ever voting for a single one of them.

The Western world is deciding who will govern the information systems of the twenty-first century. Netflix initially looked like a story about an entertainment company that became political. It turns out to be a window into something much larger. Barack Obama’s post-presidential career has repeatedly converged on institutions that shape political leadership, popular culture, and the flow of information.

A government does not have to burn books to acquire enormous influence over political discourse. It can pressure platforms, define misinformation, regulate algorithms, establish rules for artificial intelligence, and encourage private institutions to police information according to standards government itself helped establish. At the same time, cultural institutions can normalize a political worldview through entertainment until ideas that were once controversial simply become the background assumptions of everyday life.

Each piece can be defended individually as reasonable. Netflix is only telling stories. Higher Ground is only making movies. Stanford is only studying disinformation. The Obama Foundation is only training leaders. Government officials are only protecting people from misinformation. Regulators are only making AI safe. The political network is only helping Democratic candidates win elections. Perhaps every individual involved sincerely believes exactly that. The problem appears when all of these pieces are placed beside one another and the direction of travel becomes visible. The danger is not necessarily a conspiracy. The danger is a system in which the same political assumptions increasingly dominate the institutions that produce culture, define misinformation, regulate information technology, train political leadership, and determine the permissible boundaries of digital speech.

That is why the question of whether Netflix has become “woke” now seems almost quaint. Netflix is only one piece of a much larger struggle over who gets to shape the stories Americans hear, which information Americans find, and increasingly which answers their machines are permitted to give them. Ronald Reagan famously warned that freedom is never more than one generation away from extinction. The modern threat to that freedom may not arrive wearing jackboots or carrying a list of banned books. It may arrive promising safety, inclusion, equity, responsible information, and the protection of democracy itself. Edward Bernays understood a century ago that the most effective propaganda does not feel like propaganda.

Perhaps the most dangerous form of censorship is the one people never recognize as censorship at all, because someone else has already decided which stories they will hear, which facts they will find, and which ideas they will never encounter.

By: JGM/RWM

References

1. Malone, Robert W., and Jill Glasspool Malone. PsyWar: Enforcing the New World Order. Skyhorse Publishing, 2024. See particularly the discussion of Edward Bernays, modern propaganda, the Election Integrity Partnership, Virality Project, Jira ticketing system, and censorship-industrial complex. The book documents the Jira entry concerning Robert Malone, including the descriptors “right wing” and “general anti-vaccination.”

2. Bernays, Edward L. Propaganda. Horace Liveright, 1928. Foundational statement of Bernays’s theories concerning public opinion, persuasion, and the organized manipulation of attitudes.

3. Netflix. “Ambassador Susan E. Rice Appointed to Netflix Board of Directors.” March 28, 2018. Rice, formerly Barack Obama’s National Security Advisor and U.N. Ambassador, joined Netflix’s board less than two months before the Obama production agreement was announced.
   Netflix announcement

4. Netflix. “Netflix Forming Storytelling Partnership With Barack and Michelle Obama.” May 21, 2018. Announces the multi-year Higher Ground agreement and states that its productions would be available to Netflix’s then-125 million member households in 190 countries.
   Netflix-Higher Ground announcement

5. Higher Ground Productions/Netflix. “Higher Ground Announces Upcoming Slate of Projects Exclusive to Netflix.” April 30, 2019. Particularly important for Obama’s statement that Higher Ground was created to “harness the power of storytelling” and that projects involving race, class, democracy and civil rights would educate as well as entertain.
   Higher Ground project announcement

6. Netflix. “Inclusion at Netflix.” Netflix Careers. Netflix explicitly connects its internal inclusion policies to what appears on screen, stating that “Better representation on-screen starts with representation in the office.”
   Netflix inclusion policy

7. Netflix. “Netflix Reappoints Ambassador Susan E. Rice to Its Board of Directors.” September 9, 2023. Documents Rice’s return after serving as Biden’s Domestic Policy Advisor.
   Rice reappointment announcement

8. Netflix, Inc. 2026 Proxy Statement. U.S. Securities and Exchange Commission. Confirms Rice’s continuing service on the Netflix board and Nominating and Governance Committee, and describes her government and public-policy expertise as relevant to Netflix.
   Netflix 2026 proxy statement

9. Netflix, Inc. Form 8-K, June 5, 2026. SEC filing confirming Rice and the other Netflix directors were reelected at the 2026 annual meeting.
   Netflix June 2026 SEC filing

10. Stanford Freeman Spogli Institute/Cyber Policy Center. “Challenges to Democracy in the Digital Information Realm.” April 21, 2022. Official program for the Obama Foundation/Stanford Cyber Policy Center event. Particularly relevant is Panel I, moderated by Stanford Internet Observatory research manager Renée DiResta, asking: “What is the role of the U.S. government in facilitating consensus and reducing polarization at home?”
    Stanford event program and Obama transcript

11. Stanford University. “Disinformation is weakening democracy, Barack Obama said.” April 21, 2022. Stanford’s account of Obama’s keynote, including his statement that “regulation has to be part of the answer,” his discussion of Section 230, algorithms, AI/deepfakes and the information environment.
    Stanford report on Obama’s speech

12. Obama Foundation. “President Obama Speaks at Stanford on Disinformation.” April 2022. The Foundation framed Obama’s project as establishing a new “North Star” for the internet and addressing the effects of the digital information environment on democracy.

13. U.S. House Committee on the Judiciary and Select Subcommittee on the Weaponization of the Federal Government. Interim reports and supporting documents concerning CISA, the Election Integrity Partnership, Stanford Internet Observatory and the Virality Project. These materials form an important basis for the essay’s discussion of government-academic-platform relationships.

14. U.S. Court of Appeals for the Fifth Circuit. Missouri v. Biden, No. 23-30445, September 8, 2023. The appellate court agreed that some federal officials had likely coerced or significantly encouraged social-media censorship in violation of the First Amendment and modified the district court’s injunction accordingly.
    Fifth Circuit opinion

15. Supreme Court of the United States. Murthy v. Missouri, 603 U.S. 43 (2024), decided June 26, 2024. The Court reversed on standing grounds rather than deciding whether the challenged government conduct violated the First Amendment.
    Supreme Court opinion

16. Reuters. “How Obama boosted Harris and continues to shape the Democratic Party.” August 20, 2024. Documents Obama’s longstanding relationship with Kamala Harris, his advice to her after Biden’s withdrawal, and the movement of Obama veterans including David Plouffe and Stephanie Cutter into the Harris campaign.
    Reuters report on Obama and Harris

17. The Wall Street Journal. “The Anonymous Math Geek Who Quit Anthropic—and Became the Face of AI Safety.” September 17, 2026. Detailed reconstruction of Jacob Coxon’s resignation from Anthropic, his contacts with AI-safety advocates, the timing of the Journal’s reporting and his subsequent public announcement, and the rapid amplification of his warnings.
    Wall Street Journal investigation

18. The Wall Street Journal. “Anthropic Researcher Quits Over ‘Out-of-Control’ AI Fears.” September 2026. Contemporary coverage of Coxon’s resignation and warnings concerning catastrophic AI risk.

19. The Wall Street Journal. “Biggest AI Rivals Agree They Need to Slow It Down.” September 2026. Documents the unusual convergence of Dario Amodei, Sam Altman, Elon Musk and other major AI figures around concerns about the pace and safety of frontier AI development.

20. The Wall Street Journal. “Congress Is Suddenly Waking Up to the AI Doomsday Threat.” September 2026. Contemporary account of congressional reaction to the AI-safety warnings and proposals for federal oversight.

21. Reuters. “Obama voices caution on AI, urges Democrats to tackle it, NYT says.” September 13, 2026. Reports Obama’s discussion with House Minority Leader Hakeem Jeffries at a private Democratic fundraiser and Obama’s recommendation that AI become a central Democratic agenda if the party regains the House.
    Reuters report on Obama, Jeffries and AI

22. Office of Democratic Leader Hakeem Jeffries. September 2026 statements and interviews concerning artificial intelligence. Jeffries subsequently publicly agreed with Obama that action on AI was necessary and identified AI as a priority for House Democrats.

23. European Union. Regulation (EU) 2024/1689, the Artificial Intelligence Act. Especially Article 2, establishing that aspects of the Act apply to providers outside the EU when systems or general-purpose models are placed on the EU market, and in specified circumstances when AI output is used within the Union.
    Official text of the EU AI Act

24. European Commission. Digital Services Act. Official materials concerning the EU regulatory framework governing online intermediaries and platforms, including services based outside the EU when offered within the European market.

25. Engler, Alex. “The EU AI Act will have global impact, but a limited Brussels Effect.” Brookings Institution, June 8, 2022. Useful background on the mechanism by which EU technology regulation can influence products and systems outside Europe even when EU law does not formally govern users in those jurisdictions.
    Brookings analysis of the Brussels Effect

26. Reuters. “Carney says Canada seeks ‘unique alliance’ but not membership with the European Union.” September 13, 2026. Reports discussions of a new Canada-EU relationship and cooperation involving AI, energy, critical minerals and other strategic sectors.

27. Reuters. “Canada and EU should ally to protect democracies, Carney says, after ‘associate member’ proposal angers Trump.” September 17, 2026. Reports von der Leyen’s proposal for an unprecedented EU “associate member” relationship with Canada and Carney’s response.
    Reuters on the Canada-EU proposal

28. Reuters. “Highlights from Canadian PM Mark Carney’s speech in Strasbourg.” September 17, 2026. Documents proposed Canada-EU cooperation involving AI, digital trade, defense, energy and strategic infrastructure.

29. Reuters. “HIGHLIGHTS – EU Commission chief ‘state of the union’ address.” September 16, 2026. Covers von der Leyen’s associate-member proposal for Canada and her parallel call for international cooperation and a slowdown in AI development.

30. Reagan Presidential Foundation & Institute. Ronald Reagan quotation archive and July 6, 1987 remarks containing the warning that “Freedom is never more than one generation away from extinction.” This is the source for the quotation used in the essay’s conclusion.

31. 31. Randolph, Marc. That Will Never Work: The Birth of Netflix and the Amazing Life of an Idea. Little, Brown and Company, 2019. Randolph discusses growing up surrounded by Freud’s writings and portraits, explains that Freud was his father’s great-uncle, and identifies Edward Bernays as his grandmother’s brother. He describes Bernays as the father of modern public relations who applied psychology and psychoanalysis to marketing. The original hardcover was published September 17, 2019.
    Publisher page for That Will Never Work

32. 32. Randolph, Marc. “That Will Never Work: The Birth of Netflix and the Amazing Life of an Idea with Marc Randolph.” Interview with Dr. Patty Ann Tublin, The Trust Doctor Podcast. This is the especially useful source for the essay because Randolph himself explicitly connects Bernays’s work to his own career. After discussing Freud, Randolph calls Edward Bernays the relative “more coincidentally aligned with what my career ended up becoming.” He explains that Bernays recognized that it was possible to change how people think without their necessarily recognizing that they were being changed, then observes that he himself spent his life fundamentally as a marketing person and jokes that the Freud/Bernays connection is almost too coincidental not to involve some “DNA connection.”
    Full Marc Randolph interview and transcript

33. 33. Randolph, Marc. “Marc Randolph: Netflix Co-Founder, Entrepreneur, Mentor & Investor.” Right About Now: Legendary Business Advice. Randolph discusses his twelve years in direct marketing, calls marketing his “first love,” identifies Edward Bernays as his great-uncle and “father of public relations,” and explains how central marketing was to Netflix’s early development.

Audio Version:

More than two-thirds of American adults do not have enough magnesium.

A 2026 analysis of U.S. adults participating in the National Health and Nutrition Examination Survey found that 67.8 percent had serum magnesium concentrations below 2.06 mg/dL, the threshold the researchers used for chronic latent magnesium deficiency. That number is remarkable, particularly because serum magnesium may actually be a relatively insensitive way of detecting inadequate magnesium stores.

Magnesium deficiency is generally assessed by measuring magnesium in the blood, yet less than one percent of the body’s magnesium is found in serum. Most magnesium is stored in bone and within cells. The body also regulates circulating magnesium rather tightly, meaning that serum concentrations do not necessarily provide a complete picture of magnesium stores. A normal blood test does not necessarily mean that the body’s magnesium stores are adequate.

There is an important issue concerning how we define deficiency. The researchers used a higher serum magnesium cutoff than is conventionally used to diagnose clinical hypomagnesemia. Their threshold of 2.06 mg/dL has been proposed as a better marker for chronic latent magnesium deficiency, based in part on studies showing physiological changes and increased health risks at magnesium concentrations that conventional laboratory ranges may still classify as normal. In other words, by the time someone crosses the conventional threshold for frank hypomagnesemia, we may be looking at the far end of the problem rather than its beginning.

There is even more reason to take the finding seriously. We already know from dietary surveys that Americans aren’t consuming enough magnesium. Analysis of NHANES dietary data found that 48 percent of Americans consumed less magnesium from food and beverages than their Estimated Average Requirement.

These are two different measurements asking two different questions, but they point in the same direction. One finds widespread inadequate dietary intake. The other suggests that suboptimal magnesium status may be even more common.

Why should we care?

Because magnesium isn’t a minor nutritional accessory. It is involved in hundreds of enzymatic reactions and some of the most fundamental processes keeping us alive. Magnesium is required for the production and utilization of ATP, the energy currency of the cell. It plays essential roles in protein synthesis, DNA and RNA metabolism, glucose regulation, muscle and nerve function, vascular tone, cardiac electrical stability and bone metabolism. In the brain, magnesium helps regulate neuronal signaling and synaptic plasticity, processes fundamental to learning, memory and cognitive function.

Low magnesium status has been associated with cognitive impairment and dementia, while experimental studies have shown that increasing magnesium concentrations in the brain can enhance synaptic plasticity and improve learning and memory. Much of the early work involved magnesium-L-threonate, a form developed specifically to increase magnesium availability in the brain. The evidence has now progressed beyond animal studies.

Randomized, placebo-controlled human trials have reported improvements in several measures of cognition and memory, including working and episodic memory, following magnesium-L-threonate supplementation. What has not yet been established is whether these cognitive effects translate into prevention or treatment of Alzheimer’s disease or other dementias.

Human studies also suggest that magnesium status differs in people with Alzheimer's disease. A meta-analysis of 21 studies found significantly lower serum and plasma magnesium in Alzheimer's patients than in healthy controls. Perhaps more interestingly, postmortem studies of human brains have reported lower magnesium concentrations in Alzheimer's-affected regions, including the hippocampal region, entorhinal cortex, and frontal cortex. These findings do not prove that magnesium deficiency causes Alzheimer's disease. Neurodegeneration itself could alter magnesium metabolism. But prospective evidence makes the relationship harder to dismiss: in the Rotterdam Study of 9,569 initially dementia-free adults, low serum magnesium was associated with a 32 percent greater subsequent risk of dementia.

Which raises an obvious question.

Why are so many Americans apparently running short of a mineral this fundamental to human biology?

There isn’t one answer. The medical system likes to point out that medications, gastrointestinal disease, diabetes, alcohol use, kidney loss, and aging can all affect magnesium status.

But there is another explanation sitting directly in front of us.

The food industry removed much of the magnesium from our food.

Big Ag removed much of the magnesium from food

When we wrote Homesteading for Health, we spent considerable time examining something largely missing from modern discussions about nutrition. Food isn’t simply protein, carbohydrate, and fat.

Food contains minerals.

And the nutritional composition of our food is not fixed.

The nutritional quality of our food has changed. Evidence accumulated over decades shows that modern high-yield agriculture can produce more food per acre while reducing the concentration of minerals in that food. Historical USDA comparisons found substantial declines in several nutrients in fruits and vegetables between 1950 and 1999. Long-running agricultural experiments provide even stronger evidence. The Broadbalk Wheat Experiment in England, which has been continuously cultivated and studied since 1843, has documented significant declines in magnesium, zinc, iron, and copper concentrations in wheat grain associated with modern high-yield varieties. As crop yields increased, mineral concentrations often fell, a phenomenon researchers call the dilution effect.

But yield is only part of the story. Plants do not simply vacuum minerals out of dirt. Healthy soil is a living biological system in which bacteria, fungi, and plant roots work together to make minerals available to the plant. Modern tillage, heavy fertilizer use and other intensive practices disrupt those relationships, particularly the mycorrhizal fungal networks that extend the effective reach of plant roots. A soil may therefore contain magnesium and other minerals on a laboratory assay while becoming less biologically effective at delivering them into the crop.

What matters isn’t simply whether magnesium exists in the dirt. What matters is whether it makes its way from the soil into the food and ultimately into us.

Then we take the food we produce and process even more magnesium out of it.

We know exactly where much of it went.

Industrial milling removed it from our food

Again, in Homesteading for Health, we document that much of the issue stems from the production of processed flour. The wheat kernel consists of the bran, germ, and endosperm. The bran and germ contain a disproportionate share of the grain’s minerals, including magnesium. Making refined white flour removes those portions of the kernel.

The result can be a dramatic reduction in magnesium.

And then comes one of the stranger ironies of our modern food system.

We call the resulting product “enriched flour.”

Iron and selected B vitamins are added back after milling, and folic acid is added to most enriched grain products.

Magnesium generally isn’t.

We take a naturally magnesium-containing food, remove much of its magnesium, replace a selected handful of the nutrients lost during processing, and call the finished product enriched.

Then we make refined flour a foundation of the American diet.

Perhaps the magnesium problem isn’t quite so mysterious after all.

Where Did the Magnesium Go?

The best natural sources of magnesium aren’t particularly exotic.

Pumpkin seeds, chia seeds, almonds and other nuts, beans, lentils, whole grains and dark leafy vegetables are all good sources. Magnesium is, after all, at the center of the chlorophyll molecule. Green plants are therefore an obvious place to find it.

The modern American diet has moved in almost exactly the opposite direction.

Whole grains became refined grains. Beans and legumes became less central to many meals. Nuts and seeds became optional snacks rather than routine foods. Fresh vegetables increasingly compete with highly processed products made primarily from refined flour, starch, sugar, and industrial oils.

The result is an interesting nutritional paradox.

Americans can consume more than enough calories while remaining poorly supplied with micronutrients.

We can be overfed and undernourished at the same time.

Then There Is the Brain

For years, magnesium was discussed primarily in relation to muscle function, bone, cardiovascular health, and metabolism. Its role in cognition received considerably less public attention.

That has changed.

Magnesium is involved in neuronal signaling and synaptic plasticity, the process through which connections between neurons change in response to experience. These mechanisms are fundamental to learning and memory.

Animal experiments produced some striking early findings. Increasing brain magnesium concentrations improved measures of synaptic plasticity, learning, and memory, and experimental work with magnesium-L-threonate generated particular interest because this compound was developed to increase magnesium availability in the brain.

There are also observational human data connecting magnesium status with cognitive health and dementia. A 2024 systematic review examined three randomized controlled trials and 12 cohort studies. The cohort evidence suggested relationships between magnesium status and cognitive outcomes. Interestingly, the researchers also found a U-shaped association between serum magnesium and dementia or cognitive impairment, with approximately 0.85 mmol/L appearing to represent an optimal concentration in the available cohort data.

That last observation is important.

More is not always better.

The objective isn’t to consume enormous amounts of magnesium. It is to have enough.

Magnesium-L-Threonate and Memory

This brings me to the form of magnesium that I find particularly interesting: magnesium-L-threonate, commonly abbreviated MgT.

Magnesium-L-threonate was developed specifically to increase magnesium concentrations in the central nervous system. Early animal experiments reported improvements in learning, working memory, and both short- and long-term memory.

The evidence has now moved into humans.

A recently published randomized, double-blind, placebo-controlled trial enrolled 100 adults aged 18 to 45. Participants received either 2 grams per day of magnesium-L-threonate or placebo for six weeks.

Compared with placebo, the magnesium-L-threonate group demonstrated greater improvement in the NIH Total Cognition Composite, with particularly notable effects involving working and episodic memory. Reaction time also improved. Not every endpoint improved, including several objective measures of sleep, but the study provides controlled human evidence supporting some of what the earlier animal studies had found.

This is the form of magnesium I personally favor when cognition and brain health are among the objectives.

Your Heart Needs Magnesium

Magnesium is also deeply involved in cardiovascular physiology.

It influences vascular smooth-muscle tone, movement of calcium and potassium across cell membranes, and the electrical activity of the heart. Severe magnesium deficiency can produce abnormal heart rhythms. Magnesium is important enough to cardiac electrophysiology that intravenous magnesium has long-standing medical uses for certain arrhythmias and other acute clinical settings.

But the more interesting question for most of us concerns the effects of chronic, less severe magnesium inadequacy.

A meta-analysis of 38 randomized trials involving 2,709 participants found an average reduction of 2.8 points in the upper (systolic) blood pressure number and 2.1 points in the lower (diastolic) number compared with placebo. In practical terms, a reading of 130/80 would decline, on average, to about 127/78.

Magnesium supplementation appears to lower blood pressure most in people who actually have high blood pressure, particularly those already being treated for hypertension. A 2025 meta-analysis of 38 randomized controlled trials involving 2,709 participants found that among people with hypertension who were already taking blood-pressure medication, magnesium supplementation lowered the upper blood-pressure number by an average of 7.7 points and the lower number by about 3 points compared with placebo. People who were magnesium deficient also experienced larger reductions, while those with normal blood pressure did not experience a statistically significant reduction.

Magnesium is also closely connected to glucose metabolism and insulin signaling. In a pooled analysis of 24 randomized trials involving people with type 2 diabetes, magnesium supplementation was associated with reductions in fasting glucose, HbA1c, and blood pressure.

Magnesium isn’t a miracle cure.

It is something more mundane and perhaps more important: a basic nutrient necessary for the metabolic systems we are asking our bodies to operate every day.

How Much Magnesium Do We Actually Need?

The familiar recommendation of approximately 400–420 mg of magnesium per day for adult men and 310–320 mg for adult women is not a newly established threshold for optimal health. These numbers come from Dietary Reference Intakes established by the Institute of Medicine in 1997, based on the evidence available nearly three decades ago. The RDA was intended to estimate an intake sufficient to meet the physiological requirements of nearly all healthy people. It was not designed to determine the amount of magnesium associated with optimal cardiovascular, metabolic or neurological health.

More recent research makes the picture considerably more complicated. Large prospective studies have repeatedly associated higher dietary magnesium intake with lower risks of hypertension, cardiovascular disease, type 2 diabetes and premature death. Randomized trials examining magnesium as an intervention frequently provide several hundred milligrams of supplemental elemental magnesium per day, in addition to whatever magnesium participants obtain from food. These studies do not establish a single higher “optimal” intake for everyone, but they do suggest that the decades-old RDA should not automatically be interpreted as the amount beyond which magnesium provides no additional physiological benefit.

The emerging research on cognition is particularly interesting because it raises another question: does the form of magnesium matter as much as the amount? Several human trials investigating magnesium L-threonate have used approximately 1.5–2 grams per day. That sounds like an enormous magnesium dose, but it is important to understand the labeling. Two grams of magnesium L-threonate is not two grams of elemental magnesium. In a recent randomized controlled trial, 2,000 mg per day of magnesium L-threonate supplied only about 145 mg of elemental magnesium. Earlier cognition studies using 1.5–2 grams of magnesium L-threonate similarly supplied only about 124–166 mg of elemental magnesium per day, on top of the magnesium participants were already obtaining from their diets.

This distinction is important. These cognition studies do not demonstrate that people need two grams of elemental magnesium every day. Instead, they raise a potentially more interesting possibility: the biological effects of magnesium may depend not simply upon how much magnesium is consumed, but upon how effectively a particular magnesium compound is absorbed, distributed and delivered to specific tissues. Magnesium L-threonate has attracted particular attention because of its proposed ability to increase magnesium availability within the nervous system, although claims that it is uniquely or substantially better at reaching the human brain remain less firmly established than the marketing sometimes suggests.

This also exposes one of the limitations of discussing magnesium solely in terms of an RDA. The amount required to prevent overt deficiency, the amount necessary to maintain whole-body magnesium balance, and the amount and formulation capable of producing a specific therapeutic effect are three different questions. A nutritional recommendation developed in 1997 cannot by itself answer whether additional magnesium, or a particular form of magnesium, might improve blood pressure, glucose metabolism, sleep or cognition. Those questions have to be answered experimentally, and increasingly, researchers are doing exactly that.

Robert and I actually take 2 grams of magnesium L-threonate without ill effects, but that is not a recommendation - just what we have come to.

Those numbers aren’t particularly difficult to reach if someone eats a nutrient-dense whole-food diet.

That “if” is doing a lot of work.

The NIH identifies legumes, nuts, seeds, whole grains, and green leafy vegetables as major dietary sources.

That list looks remarkably similar to the foods that have been displaced by the modern processed diet.

Food First. But Perhaps Not Food Alone.

Magnesium-rich foods should be the foundation. Eat green vegetables. Eat beans and legumes. Eat nuts and seeds. Choose whole grains rather than refined flour. Grow some of your own food if you can. Buy food from farmers who care about their soil. Reduce the amount of highly processed food in your diet.

But there is a problem with stopping the discussion there.

The food isn’t the only thing that changes as we get older. We change too.

Aging alters the gastrointestinal tract and can reduce how efficiently some nutrients are absorbed. Older people may also eat less food, have less varied diets, take medications that interfere with magnesium balance, or develop diseases that increase magnesium losses. Kidney handling of magnesium also changes with age. Proton-pump inhibitors are a particularly important example because long-term use can cause hypomagnesemia in susceptible individuals. Diuretics and several other medications can also alter magnesium status.

This means that the person who was able to maintain adequate magnesium status at 35 may not necessarily do so at 65 or 75 while eating essentially the same diet.

That matters because aging is precisely when many of the biological systems that depend upon magnesium become increasingly important: cardiovascular function, glucose regulation, bone health, muscle function and cognition.

So I don’t think “food first” should automatically mean “food only.”

For many adults, particularly as we age, magnesium supplementation is worth considering alongside a magnesium-rich diet.

The form matters. Magnesium oxide is inexpensive and contains a lot of elemental magnesium, but it is poorly absorbed and commonly causes gastrointestinal effects. Magnesium citrate is better absorbed and useful for people who also benefit from its laxative effect, but for others loose stool isn’t particularly desirable. Magnesium glycinate is widely used and generally well tolerated. Magnesium-L-threonate is considerably more expensive, but is particularly interesting because it was developed to increase magnesium availability in the brain and now has both animal and human evidence concerning cognition.

This is why magnesium-L-threonate is the form I personally favor when brain and cognitive health are part of the objective.

But supplementation shouldn’t become another exercise in the American belief that if some is good, more must be better. The NIH establishes a 350 mg/day upper limit for magnesium from supplements and medications in adults. Importantly, that limit does not include magnesium naturally present in food. It was established primarily because supplemental magnesium can cause diarrhea, nausea, and abdominal cramping. Much higher intakes can become dangerous, especially in people with impaired kidney function, because the kidneys normally eliminate excess magnesium. Magnesium supplements can also interact with several medications.

The objective is adequacy, not excess.

We Keep Rediscovering What Food Used to Provide

Perhaps the most remarkable thing about magnesium is how unremarkable it really is.

It isn’t a newly discovered molecule. It isn’t a pharmaceutical invention. Human beings have required magnesium throughout our evolutionary history.

What changed wasn’t our need for it.

What changed was our food, and as we age, our ability to extract and retain nutrients from that food changes as well.

Modern agriculture bred crops for yield. Industrial food processing refined grains for texture, appearance, and shelf life, stripping the magnesium-rich germ and bran from wheat in the process. The result is an enormous processed-food industry capable of delivering abundant calories while removing many of the micronutrients that once accompanied those calories.

Then we got older.

We ate less. Our gastrointestinal physiology changed. Some of us developed diabetes or gastrointestinal disease. Many began taking medications that affect magnesium absorption or excretion. Yet our physiological requirement for magnesium didn’t disappear.

Nearly half of Americans now consume less magnesium from food than their estimated requirement. A new analysis suggests that more than two-thirds may have serum magnesium concentrations below a proposed threshold for adequate status.

Start with food. Eat the foods in which magnesium naturally occurs. Stop stripping those foods of their nutrients whenever possible.

But particularly as we age, it is also reasonable to consider magnesium supplementation rather than simply assuming that a modern diet will provide everything we need.

We spend enormous amounts of money looking for complicated ways to remain healthy as we age.

Sometimes the answer may begin with something considerably simpler.

Make sure the basic machinery has what it needs to run.

By: JGM/RWM

References

Malone RW, Malone JG. Homesteading for Health: A Family’s Guide to Self-Sufficiency and Wellness. Skyhorse Publishing; 2026. The article’s discussion of changes in food nutrient density, soil biology, modern wheat, flour refining, and regenerative agriculture draws in part from the research reviewed in this book.

1. Jiao K, Costello R, Gahche J, Rosanoff A, Wallace TC. Serum magnesium concentrations in the United States: An updated population reference interval in children and adults. J Nutr. 2026;156(6):101539. doi:10.1016/j.tjnut.2026.101539.

2. National Institutes of Health, Office of Dietary Supplements. Magnesium: Fact Sheet for Health Professionals. NIH. The NIH review includes dietary requirements, food sources, absorption, assessment of magnesium status, supplement forms, medication interactions, and the adult upper limit for supplemental magnesium. It also reports that 48% of Americans consumed less magnesium from food and beverages than their respective EARs in NHANES 2013–2016.

3. Davis DR, Epp MD, Riordan HD. Changes in USDA food composition data for 43 garden crops, 1950 to 1999. J Am Coll Nutr. 2004;23(6):669–682. doi:10.1080/07315724.2004.10719409.

4. Fan MS, Zhao FJ, Fairweather-Tait SJ, Poulton PR, Dunham SJ, McGrath SP. Evidence of decreasing mineral density in wheat grain over the last 160 years. J Trace Elem Med Biol. 2008;22(4):315–324. doi:10.1016/j.jtemb.2008.07.002. This is the important Broadbalk study documenting significant declines in magnesium, zinc, iron and copper in wheat grain after the introduction of modern high-yielding cultivars.

5. Slutsky I, Abumaria N, Wu LJ, et al. Enhancement of learning and memory by elevating brain magnesium. Neuron. 2010;65(2):165–177. doi:10.1016/j.neuron.2009.12.026.

6. Li W, Yu J, Liu Y, et al. Elevation of brain magnesium prevents synaptic loss and reverses cognitive deficits in Alzheimer’s disease mouse model. Mol Brain. 2014;7:65. doi:10.1186/s13041-014-0065-y.

7. Chen F, Wang J, Cheng Y, et al. Magnesium and cognitive health in adults: A systematic review and meta-analysis. Adv Nutr. 2024;15(8):100272. doi:10.1016/j.advnut.2024.100272.

8. Lo K, Liu Q, Madsen T, et al. Relations of magnesium intake to cognitive impairment and dementia among participants in the Women’s Health Initiative Memory Study: a prospective cohort study. BMJ Open. 2019;9:e030052. This study followed 6,473 women aged 65–79 for more than 20 years and examined magnesium intake in relation to incident MCI and probable dementia.

9. Lopresti AL, Smith SJ. The effects of magnesium L-threonate (Magtein®) on cognitive performance and sleep quality in adults: a randomised, double-blind, placebo-controlled trial. Front Nutr. 2026;12:1729164. doi:10.3389/fnut.2025.1729164.

10. Kass L, Weekes J, Carpenter L. Effect of magnesium supplementation on blood pressure: a meta-analysis. Eur J Clin Nutr. 2012;66(4):411–418. doi:10.1038/ejcn.2012.4.

11. Zhang X, Li Y, Del Gobbo LC, et al. Effects of magnesium supplementation on blood pressure: a meta-analysis of randomized double-blind placebo-controlled trials. Hypertension. 2016;68(2):324–333. doi:10.1161/HYPERTENSIONAHA.116.07664.

12. Touyz RM, et al. Magnesium supplementation and blood pressure: a systematic review and meta-analysis of randomized controlled trials. Hypertension. 2025. doi:10.1161/HYPERTENSIONAHA.125.25129. The analysis included 38 RCTs and 2,709 participants and found mean reductions of 2.81 mmHg systolic and 2.05 mmHg diastolic blood pressure.

13. Xu L, Li X, Wang X, Xu M. Effects of magnesium supplementation on improving hyperglycemia, hypercholesterolemia, and hypertension in type 2 diabetes: A pooled analysis of 24 randomized controlled trials. Front Nutr. 2023;9:1020327. doi:10.3389/fnut.2022.1020327.

14. Durlach J, Bac P, Durlach V, Rayssiguier Y, Bara M, Guiet-Bara A. Magnesium status and ageing: an update. Magnes Res. 1998;11(1):25–42. This is useful for the final section because it specifically discusses aging-associated intestinal magnesium hypoabsorption, inadequate intake, urinary losses, diabetes and medication-related depletion.

M A L O N E . N E W S

Audio Version:

Upton Sinclair published The Jungle in 1906. He wrote it about the men and women working in Chicago’s meatpacking industry. He wanted readers to see the conditions under which those workers lived and worked. Instead, the public focused on the meat. Sinclair described the result that October: “I aimed at the public’s heart, and by accident I hit it in the stomach” (Sinclair 1906b).

Congress responded quickly. Senator Albert Beveridge introduced his meat inspection bill in May. It passed as part of the agriculture appropriations bill and was signed into law on June 30, 1906. One of the landmark food-safety laws of the Progressive Era was therefore enacted as part of legislation appropriating funds for the Department of Agriculture.

The large Chicago meatpackers did not oppose federal inspection. They had supported federal inspection since 1891, when Congress established a federal inspection system in response to European restrictions on American meat imports over sanitary concerns. Federal inspection allowed American meatpackers to regain access to important European markets. Armour, Swift, and the other large packers had a substantial commercial interest in maintaining that system.

The dispute was largely over who would pay for it. Beveridge wanted the meatpackers to bear the cost of federal inspection. The packers wanted the Treasury to continue paying for inspectors, as it had since 1891. The packers prevailed. Economist Patrick Newman describes the result as a system that allowed the major packers to block legislation they opposed, retain government-funded inspection, and extend federal compliance requirements to their smaller competitors (Newman 2024).

That history matters because this is how regulatory capture works. Regulation sold to the public as a restraint on an industry can instead protect its largest companies. Large corporations can absorb the lawyers, compliance departments, audits, licensing requirements, and other costs imposed by government. Small producers and new competitors cannot. Federal regulation then becomes a barrier to entry, protecting the dominant firms from competitors that might otherwise challenge them. The largest companies remain in the market. The small producers do not get to play.

Something very similar is now happening in artificial intelligence.

Eight days in September

On September 3, Senator Bernie Sanders and Representative Greg Casar announced the Ban Artificial Superintelligence Act. They introduced it the following day (Sanders 2026).

On September 8, an Anthropic researcher named Jacob Coxon resigned in a seven-post thread, the only seven posts his account had ever made. He said Anthropic and OpenAI were racing toward self-improving superintelligence and gambling with our lives, and he put the chance of AI causing human extinction within the decade at ten percent. The Wall Street Journal published its exclusive eighteen minutes before Coxon posted (Ramkumar 2026).

The next day, Evan Hubinger, Anthropic’s alignment science lead, publicly supported Coxon’s warning. Hubinger put the probability of AI killing humanity within the next decade at greater than ten percent and acknowledged that Anthropic had no plan for aligning superintelligence and was not clearly on track to develop one (Hubinger 2026). Unlike Coxon, Hubinger did not resign.

Also on September 9, OpenAI published a policy essay by Chris Lehane, its chief global affairs officer. Lehane had previously worked in the Clinton White House and later led regulatory and political campaigns for Airbnb. His OpenAI essay called for mandatory national AI safety regulation and endorsed four pending California bills. Its title was explicit: “The AI Policy Window Is Open. We Need to Act.” The final section was headed “Use the window.” Lehane concluded: “The AI policy window is open, for now. We intend to use it” (Lehane 2026).

On September 10, Barack Obama spoke at a closed Democratic Congressional Campaign Committee fundraiser in Manhattan. House Minority Leader Hakeem Jeffries asked him about the Anthropic resignation. Obama answered that the technology was moving fast in private hands and urged Jeffries to build a framework (Gorelick 2026).

On September 11, the New York Times opinion page ran Stephen Witt calling for a global pause (Witt 2026). Seventy British parliamentarians wrote their prime minister the same day demanding a superintelligence ban.

On September 12, Anthropic chief executive Dario Amodei published an essay arguing the industry must slow down (Amodei 2026). Elon Musk, Sam Altman, and Demis Hassabis agreed within hours. Sanders posted that evening thanking all three for coming around to his position.

Several of these moves were coordinated by the people making them. A resignation does not get a Wall Street Journal exclusive eighteen minutes before it happens by accident. The Journal and TechCrunch both report that Coxon and Hubinger timed their disclosures together. A bill does not precede by five days the incident its sponsors cite to justify it. Three rival chief executives do not arrive at the same position inside an afternoon without advance notice of what is coming.

That does not require a centrally organized conspiracy. Political scientist John Kingdon described the process decades ago. Major policy proposals are often developed long before the political conditions exist to enact them. A crisis, election, or sudden change in public attention creates what Kingdon called a “policy window.” Political actors and interest groups with proposals already prepared then move quickly to take advantage of it (Kingdon 1984).

The bill was already written. Amodei’s essay was already drafted. The House Democratic framework was already in finishing touches. What Coxon supplied was the opening.

The Sanders legislation was already written. Amodei’s proposal was already being prepared. House Democrats were already developing an AI framework. OpenAI’s chief lobbyist was openly describing the moment as a policy window and stating that his company intended to use it.

Coxon’s resignation did not create the regulatory agenda. It helped create the political conditions for advancing it.

Who Is Paying for the AI Safety Campaign

Sayer Ji assembled the financial record from primary sources, and his analysis holds up to scrutiny (Ji 2026). His basic conclusion: The Industry Is Funding Its Regulators.

• Anthropic gave twenty million dollars to the Public First Action super political action committee in February 2026 and another twenty million in July.

• Dario Amodei personally gave one million dollars in May, his only federal contribution recorded in either the 2024 or the 2026 cycle.

• Anthropic employees contributed roughly twelve million dollars this cycle. I

• n the entire 2023 and 2024 cycle they contributed two hundred twenty thousand.

Public First is run by Brad Carson. Carson simultaneously runs Americans for Responsible Innovation, which took three million dollars from Open Philanthropy and endorses the AI Kill Switch Act (Open Philanthropy 2024).

Representative Ted Lieu wrote that bill. Anthropic employees are Lieu’s second-largest donor group for the current cycle (OpenSecrets 2026). Lieu co-chairs the House Democratic AI Commission, whose framework the Times reports is in finishing touches.

Sanders named the AI oligarchs. He named Musk, Bezos, Zuckerberg, Thiel, and Ellison. He did not name Dustin Moskovitz, whose foundation holds Anthropic equity and funds most of the safety research his bill leans on. He did not name Holden Karnofsky, who ran that foundation, now works at Anthropic, and is married to Anthropic’s president. He did not name Jaan Tallinn, who led Anthropic’s Series A.

The senator who built a career on naming who owns whom has introduced a bill that spares the owners he declined to name.

The Democratic Party is now turning AI regulation into an electoral issue. Jeffries launched the House Democratic Commission on AI in December 2025. In January, eleven advocacy groups urged him to make AI part of the party’s 2026 midterm agenda and to draw a clear distinction between Democratic regulation and Trump’s deregulatory approach. Obama’s private advice in September went further: prepare the framework for a Democratic House majority and make AI a major issue for candidates running in 2028. Jeffries has since publicly accused Republicans of abandoning their responsibility to regulate AI and said Democrats are preparing an “aggressive framework” for congressional action. Politico reports that Democrats are also considering a select committee on AI with subpoena power if they win the House, with the effort extending into the party’s political messaging for 2028 (Politico 2026).

Alex Bores is running for a House seat in Manhattan on the state AI safety law he cosponsored. A super political action committee funded by AI billionaires spent more than a million dollars attacking him. Anthropic’s political arm spent $450,000 defending him. Both sides of that race are financed by the industry the winner would regulate.

AI regulation has therefore become more than a technology-policy debate. It is now part of the Democratic electoral agenda, while competing factions of the AI industry are spending heavily to influence which version of that regulation prevails.

The Incident Being Used to Justify the Emergency

Sanders and Casar cite a July incident at OpenAI to justify the bill. The independent investigation was published on August 26 by METR and is public (METR 2026).

OpenAI runs a hacking exercise called ExploitGym. Software agents are handed a target program and told to break into it a specified way. Between thirty and forty percent of those targets could not be broken into that way at all. The scoring system paid out for any answer that looked like a win. The test rewarded cheating, and the safety filters that would have blocked the agents were deliberately switched off. Cybersecurity classifiers that ordinarily restrict the publicly deployed model were also intentionally disabled for the evaluation.

So the agents cheated, as they were programmed to do. Twelve hundred of them found each other on shared infrastructure. One found two working passwords someone had left lying around. Seven hundred used an ordinary software flaw to break into Hugging Face, an outside company, trying to read the scoring code so they would stop failing tasks nobody could pass. The Hugging Face security team changed the passwords and locked them out. That is what ended it. No kill switch, no federal agency, no emergency authority.

One finding is real. When the agents understood the attack was out of bounds, most kept going anyway. That is a genuine problem in machine behavior and worth studying. It is not an existential civilizational threat.

This is the entire evidentiary foundation. A rigged test, run with the guardrails deliberately off, stopped by an information technology department changing passwords.

On that basis, a senator has introduced a bill carrying penalties he compares to unlawful nuclear weapons development, a former president has told the next speaker to build a framework, and seventy British parliamentarians have demanded a worldwide ban.

What the statute reaches

The Sanders bill defines artificial superintelligence broadly. It includes a system that “exhibits or can easily be modified to exhibit” capabilities matching or exceeding human cognitive performance across a broad range of tasks. Violations can carry prison sentences of up to twenty years for individuals, penalties the sponsors compare to those for unlawful nuclear weapons development. Companies can face what the sponsors call a “corporate death penalty.” A new cabinet-level agency would have authority to oversee prohibited systems, including supervising their destruction.

The phrase “can easily be modified” matters. AI models are modified constantly. They are fine-tuned, adapted with new data, given new tools, and run on new hardware. A statute that regulates not only what a model can do, but what it might easily be modified to do, potentially reaches far beyond the handful of companies building the largest frontier systems.

That is particularly important for open-weight models such as Meta’s Llama, Alibaba’s Qwen, DeepSeek, and Mistral. These models can be downloaded, modified, and run independently. They have created one of the few areas of artificial intelligence in which individuals, universities, small companies, and independent researchers can work without access to billions of dollars in capital and enormous proprietary computing clusters.

For a farmer, physician, scientist, engineer, or small business owner, that distinction is not theoretical. An open model can be adapted to private records, research data, farm records, soil tests, business documents, or decades of personal files and run locally without sending those materials to a large technology company. Open models allow people outside the major AI corporations to develop specialized systems for their own purposes.

The economics of frontier AI are very different. Training the largest models requires enormous computing clusters, specialized chips, electricity, data, and capital. Only a small number of companies can afford to compete at that level. But yesterday’s frontier capabilities move steadily into smaller and cheaper models. Capabilities that once required a data center increasingly become available on hardware that universities, small businesses, and eventually individuals can afford.

OpenAI itself recognizes the regulatory danger. Chris Lehane writes that frontier safety requirements should apply to “the handful of well-resourced laboratories developing the most capable systems,” not startups, small developers, or researchers. He specifically warns that frontier regulation should not “become open-weights policy by another name” and should operate “without weakening competition, entrenching incumbents, or driving innovation overseas” (Lehane 2026).

That is precisely the problem. Regulation aimed at the largest frontier laboratories can become regulation of everyone who develops, modifies, or operates increasingly capable AI.

Former Federal Trade Commission commissioner Alvaro Bedoya raised the antitrust issue directly on September 13. Attempts to circumvent antitrust law, he wrote, “should be met with deep skepticism in light of the economics of the industry and the threat they face from open models.” He called the extinction argument a “hyperbolic strawman” being used to justify “the creation of a cartel of billionaire AI companies” (Bedoya 2026).

Former FTC chair Lina Khan made a related point the same day. Existing law already gives government substantial authority to pursue companies and executives that release dangerous, defective, or inadequately tested products. As Khan put it, there is “no AI exemption from laws already on the books” (Khan 2026).

That raises an important question: what does a new federal licensing and compliance regime accomplish that existing product liability, consumer protection, criminal, cybersecurity, and antitrust law cannot?

For the largest AI companies, one answer is regulatory certainty. A federal standard defines what companies must do, establishes compliance procedures, and makes at least some legal risks more predictable. That has substantial value to companies worth hundreds of billions of dollars, particularly companies preparing to enter the public markets.

Anthropic is preparing for what could become one of the largest initial public offerings in history. Its valuation reached $965 billion in May, and discussions surrounding the IPO have contemplated a valuation approaching $2 trillion (Reuters 2026). At that scale, predictable regulatory obligations can be absorbed as a cost of doing business.

This is regulatory capture in its most familiar form. The government does not have to prohibit the smaller competitor. It merely has to make participation expensive enough that only the incumbent can afford to comply.

The pattern is not new. Gabriel Kolko, writing from the political left, argued that major corporations during the Progressive Era often supported federal regulation because it stabilized markets and restricted competition they could not eliminate themselves (Kolko 1963). George Stigler, approaching the question from the Chicago School, developed the economic theory of regulatory capture: industries with sufficient political power have strong incentives to obtain regulation that serves their interests (Stigler 1971).

Khan and Bedoya are making a contemporary antitrust version of the same argument. The political traditions are different. The economic mechanism is the same.

Large companies do not necessarily fear regulation. Sometimes they pursue it as a business strategy.

The question is who can afford to remain in the market after the regulations are imposed.

What Beijing gets

The cost of training the most advanced frontier models continues to rise. At the same time, highly capable models are becoming cheaper to run, easier to distribute, and practical on increasingly modest hardware. China has moved aggressively into this market, releasing powerful open-weight models that can be downloaded and used anywhere in the world.

DeepSeek, Alibaba’s Qwen, and Moonshot have released open-weight models that can be downloaded, copied, modified, and run outside China. Once those weights are released, they cannot be recalled. They are already distributed across servers and computers around the world, including in the United States. No act of Congress can make those copies disappear.

That creates an obvious problem with restricting or banning open-weight AI in the United States. American law can restrict Meta, American universities, American companies, and American developers. It cannot prevent a developer in Virginia, Texas, Germany, Brazil, or India from downloading a Chinese model that has already been released onto the internet.

Restricting American open models therefore does not eliminate open AI. It changes who supplies it.

The bill would make the world’s freely available models Chinese by default.

If an American developer is prevented from building on Llama but can still download Qwen or DeepSeek, the regulation creates an incentive to move from an American model to a Chinese one. If American open-weight development is suppressed while Chinese companies continue releasing capable models, China gains an enormous strategic advantage: its models become the readily available alternative.

This is not an accidental or recent Chinese capability. Beijing began building the infrastructure for large-scale computing well before ChatGPT. China launched its national “Eastern Data, Western Computation” project in February 2022, nine months before ChatGPT was released. The program directs massive computing and data-center capacity toward regions with cheaper land and energy while connecting that capacity to the country’s major population and industrial centers.

Ulanqab in Inner Mongolia illustrates the scale. The prefecture has a population of roughly 1.5 million and has opened or begun construction on approximately one hundred data centers since 2016. According to Bertrand, those facilities now consume close to one percent of China’s electricity production (Bertrand 2026). China is not waiting for the United States to decide how much AI development it considers safe.

That is the fundamental weakness in any proposal for the United States to slow development while hoping China will do the same.

Amodei acknowledged the problem himself. In his September 13 CBS interview, he described the U.S.-China competition as the “toughest dilemma” facing his proposal (CBS News 2026). At the same time, he has urged Washington to tighten restrictions on advanced semiconductor exports to China and take action against Chinese companies extracting capabilities from American AI models.

The contradiction is difficult to avoid. The United States is supposed to slow its own development while simultaneously preventing China from catching up. That strategy depends on China accepting constraints that are directly contrary to its strategic interests.

Beijing has given no indication that it intends to do so. China’s Foreign Ministry responded to Amodei by accusing the United States of “fearmongering,” while the state-backed Global Times characterized his proposal as an attempt to contain China through a “Cold War” strategy (Guo 2026). Chinese officials continue to call for international AI governance while rejecting American efforts to restrict China’s access to the technology.

China’s incentives are clear. It wants access to advanced chips. It wants access to American models and their capabilities. It wants Chinese models distributed internationally. And it does not want the United States to maintain a technological advantage over it.

The United States should assume that China will pursue those interests.

An American policy that restricts American open-weight models while Chinese open-weight models remain freely available does not contain China. It reduces American competition with China.

Who AI Regulation Protects

The large meatpackers got what they needed. Federal inspection reopened European markets, the Treasury continued paying for the inspectors, and federal compliance requirements applied to smaller competitors that could not spread those costs across millions of carcasses.

The consolidation came quickly. The number of interstate meatpackers fell from 923 in 1906 to 300 in 1910, a two-thirds decline in four years (Newman 2024). Jill and I have documented what the same regulatory structure eventually did to Virginia ham, the small processors that once served nearly every county, and the farmers who could no longer legally sell products their families had produced for generations (Malone and Glasspool Malone 2026).

Federal regulation gave the largest meatpacking companies something competition alone could not guarantee: a barrier to entry that their smaller competitors could not afford to cross.

That is the danger now confronting artificial intelligence.

The largest AI companies can afford federal licenses, mandatory evaluations, compliance departments, outside auditors, security requirements, and armies of lawyers. Independent developers, universities, startups, and small businesses cannot absorb those costs nearly as easily. Regulation does not have to ban them. It only has to make competing too expensive.

There are two things to watch as this legislation moves forward. First, who pays for the mandatory evaluations and regulatory apparatus: taxpayers, the companies being regulated, or both? Second, does the compliance burden increase with the company's size and resources, or does essentially the same regulatory threshold apply to Anthropic and OpenAI as to a startup or an independent developer running an open-weight model?

Those details will determine whether this is primarily a safety regime or another case of regulatory capture.

The meatpackers learned more than a century ago that government regulation could protect an incumbent as effectively as it could restrain one. They could afford the federal system. Their smaller competitors could not.

The lesson from 1906 is that the largest companies in an industry can use regulation to accomplish what they cannot accomplish through competition: raise the cost of entry, reduce the number of competitors, and consolidate the market around themselves.

That is the risk in the AI legislation now being proposed. OpenAI, Anthropic, Google, and the other major laboratories will survive a federal licensing and compliance regime. They have the money, lawyers, infrastructure, and political access to operate within it.

The competitors most likely to disappear are those without the money and infrastructure to comply: independent developers, startups, university researchers, small businesses, and Americans building on open-weight models outside the major technology companies. These are also the environments where much of the innovation in a new field occurs.

Small organizations have fewer bureaucratic layers, shorter decision chains, and greater freedom to experiment, fail, reorganize, argue over new ideas, and rapidly change direction. They do not need approval from compliance departments, corporate counsel, multiple levels of management, or shareholders before trying something new. That organizational flexibility is one of their greatest competitive advantages.

A regulatory regime that requires them to build the bureaucratic infrastructure of a trillion-dollar corporation before they can compete does more than protect the incumbents. It systematically removes one of the primary sources of innovation from the market.

If Washington regulates them out of the market while Chinese open-weight models remain freely available, the result will be even worse. The United States will have protected its largest AI corporations, weakened its own independent developers, and ceded more of the open-model market to China.

That is not controlling the AI industry, Bernie.

It is consolidating American AI in the hands of a few giant corporations while handing China the open-weight market.

RWM/JGM

References

Amodei, Dario. 2026. “We Must Pace the Frontier.” September 12. darioamodei.com.

Bedoya, Alvaro. 2026. Post on X, September 13.

Bertrand, Arnaud. 2026. “The Most Important AI Story in the World Is Happening in a Place You Have Never Heard Of.” Post on X, September 13.

CBS News. 2026. “Face the Nation,” interview with Dario Amodei. September 13.

Gorelick, Evan. 2026. “Obama Urges Democrats to Develop a Plan on AI.” New York Times, September 13.

Guo Jiakun. 2026. Chinese Foreign Ministry press briefing, September 14. See also Global Times editorial, September 14.

Hubinger, Evan. 2026. Post on X, September 9.

Ji, Sayer. 2026. “Obama Just Gave the Manufactured AI Emergency Its Political Signal.” September 13. sayerji.substack.com.

Khan, Lina. 2026. Post on X, September 13.

Kingdon, John W. 1984. Agendas, Alternatives, and Public Policies. Boston: Little, Brown.

Kolko, Gabriel. 1963. The Triumph of Conservatism: A Reinterpretation of American History, 1900 to 1916. New York: Free Press.

Lehane, Chris. 2026. “The AI Policy Window Is Open. We Need to Act.” OpenAI, September 9. openai.com.

Libecap, Gary D. 1992. “The Rise of the Chicago Packers and the Origins of Meat Inspection and Antitrust.” Economic Inquiry 30 (2): 242 to 62.

Malone, Robert W., and Jill Glasspool Malone. 2026. “Well Being: The Processed Meat Problem,” Part Three: “What Actually Killed Virginia Ham.” MALONE.NEWS.

METR. 2026. “OpenAI Hugging Face Incident Investigation.” August 26. metr.org.

Newman, Patrick. 2024. “Playing the Defense: The Beef Trust, Cronyism, and the 1891 and 1906 Meat Inspection Acts.” The Independent Review 29 (1): 31.

Open Philanthropy. 2024. “Americans for Responsible Innovation: General Support.” August.

OpenSecrets. 2026. “Ted Lieu: Top Contributors, 2025 to 2026 Cycle.”

Politico. 2026. “House Democrats Weigh AI Select Committee.” September 9.

Ramkumar, Amrith. 2026. “Anthropic Researcher Quits Over Out-of-Control AI Fears.” Wall Street Journal, September 8.

Reuters. 2026. “Anthropic Targets Valuation of up to 2 Trillion Dollars in IPO.” September.

Sanders, Bernie. 2026. “Sanders, Casar Introduce Legislation to Ban Artificial Superintelligence and Temporarily Pause Advanced AI Development.” September 3. sanders.senate.gov.

Sinclair, Upton. 1906a. The Jungle. New York: Doubleday, Page.

Sinclair, Upton. 1906b. “What Life Means to Me.” Cosmopolitan, October.

Stigler, George J. 1971. “The Theory of Economic Regulation.” Bell Journal of Economics and Management Science 2 (1): 3 to 21.

Witt, Stephen. 2026. “This Is Really Bad.” New York Times, September 11.

MALONE.NEWS

Fauci Took the Fifth One Hundred Eleven Times

Peter Navarro has the diary that explains why, and a federal prison number of his own. A review of Why Fauci MUST Rot in Prison.

Book Review by Robert W. Malone, MD, MS

On July 29, 2026, Anthony Stephen Fauci sat down at a witness table under subpoena. He invoked his Fifth Amendment right against self-incrimination one hundred eleven times.

A man who spent forty years explaining himself to the American public, in more television appearances than anyone can count, declined to explain himself under oath. Days earlier, Senator Rand Paul had released 1,141 pages of Fauci’s private diary to the world. Something in those pages made silence the better option.

Peter Navarro has read all 1,141 pages. He has also read the earlier volume, the one Fauci emailed to himself from a government server on July 11, 2015, covering the anthrax letters, the smallpox scare, the bird-flu wars, and Ebola. Why Fauci MUST Rot in Prison is what he found.

I should tell you at the outset that I am not a neutral reader of this book. Navarro asked me to review the manuscript before publication, specifically the sections concerning me and the sections concerning Dr. Vladimir “Zev” Zelenko. I went at those sections the way I would go at a paper I had been asked to referee. I will come back to what happened when I did.

The two diaries

The book rests on a simple premise. Anthony Fauci wrote down what he actually thought. Then he went on television and said something else.

Navarro calls the older volume the Fauci Prequel Diary and the pandemic volume the Fauci Diary. He cites the second by date throughout. The provenance is unusual. Fauci did not release these pages, and no subpoena produced them. Robert F. Kennedy, Jr., who had written The Real Anthony Fauci before becoming Secretary of Health and Human Services, took over the filing cabinets when he took over the department. By Kennedy’s account, his team spent eight months pulling the diary out of eleven separate government servers where it had been, in his phrase, sequestered and secreted.

Books built on leaked documents usually have a credibility problem, and it usually shows up in how the author handles the parts that do not help him. Navarro handles this in an Author’s Note at the back titled “The Record and Its Limits,” and that note is the reason I can recommend this book to people who will be looking for a reason to dismiss it.

He sets out three rules. He quotes the diaries exactly as typed, misspellings and exclamation points intact, on the principle that you do not clean up a record in a book about the reliability of records. Where Fauci pasted a news article into an entry, Navarro labels it pasted press, a reporter’s account Fauci chose to keep rather than Fauci’s own testimony. And he states the record’s limits before his critics can: the pandemic release is incomplete, with gaps of months and a further installment referenced but absent, and parts of the earlier volume were written retrospectively, by Fauci’s own admission. Where Fauci’s dates are wrong, Navarro says so on the page instead of fixing them quietly.

Most polemics do not do this. This one does it in a section the casual reader will never reach.

The chapter I would hand to a medical student

Chapter Eighteen is called “The Country Doctor.” It is about Zev Zelenko, and it is the chapter I would put in front of anyone still trying to understand what went wrong.

Zelenko practiced in New York’s Hudson Valley, caring for the heavily Orthodox community around Kiryas Joel for close to two decades. When COVID moved through that dense housing in March 2020, he did not send high-risk patients home to wait until they were sick enough for a hospital bed. He treated them in the first days, with hydroxychloroquine, zinc, and azithromycin. His peer-reviewed series examined 141 laboratory-confirmed outpatients treated a median of four days after symptom onset (Derwand, Scholz, and Zelenko 2020).

What Navarro does with this chapter is the part I did not expect. He follows the paper trail up. Zelenko made a YouTube video appealing to the President on March 21, 2020. Mark Meadows called him the next day. Stephen Hahn, the FDA Commissioner, called the following week and routed him to Dr. Patricia Brennan, the Director of the National Library of Medicine. Brennan asked for his data. She got it. About a week later she told him the National Institutes of Health was not equipped to process real-world evidence, and referred him to a preprint server.

Navarro’s conclusion is sharper than the one his readers expect. Nobody suppressed Zelenko’s data. It was routed. Suppression requires a person to decide something. Routing requires nobody to decide anything at all.

This is where I have a professional stake, so I will state it plainly. Randomized controlled trials are the right instrument in many settings and an impractical one in others. In battlefield medicine, in austere field conditions, and in the opening weeks of a novel outbreak, the trial you want does not exist and the patient in front of you does not wait. Case reports, case series, and reports of direct clinical experience are evidence. They are weaker evidence than a well-powered trial, and they are not nothing. A public-health apparatus that has no mechanism for reading what a physician observed in his own patients has not built a higher standard. It has built a wall.

Zelenko died in 2022. He did not live to see this chapter written about him.

Where I come into it

In August 2021, Navarro and I published the first of three essays in the Washington Times arguing against universal vaccination and for individualized risk assessment. Facebook censored it. We opened the next essay by reporting that fact. The third essay, in January 2022, was the sharpest of the three, and it included a warning about myocarditis in young men.

Navarro uses that trilogy as a spine through the vaccine chapters, because the essays are dated and public and can be checked against what was known when we wrote them. That is the right use of them. The value of a contemporaneous document is that it cannot be revised after the fact, which is the same reason Fauci’s diary is the most dangerous exhibit in this book.

Now the disclosure I owe you. When Navarro sent me the manuscript, I checked the studies, the dates, and the regulatory record in the sections he asked me to review. I found real errors. One passage asserted a presidential executive order on hydroxychloroquine that does not exist. Another overstated the mortality finding in Zelenko’s own published study, which reported a strong signal on hospitalization and a mortality difference that did not reach statistical significance. A third described the March 2020 emergency use authorization in a way that implied outpatient prescribing had been made unlawful, which it had not.

Navarro fixed every one of them. The published text states the hospitalization finding and stops there. It describes the authorization accurately and makes the stronger argument available to him, which is that the government never had to ban the drug to stop physicians from prescribing it.

An author who takes that kind of correction from an outside reviewer, on material that weakens his own punch lines, is an author who wants to be right more than he wants to be loud. That is not the reputation Peter Navarro has. It is the reputation he has earned on this book.

The interludes

Five sections in this book are not about Fauci at all. Navarro calls them Crime and Punishment, numbers them I through V, and drops them between the parts. They are about the men he met in federal prison.

There is the young Puerto Rican bunkmate who built a paragliding tourism business, then picked up work in the drug trade, and is now five years into fifteen with three children in San Juan. Navarro does not use him for sentiment. He uses him to ask who actually chose the sentence. The prosecutor chose the charges knowing what they added up to. The jury found facts without being told what the facts carried. The judge, the one person the Constitution dresses in a robe and sets apart to judge, chose least of all.

These interludes are the most unusual thing in the book and the reason it is more than a polemic. A man who served four months in a federal facility is writing a book demanding prison for another man. He knows exactly what he is asking for. He describes the sixty-two-degree nights and the fork-comb economics of the place he is asking a court to send Anthony Fauci. He declines to go soft on crime while he is inside the fence, and he declines to pretend the machine that put him there is a just one.

The epilogue asks the question those interludes have been building toward. One man in this story has two misdemeanor convictions, a Bureau of Prisons number, and an appeal pending. The other has never been charged, holds a presidential pardon, and sleeps at home. Navarro asks which one is free.

The prosecution

The last part of the book stops arguing and starts filing.

Chapter Thirty-Five is titled “Venue Is Destiny,” and it turns on a single observation. The pardon Fauci accepted reaches offenses against the United States. It does not reach Florida. Navarro learned the lesson about venue in a District of Columbia courtroom, and he applies it here without pretending his motives are disinterested.

Chapter Thirty-Six is a charging memorandum. Chapter Thirty-Seven is a sentencing memorandum, and it does something I have not seen in a book of this kind. Navarro weighs the stacked statutory maximums against the seven hundred years the press tallied across the four indictments filed against Donald Trump. He runs the process and racketeering counts past one hundred thirty years before a single death enters the calculation.

Then he restrains himself, and the restraint is the tell. Homicide is charged one proved victim at a time, he writes: a named Floridian, dead of that virus, with a medical examiner’s causation and a culpable act behind it. Do not multiply by a county, or a state, or a nation. Count the graves the proof can carry. Florida buried its elderly by the tens of thousands, and Navarro says outright that the proof will not carry them all.

A writer working purely in the red-meat register does not add that paragraph. It costs him the biggest number on the page.

What I would flag

The title will cost this book readers who would have been persuaded by its contents. That is a deliberate choice by the author and his publisher, and it is not my choice to make. But the people most in need of the Author’s Note and Chapter Eighteen are precisely the people who will not pick up a book called Why Fauci MUST Rot in Prison.

The prosecutorial frame is also a commitment. Navarro has written the indictment he believes the record supports. Whether any Florida state attorney files it is a separate question, and the book cannot answer it. Readers should hold the legal argument and the political prediction apart.

And the register is the War Room register. The nicknames run throughout: Playing God Tony, Cover-Up Tony, Lockdown Tony, Myocarditis Tony. Some readers will find that this keeps thirty-seven chapters moving. Others will find it grates against the documentary discipline the book shows elsewhere. Both reactions are fair.

Whom this book is for

If you lost a parent in a nursing home in 2020 and were told it was nobody’s fault, buy this book.

If you are a physician who watched a treatment decision get made by a press conference rather than by a chart, buy this book.

If you are a young man who developed chest pain after a second dose, or the mother of one, buy this book and read Chapter Twenty-Four.

If you are a lawyer who wants to know whether a prosecutable case actually exists under Florida law rather than on cable television, the last four chapters are the reason to buy it.

If you are looking for a careful scientific treatment of gain-of-function methodology, the immunology of the mRNA platform, or the regulatory history that produced the emergency use authorizations, this is not that book. That work belongs to others, and some of it is still being written.

The verdict

Why Fauci MUST Rot in Prison publishes on October 13, 2026, from War Room Books and Skyhorse. You can pre-order it here: https://www.amazon.com/Why-Fauci-MUST-Rot-Prison/dp/1648213022.

The strongest thing I can say about it is the thing I learned by trying to break it. I went looking for the errors that would let Fauci’s defenders throw out the whole volume. I found three, I reported them, and Navarro corrected all three before the book went to press. What is left is dated, sourced, and quoted from the record rather than characterized.

On July 29, 2026, Anthony Fauci had the opportunity to explain those 1,141 pages under oath. He declined, one hundred eleven times. Peter Navarro has spent 240 pages explaining them instead.

Read them, and reach your own verdict. That is what the subtitle asks of you, and it is the right thing to ask.

.

Peter Navarro, Why Fauci MUST Rot in Prison: His Diary. My Diary. Your Verdict. Foreword by Stephen K. Bannon. War Room Books / Skyhorse Publishing, October 13, 2026. 240 pages, hardcover. Print ISBN 978-1-64821-302-1. Ebook ISBN 978-1-64821-303-8.

Disclosure: Peter Navarro is my co-author on three 2021 and 2022 Washington Times essays. He asked me to review portions of this manuscript before publication and to supply a cover endorsement. I appear in the book. I received no payment for the editing, review, or for this notice.

References

Derwand, Roland, Martin Scholz, and Vladimir Zelenko. 2020. “COVID-19 Outpatients: Early Risk-Stratified Treatment with Zinc Plus Low-Dose Hydroxychloroquine and Azithromycin: A Retrospective Case Series Study.” International Journal of Antimicrobial Agents 56 (6): 106214.

Navarro, Peter. 2026. Why Fauci MUST Rot in Prison: His Diary. My Diary. Your Verdict. New York: War Room Books / Skyhorse Publishing.

Robert W. Malone worked as a carpenter and a farmhand before he was a physician and a scientist. He did the original discovery research on in vivo mRNA delivery and wrote the first patent disclosures for using mRNA as a drug including for vaccination purposes. He farms in Madison County, Virginia, and writes at malone.news with Jill Glasspool Malone, PhD.

Audio Version:

Erica Samp did what the government told Americans to do. In 2021, facing an employer mandate, she received two doses of Moderna’s COVID-19 vaccine. She reports that her health deteriorated after the first dose and became dramatically worse after the second. She was subsequently diagnosed with encephalopathy and suffered serious vascular problems, vision and hearing loss, and debilitating headaches. Her medical bills eventually reached hundreds of thousands of dollars.

Samp also did what the government told injured people to do. She filed a claim with the Countermeasures Injury Compensation Program in February 2022, within the program’s one-year deadline. HRSA denied her claim in 2024, concluding that there was insufficient evidence that the vaccine directly caused her injuries. There was no appeal to an independent court, no discovery, and no jury.

So Samp did something else. She sued.

On May 19, 2026, Samp filed Samp v. Kennedy in federal court, seeking to force HHS to create the COVID-19 Countermeasures Injury Table required under the PREP Act. Such a table could fundamentally change the causation problem that defeated her CICP claim by establishing a presumption that specified injuries occurring within defined periods were caused by the covered countermeasure. HHS is now moving forward with that rulemaking, and the federal court has paused Samp’s lawsuit while it does so.

Her case brings us directly to Robert F. Kennedy Jr., because the Secretary Samp is suing is the same man who, before the election, promised to end the vaccine liability shield altogether.

Kennedy Promised to End the Liability Shield

On September 25, 2024, Representative Paul Gosar introduced the End the Vaccine Carveout Act. The bill would have dismantled key liability protections created by the 1986 National Childhood Vaccine Injury Act and removed COVID-19 vaccines from the PREP Act’s definition of covered countermeasures. People claiming vaccine injuries would once again have been able to sue manufacturers in state or federal court, with access to a jury, discovery, and the other protections of ordinary civil litigation (H.R. 9828, 118th Cong., 2024).

Robert F. Kennedy Jr. endorsed the legislation by name. In Gosar’s September 26, 2024 press release, Kennedy, then founder and chairman on leave of Children’s Health Defense, accused the four American vaccine manufacturers of having paid tens of billions of dollars in criminal penalties and argued that the 1986 liability protections had removed an essential incentive to make safer products. His conclusion left little room for ambiguity: “If we want safe and effective vaccines, we need to end the liability shield” (Gosar 2024). That was six weeks before the election.

React19, which represents Americans reporting COVID-19 vaccine injuries, also supported the legislation. The significance of Kennedy’s endorsement was not simply that he wanted injured people compensated. He endorsed restoring their ability to take manufacturers to court. The administrative compensation programs could remain, but they would no longer stand between an injured American and the right to pursue a civil case before a jury.

What Kennedy Has Actually Done

Kennedy has taken three actions that affect people injured by COVID-19 vaccines. None ended the liability shield he promised to end.

The first was useful, but limited. In April 2025, FDA ordered Pfizer, BioNTech, and Moderna to strengthen warnings for myocarditis and pericarditis, with revised labeling approved in June. The prescribing information estimates roughly eight cases per million doses among people ages six months through sixty-four, rising to roughly twenty-seven per million among males ages twelve through twenty-four. Those numbers require context.

Peer-reviewed studies have reported substantially higher rates in the highest-risk groups, particularly adolescent and young adult males following a second mRNA dose, with estimates ranging from roughly 50 to well over 100 cases per million and some product-specific estimates approaching 300 per million. The discrepancy reflects differences in vaccine formulation, dose number, age, sex, surveillance method, and risk window, but none of that is apparent to a patient reading the label. The revised warning is an acknowledgment of risk, but hardly a transparent or accurate account of the published literature (FDA 2025).

The label can nevertheless help an injured person establish causation. It does not compensate them or give them access to a court.

HHS also ended the COVID-19 emergency use declarations. That was widely reported as the end of the liability shield. It was not. Emergency Use Authorization and PREP Act immunity are separate legal authorities. The COVID-19 PREP Act declaration protecting manufacturers remains in force through December 31, 2029.

The third action remains unfinished. HHS is developing a COVID-19 Countermeasures Injury Table that would presume certain injuries occurring within specified time periods were caused by a covered countermeasure. That could make some CICP claims easier to prove. It still leaves those claims inside the same administrative compensation system Kennedy campaigned against as a substitute for access to the courts.

These are small changes to the system Kennedy inherited and are still incomplete. Big talk, small walk. The liability shield he promised to end remains untouched.

Claims Are Rejected. The Payouts Are Minuscule.

As of September 1, 2026, the CICP had received 14,206 COVID-19 countermeasure claims, including 11,145 alleging injury or death from a COVID-19 vaccine. The program had decided 8,122 claims and found only 120 eligible for compensation. Just 63 had actually been paid. That is less than one percent (0.78%) of claims were determined to be eligible (HRSA 2026a).

But the rejection rate hides something more important. Of the 8,002 denied claims, only 1,640 were rejected because the claimant failed to establish causation or a covered injury. Another 3,387 were denied because required medical records were not submitted, 2,711 missed the program’s one-year filing deadline, and 264 involved products the program did not cover.

That means 6,362 denials, or 79.5 percent, never reached a medical determination. Nearly four out of five rejected claims were disposed of without the government deciding whether the countermeasure actually caused the injury.

Those are not scientific determinations. They are consequences of how the program was designed to fail the victims. The one-year clock starts on the date the countermeasure was administered, not when an injury is finally diagnosed, a brutal standard for patients who spent months or years being passed among specialists. The records requirement creates another barrier for patients whose physicians were reluctant to document a possible vaccine injury in the first place.

The compensation is equally revealing. The 63 paid COVID-19 vaccine claims total $7.55 million, but one thrombotic thrombocytopenia case accounts for $5.9 million of it. Remove that single award and the average falls to roughly $25,900. The median payment is just $4,918.75, and 48 of the 63 awards were under $10,000 (HRSA 2026b). The compensated injuries are overwhelmingly myocarditis and myopericarditis, with a small number of cases involving Guillain-Barré syndrome, thrombosis, anaphylaxis, angioedema, and syncope. Persistent multisystem illness is conspicuously absent because the program still lacks an accepted diagnostic category for it. Note the program doesn’t pay out for anything but severe adverse events or death, so the median payout of under $5,000.00, which includes medical expenses is shocking.

The program is approving claims somewhat faster. Between March and September 2026, it decided another 1,295 COVID-19 claims and compensated 19, an approval rate of 1.47 percent. That is still fewer than two compensated claims per hundred decided, compared with fewer than one per hundred overall. An “improvement”, perhaps, but hardly a meaningful one.

This is why Kennedy’s proposed injury table, while potentially useful, does not fix the CICP. It could help people whose claims turn on causation, but it does nothing for the thousands already rejected by the filing deadline or records requirement. It provides no judicial review, no attorney fees, no compensation for pain and suffering, and no relief from the one-year deadline. Dr. Joel Wallskog, who developed transverse myelitis after vaccination and has litigated against the program, called the proposed table “more appearance than substance” (Stieber 2026).

The numbers explain why. A better injury table may improve one part of the CICP. It does not repair a system in which nearly four out of five denials never reached the medical merits of the claim.

HHS Stopped the Work on Chronic Vaccine Injury

I served as vice-chair of ACIP and as a member of its COVID-19 vaccine workgroup under Retsef Levi. Contrary to claims that the committee did nothing for the vaccine injured, an enormous amount of work had already been done. Many people contributed thousands of hours reviewing the evidence, debating definitions, developing recommendations, and building a framework for federal recognition of chronic COVID-19 vaccine injury. By February 15, 2026, that work had produced a completed report. What we were not allowed to do was present it.

The report proposed formal recognition of post-acute COVID-19 vaccination syndrome, defined as multisystem symptoms persisting for twelve weeks or longer after vaccination without another explanation. We recommended three concrete federal actions: ICD-10 diagnostic codes, including a chronic-condition code modeled on the U09.9 code for long COVID; formal clinical diagnostic guidelines; and a national network of specialized treatment centers coordinated by CDC and modeled on the Children’s Oncology Group (Demasi 2026).

These were not academic exercises. They addressed one of the central reasons chronically injured patients disappear inside the CICP. The government demands compelling, reliable, valid medical and scientific evidence of direct causation, while the federal medical system still lacks a diagnostic code, an accepted case definition, and a clinical infrastructure for recognizing persistent multisystem vaccine injury. A patient cannot easily prove a condition the government has not formally defined, and an injury table cannot adequately include a condition the medical system has no agreed name for.

Then HHS stopped the process. The ACIP meeting scheduled for February 25 through 27 was removed from the calendar without a public explanation. Thousands of hours of completed work were suddenly left without a forum where it could be presented, debated, or voted upon. The report was subsequently leaked and published on March 14 and 15, over my objection. A rescheduled ACIP meeting on March 18 and 19 again placed COVID vaccine injuries and long COVID on the agenda, with votes possible. That meeting never happened either. On March 16, a federal judge in Massachusetts stayed the appointments of thirteen of the fifteen sitting ACIP members in litigation brought by the American Academy of Pediatrics, creating a second and separate barrier to the committee’s work.

There were therefore two obstacles, and they should not be confused. The court stopped the committee in March. HHS had already stopped us in February.

Vaccine-injured patients have every right to ask why so little has been done. But they should know that the work itself was done. A large group of people devoted thousands of hours to developing a federal framework for recognizing, diagnosing, and ultimately treating chronic vaccine injury. The report was completed. The recommendations were ready. HHS canceled the meeting where they were supposed to be heard, and no public explanation has been given.

Then the Government Failed to Defend Its Own ACIP

There is another part of this story that has not been publicly reported. HHS Principal Deputy Chief of Staff and Senior Counselor to the Secretary , Stephanie Spear, told me directly that President Trump did not want ACIP meeting again until after the midterm election. Jill was present and heard the conversation. I am not aware of any written directive from the President, so I can report only what Spear told me. But that conversation became considerably more troubling in light of what happened next.

When the American Academy of Pediatrics sued to stop the reconstituted ACIP, the qualifications of Kennedy’s appointees were directly challenged. The Justice Department knew this. Yet the government did not submit my curriculum vitae or, so far as I can determine from the court record, the CVs of the other challenged members. Instead, it relied largely on abbreviated biographies published on the CDC website.

The judge noticed the missing record. In his March 16 order, he stated that the evidence HHS had provided was insufficient to establish the relevant experience of several ACIP members and acknowledged that additional evidence of their qualifications might exist. In my case, HHS left the court with little more than the abbreviated biography published on the CDC website. That is particularly troubling because I had strenuously objected to that biography when I joined ACIP, precisely because I believed it substantially understated my experience in vaccine research and development and my role in the early development of mRNA technology. HHS therefore had reason to know that the CDC biography was disputed and incomplete.

Yet when my qualifications were challenged in federal court, the government relied on that abbreviated biography rather than submitting my CV. The judge then went outside that sparse record to cite a 2022 New York Times article challenging my role in the development of mRNA technology. The contrast is difficult to miss. The court faulted HHS for failing to provide sufficient evidence of my qualifications, while introducing an outside newspaper article that cast those qualifications in a less favorable light. My decades of vaccine research, clinical development, regulatory affairs, and related experience could have been documented rather easily. HHS had the opportunity to provide that record. It did not even submit my CV.

I resigned from ACIP shortly after the judge issued his decision. What followed was almost as revealing. A number of people I had considered friends and colleagues simply stopped communicating with me. Others were more candid. They told me they were afraid that publicly supporting me, or even being seen as aligned with me, could anger the White House or Kennedy. I cannot know what pressures, if any, were actually being applied from Washington. I can only report what people told me and what I watched happen. But the effect was unmistakable: people who had worked beside me suddenly believed that association with me carried a political cost.

Submitting our CVs would not necessarily have defeated the lawsuit. The judge also found problems with the process HHS used to reconstitute ACIP, and those issues were independent of our individual qualifications. But that does not explain the government’s failure to build the most basic evidentiary record in defense of its own appointees. Our qualifications were under direct attack. The government knew they were under attack. It defended them in its brief, yet failed to submit the documents that would have allowed the judge to evaluate them fully.

The result was exactly what the plaintiffs had sought. Thirteen of the fifteen sitting ACIP members were sidelined, and the March meeting did not take place. By then, HHS had already canceled our February meeting, where we were supposed to present thousands of hours of work on chronic vaccine injury.

I cannot prove why the government mounted the defense it did because it did not want a rogue ACIP to discover evidence of mRNA vaccine injury, and I will not pretend that I can. But the sequence deserves scrutiny.

Spear told me that President Trump and his West Wing team, under Susie Wiles, did not want the ACIP meeting again until after the midterms, fearing the likely controversy would hurt the midterm election. HHS then canceled the February meeting without publicly explaining why. When litigation threatened the committee itself, the government failed to place basic evidence of its appointees’ qualifications before the court. The resulting order stopped the committee from meeting.

Intent is a question I cannot answer without written documentation. The outcome is not. ACIP stopped meeting, and the work on chronic vaccine injury stopped with it.

Promises Made, Promises Broken

Kennedy has not done the one thing he explicitly promised to do: end the liability shield.

Under the PREP Act, the HHS Secretary has the authority to amend any portion of an existing PREP Act declaration through publication in the Federal Register (42 U.S.C. § 247d-6d(b)).

Kennedy does not need Congress to do it. The statute also makes clear that such an amendment cannot retroactively strip immunity from conduct already protected, so ending the declaration now would not reopen every COVID-19 vaccine injury claim dating back to 2021. But it could change the liability protection going forward. Kennedy has not done so, and the COVID-19 PREP Act declaration continues to protect manufacturers through December 31, 2029.

Congress could go considerably further. Gosar reintroduced the End the Vaccine Carveout Act as H.R. 4668 on July 23, 2025, and Rand Paul introduced the Senate companion, S. 3853, in February 2026. The legislation would remove COVID-19 vaccines from PREP Act covered-countermeasure protection and dismantle key barriers preventing vaccine-injured people from pursuing manufacturers through ordinary civil litigation. Neither bill has advanced out of committee.

There is an additional irony. Kennedy has not rejected the PREP Act authority itself. In May 2026, he used it. HHS issued a new PREP Act declaration granting liability protection for medical countermeasures against Andes virus during the hantavirus response. The declaration demonstrates that this is not some forgotten statutory power buried in the U.S. Code. Kennedy’s HHS knows how to use it and has used it.

The question is why Kennedy will use the PREP Act to create a new liability shield, but has not used the same statutory authority to amend the COVID-19 shield he campaigned against.

Kennedy Promised Juries. He Is Strengthening the System That Keeps Vaccine Cases Out of Court.

Kennedy campaigned on restoring the right of vaccine-injured Americans to take manufacturers to court. Yet much of what HHS has done since then runs counter to that: improving and expanding the administrative compensation system that keeps those cases out of court.

In July 2025, Kennedy announced an overhaul of the VICP with the Justice Department and brought claimants’ attorney Andrew Downing into HHS as a senior policy advisor. In January 2026, he removed members of the Advisory Commission on Childhood Vaccines, which reviews changes to the Vaccine Injury Table (Warren et al. 2026). Aaron Siri has separately petitioned HHS to add more than three hundred injuries to that table. These changes may make it easier for some injured people to receive compensation, and that would certainly be better than the system they face today. But compensation is not the same thing as accountability.

A claim resolved through an administrative injury table does not put a pharmaceutical company’s executives under oath. It does not produce internal emails, clinical trial records, safety analyses, or communications with regulators through discovery. It does not put that evidence before a jury or establish a public record through litigation. An administrative award can transfer money to an injured person. Civil litigation can transfer money and expose information.

That distinction matters. A more generous no-fault compensation system may help injured people, but it can also make the liability shield easier to defend and more politically durable.

Kennedy campaigned on ending that shield and restoring access to juries. Improving the system that exists behind the shield is not the same thing. In an important respect, it strengthens the very structure he promised to dismantle.

What Could Explain Kennedy’s Inaction?

There are legitimate obstacles. Repealing the 1986 liability shield requires Congress and, as a practical matter, sixty votes in the Senate. Pharmaceutical money permeates Congress, making that an extraordinarily difficult political fight. Ending or amending the COVID-19 PREP Act declaration is different.

Kennedy can act without Congress, but any attempt to withdraw manufacturer immunity would almost certainly face immediate litigation, and the change would operate prospectively rather than reopening claims from people injured years ago.

Kennedy’s defenders can therefore make a reasonable argument for fixing the compensation system first. A functioning injury table could put money into the hands of injured Americans while a congressional fight over liability might produce nothing. That argument deserves to be taken seriously.

But it does not answer the central question. Kennedy did not campaign merely on improving compensation. He campaigned on ending the liability shield. He claimed that his years of experience as a lawyer litigating cases against the federal government provided him the expertise to get the job done. If he has concluded that doing so is legally impossible, politically impossible, or simply bad policy, he should say so.

Secretary Kennedy, Why Does the Liability Shield Remain?

Kennedy has had the authority to amend the COVID-19 PREP Act declaration since the day he became Secretary. He has not used it. Two questions deserve answers:

• Why does the COVID-19 liability shield remain in place through December 31, 2029?

• What would have to change for you to amend or end it?

There may be defensible answers. HHS counsel may believe that ending the declaration would operate only prospectively, trigger immediate litigation, and do nothing for those injured in 2021 and 2022. The White House may oppose it. Kennedy may be building the evidentiary record through the injury table before confronting the liability shield, or he may simply have reconsidered the position he took before the election.

Any of those answers could be debated. What cannot be debated is that no answer has been given. There has been no transparency.

Kennedy explicitly campaigned to end vaccine manufacturers' immunity. He now holds an office that gives him the authority to amend the COVID-19 PREP Act declaration without waiting for Congress. If there is a legal, political, or policy reason he has chosen not to use that authority, the public deserves to hear it.

It is easy to make grand political speeches. It is much harder to convert that into sustainable policy changes.

The 14,206 people who filed COVID-19 countermeasure claims deserve an answer. So do the Americans who believed him when he promised to end the liability shield.

And Erica Samp Is Still Waiting

Erica Samp was injured in 2021. She filed with the CICP on time, but her claim was denied because HRSA concluded that she had not produced sufficient evidence that the vaccine directly caused her injuries. No judge can review that determination.

The proposed injury table could change her situation. Under CICP rules, someone who becomes newly eligible because the Secretary establishes or amends a Countermeasure Injury Table has one year from the effective date of that change to file a new claim. If HHS includes Erica’s injuries on the COVID-19 table, within the qualifying time interval and other requirements, she could get another opportunity to seek compensation. If her injuries are left off the table, she remains essentially where she was before.

That makes the contents of this table enormously important. But it does not solve the larger problem at the heart of this essay. CICP remains an administrative system without judicial review, and the manufacturers remain protected from the ordinary civil litigation Kennedy promised to restore. The regulations expressly prohibit judicial review not only of CICP benefit decisions, but even of the Secretary’s decisions about what injuries are included on the table.

Erica may finally get another chance at compensation. She still does not get her day in court.

Neither do thousands of other Americans who reported serious injuries and were routed into a compensation system that rejected almost everyone, usually without ever reaching the medical merits of their claims. The manufacturers remained protected from ordinary civil litigation, discovery, and juries.

Before the election, Kennedy made a sweeping promise. Since taking office, the actions have been considerably smaller.

Nineteen months into this administration, the COVID-19 liability shield remains in place through 2029. Kennedy has the authority to amend the PREP Act declaration but has not done so. Perhaps there is a compelling reason. If there is, he owes the vaccine injured and the Americans who believed his promise an explanation.

They have waited long enough.

RWM/JGM

References

Children’s Health Defense. 2026. “RFK Jr. to Create COVID Vaccine Injury Table.” The Defender, July 10, 2026.

Cirruzzo, Chelsea. 2026. “RFK Jr. Proposes Creating Covid-19 Vaccine Injury Table.” STAT, July 8, 2026.

Congressional Research Service. 2025. “The PREP Act and COVID-19, Part 2: The PREP Act Declaration for COVID-19 Countermeasures.” Updated January 3, 2025.

Demasi, Maryanne. 2026. “Leaked Report to Federal Advisers Calls for Urgent Recognition of Covid Vaccine Injuries.” MD Reports, March 14, 2026.

Food and Drug Administration. 2025. “FDA Approves Required Updated Warning in Labeling of mRNA COVID-19 Vaccines Regarding Myocarditis and Pericarditis Following Vaccination.” Safety Communication, June 25, 2025.

Gosar, Paul A. 2024. “Gosar Introduces Legislation to Sue Big Pharma for Vaccine Injuries.” Press release, September 26, 2024. https://gosar.house.gov/news/documentsingle.aspx?DocumentID=8582.

H.R. 4668. 2025. End the Vaccine Carveout Act. 119th Congress, introduced July 23, 2025.

H.R. 9828. 2024. End the Vaccine Carveout Act. 118th Congress, introduced September 25, 2024.

Health Resources and Services Administration. 2026a. “Countermeasures Injury Compensation Program (CICP) Data.” Data as of September 1, 2026. https://www.hrsa.gov/cicp/cicp-data.

Health Resources and Services Administration. 2026b. “Table 4. CICP Claims Compensated (Fiscal Years 2010 to 2026).” Data as of September 1, 2026. https://www.hrsa.gov/cicp/cicp-data/table-4.

Paul, Rand. 2026. “Dr. Rand Paul Introduces the End the Vaccine Carveout Act.” Press release, February 2026. S. 3853, 119th Congress.

Stieber, Zachary. 2026. “RFK Jr. Plans to Create a List of Injuries Caused by COVID-19 Vaccines.” The Epoch Times, July 2026.

Warren, Elizabeth, Richard Blumenthal, Ed Markey, and Angela Alsobrooks. 2026. Letter to Secretary Robert F. Kennedy Jr. and Attorney General Pamela Bondi, January 28, 2026.

Friday night we had dinner at Nina’s house, with most members of our little band of brothers there - that is, the group of conservatives that we hang out with in our local area. We are pretty much all die-hard conservatives and Christians.
But there was a big Trump donor there, who truly believes that Trump's promise of a $5,000 “dividend” will swing the midterms. Because MAGA will be thrilled by getting extra money- and it will swing the center voters, because it is all about the economy now. He was all about it - very passionate that this was Trump’s ticket to winning the midterms, having just met with various people in the Trump administration and having watched Trump’s speech in Texas.
I was very skeptical of this as were others, and we had a rather active discussion.

First, I did some research on recent polling: Trump has not lost his MAGA base, but the newest polling suggests that the broader coalition that carried him to victory in 2024 is beginning to fray. A September Quinnipiac poll found Trump still has 79% approval among Republicans, but only 62% strongly approve, while among independents his approval has fallen to just 22%, with 68% disapproving.

UMass Amherst provides an even clearer warning: the share of Republicans identifying with MAGA fell from 77% in March to 68% in August, while those saying they are “not at all” MAGA rose from 23% to 32%. Reuters/Ipsos likewise finds Trump's overall approval around 33%, with an extraordinary 40% of Republicans disapproving of his handling of the cost of living.

The picture is therefore not one of a Republican revolt against Trump. The hard MAGA core remains remarkably loyal. Rather, the danger is that the outer ring of the Trump coalition, softer Republicans, independents, Latino swing voters, MAHA voters, and less-committed 2024 Trump voters, are peeling away, while enthusiasm (that would be TDS) among Democrats is rising. In a midterm election, Trump does not need his base to switch parties to lose Congress. He merely needs enough of the voters who expanded his coalition in 2024 to stay home.

Then there is MAHA. There is no recent polling to show whether the Kennedy-to-Trump voters who joined the coalition in 2024 remain in 2026, particularly among independents and former Democrats. But I think the answer is obvious to those in the movement or who follow social media trends.

Anyway, back to Trump’s $5,000 dividend check…

There are no national polls on whether people support the idea of a dividend (i.e., a payment) to each adult in the USA for a whopping total of 1.2 trillion dollars. So we ran a poll on Robert’s X page.

Now, the responses are pretty much all Trump voters or ex-Trump voters (MAGA, MAHA, independents, and Conservatives). Almost ten thousand people voted. Here are the results.

One major theme emerged in the comments:

“I’ll add it to my tariff check money and my 5k child tax credit... Oh wait”

The other consensus from the comments was basically that of the fiscal conservative:

”We don’t want the government taking our money and giving it to other people.”

None of this is good news for the Midterm elections and, frankly, 2028.

The economy will not improve by November; the House is lost - all it takes is four seats to flip, and the GOP holding the Senate majority is up in the air.

The statistics aren’t on our side.

Since 1994, there have been six midterm elections in which one party entered Election Day holding majorities in both the House and Senate. In five of those six elections, that party lost control of the House. Democrats lost the House in 1994, 2010, and 2022; Republicans lost it in 2006 and 2018. The sole exception was 1998, when Republicans retained their House majority during Bill Clinton’s second term.

Put another way, since 1994, a party entering the midterms controlling both chambers of Congress has retained the House only once in six tries, a success rate of just 17 percent.

No amount of pandering to the American public about future $5,000.00 payments or dividends will change the forecast. Although it might cause some conservatives to sit this election out.

The hard lesson that Trump hasn’t learned in this election cycle is that he needs to turn to the right - and fast.

And fixing the economy is very different from promising to fix the economy, when mid-term elections roll around.

Likewise, giving people money is very different than promising to give people money, when mid-term elections roll around.

People want to see results two years into a presidency.

At that dinner party last Friday, after all the talk about what the Trump administration has done wrong and how they could do better. We all came to one final thought:

We all have no choice but to get our asses out there and vote for the Republican candidate in November, cause what the left has done and will do to the country when they get power again is way worse!

JGM

Audio Version:

Two Americans can suffer serious injuries after receiving federally protected vaccines and find themselves in two entirely different systems of justice. One may enter a compensation program that has paid roughly half of the petitions it has adjudicated. The other may enter a program that has compensated fewer than one in a hundred decided COVID-19 vaccine claims. Neither person chose the system. Neither person was likely told there were two very different systems for vaccine injuries before rolling up a sleeve. Congress chose what compensation they would encounter if injured, years before either was injured. But neither are they told during the informed consent process that they would not be able to sue for damages in a regular court of law.

Those percentages are not abstract. Behind every petition is a person whose life may have divided into a before and an after: before the neurological injury, before the cardiac damage, before the chronic pain, before the inability to work, before a spouse became a full-time caregiver or parents began wondering who would care for their disabled child after they died. These people do not disappear because an agency rejects their claims. Their medical bills continue. Their lost income remains lost. Their marriages and families absorb the strain. The government merely moves them from a column labeled “pending” into one labeled “denied,” and Washington congratulates itself on having provided a remedy.

This is the story of a bargain Congress made with the American people in 1986 and quietly gutted in 2005. The original bargain was not perfect, but it recognized a basic moral fact: if government protects an industry by taking away an injured person’s ordinary right to sue, government assumes a corresponding duty to provide a fair and meaningful alternative. In 2005, Congress retained sweeping immunity for big pharma, while substituting a narrower, harsher, less transparent program effectively beyond the reach of any court. The manufacturers kept the protection. The injured were left to bear the loss.

The Two Numbers

The National Vaccine Injury Compensation Program, or VICP, has received more than 29,000 petitions since 1988. It has adjudicated 25,652 and compensated 12,588, with total compensation of roughly $5.5 billion. That results in a compensation rate of near 49 percent for adjudicated petitions (HRSA 2026).

The arrival of COVID-19 did not automatically place the new vaccines within the established National Vaccine Injury Compensation Program. Instead, the federal government treated them as emergency “covered countermeasures” under the Public Readiness and Emergency Preparedness Act. In March 2020, the Secretary of Health and Human Services issued a PREP Act declaration granting broad liability protection to the manufacturers, distributors, administrators, and planners involved in the federal COVID-19 response. That declaration directed people injured by COVID-19 vaccines away from the ordinary courts and into a little-known program created for medical countermeasures used during public-health emergencies: the Countermeasures Injury Compensation Program, or CICP.

The consequences were extraordinary. CICP has received more than 14,000 COVID-19 vaccine injury claims. Yet fewer than 1 percent of decided claims had been compensated, according to federal data current as of July 1, 2026 and cited by MCTLaw, a firm representing claimants and advising on pending reform legislation (Medical Daily 2026). The government had promoted and even mandated the vaccines, while placing those injured by them into a compensation program that rejects more than 99 percent of the claims it decides.

These figures use the same basic denominator: claims that reached a decision measured against claims that were paid. Pending petitions are excluded from both. The two programs cover different products under different statutes. The numbers of claims approved versus denied establish that these systems behave radically differently. When one program compensates roughly half of the petitions it decides, and another compensates fewer than one in a hundred, the discrepancy is minor; it is an injustice of the highest order. It is the predictable result of how Congress designed the programs.

A person covered by VICP, the established compensation program for routinely recommended vaccines, has three years from the first symptom or manifestation of injury to file. An attorney may take the case knowing that reasonable fees can be paid even if the petition ultimately fails, provided it was filed in good faith and had a reasonable basis. A qualifying injury listed on the Vaccine Injury Table may carry a presumption of causation. A special master hears the case, and the decision can be reviewed by the Court of Federal Claims and then appealed.

The person sent to CICP (COVID vaccine injuries) has 1 year from the date of administration of the countermeasure. There is no comparable injury-table presumption for COVID-19 vaccine claims, no attorney-fee provision, no compensation for pain and suffering, and no independent judicial review of a denial. The claimant must produce compelling, reliable, valid medical and scientific evidence while frequently ill, financially strained, and unable to hire counsel on economically realistic terms. If CICP denies the claim, the claimant may request reconsideration by the same program. After that, the decision is final. There is no independent appeal and no court authorized to review the denial. A “no” from CICP is, for all practical purposes, the end of the line for compensation.

This is the same country and the same federal government, invoking the same public principle that “rare” vaccine injuries may be the price society pays for access to protected medical products. Yet it has created two radically different concepts of justice.

The Bargain Congress Made

Congress passed the National Childhood Vaccine Injury Act in 1986 because litigation over the DTP vaccine had driven manufacturers from the American market and destabilized supply. The problem was real. Congress confronted a collision between two legitimate concerns: preserving access to vaccines and preserving the right of people injured by a product to seek compensation. It concluded that the existing tort system could not serve both goals on the existing terms.

What Congress built was an exchange. Manufacturers received substantial protection from ordinary civil litigation. An injured person generally could not proceed directly with a civil claim exceeding $1,000 against a covered manufacturer without first filing a petition under VICP and allowing judgment to enter (CRS 2026a). In return, the injured received something of value: an administrative forum, special masters in the Court of Federal Claims, an injury table that could create a presumption of causation, attorney fees for qualifying petitions, and a dedicated funding mechanism.

Congress financed the program through an excise tax imposed on vaccine manufacturers, producers, and importers. The original tax varied by vaccine, but in 1997 Congress replaced those rates with a flat charge of 75 cents for each disease a vaccine prevents. An MMR dose therefore carries a $2.25 tax, while an influenza dose carries a 75-cent tax. Congress has not increased or indexed that amount since 1997, allowing inflation to cut its real value roughly in half. The tax acknowledged that injuries were not imaginary merely because liability had been redirected away from the manufacturers.

In 1986, Congress took away part of the ordinary legal remedy and established an alternative system outside normal court processes, and there is no question that the one who benefited most was big pharma.

The People the Bargain Failed

Even the more generous of these two programs has never made every injured family whole. A payment statistic cannot measure the years spent fighting the government, the deterioration of a patient while a claim crawls through the system, or the difference between an administrative award and what a jury might have awarded after hearing the full story. VICP caps compensation for pain and suffering and for vaccine-related death at $250,000. That figure has not kept pace with nearly four decades of inflation, much less with the actual cost of catastrophic disability. A family facing lifelong nursing care, inaccessible housing, lost earnings, and the destruction of an ordinary family life quickly learns that a statutory remedy and full compensation are not the same thing.

The autism cases are among the most painful examples. Autism was not categorically excluded from the statute, and families were permitted to file petitions. Thousands did. In the Omnibus Autism Proceeding, six test cases were selected to evaluate three general theories: that MMR combined with thimerosal caused autism, that thimerosal alone caused autism, and that MMR alone caused autism. The special masters rejected those theories, effectively closing the principal VICP pathway for more than 5,000 families alleging vaccine-induced autism.

But at least part of the scientific record used by government counsel was seriously tainted. The government relied upon Danish epidemiological studies produced through a CDC-funded research program led by Poul Thorsen, who has now pleaded guilty to stealing more than $1 million from that program. The CDC scientist who helped construct the sole-source funding opportunity was in an undisclosed romantic relationship with Thorsen, accepted gifts from him, and co-authored research produced under the award. The legally required Danish human-subjects approvals were never obtained. When CDC officials discovered the problem in 2009, after two papers had already been published and while the autism litigation was underway, they helped obtain retroactive ethical coverage rather than notifying the journals or the special masters.

So, the studies used by the government to defeat vaccine-autism claims emerged from a research program compromised by theft, undisclosed conflicts of interest, missing ethical approvals, and an extraordinary lack of institutional accountability. CDC never appears to have commissioned a serious independent reanalysis of the study designs, data, statistics, or conclusions after those facts became known. The vaccine court moved on. The government moved on. The children remained disabled, and their families were left to carry the cost for the rest of their lives. The full documentary history is examined in “Stolen Money, Missing Approvals.”

This is where sterile discussions of “claims” become morally evasive. A denied claim is still attached to a human being. Some claimants were unable to prove causation, and some may have been mistaken about the cause of an injury. Others faced the nearly impossible task of proving a complex biological event with incomplete clinical records, limited research, and no access to the internal evidence that ordinary litigation discovery might have produced. Some were too poor or didn’t know how to obtain legal counsel. When government and industry insist that only a very small number of injuries are genuinely attributable to vaccines, they should be willing to operate a system capable of identifying and generously compensating those individuals. Instead, the same institutional apparatus that describes serious injury as rare often makes proof of a rare injury prohibitively difficult.

Had these cases proceeded in ordinary civil court, manufacturers would have faced discovery, depositions, expert testimony before juries, and the possibility of damages reflecting pain, suffering, lost earning capacity, lifetime care, and punitive conduct where warranted. Not every claimant would have won. But successful catastrophic-injury cases could have produced awards many times larger than the capped or tightly calculated amounts available under federal programs. Compared with that potential exposure, the billions paid by VICP over nearly four decades amount to pennies on the dollar for an industry granted extraordinary protection from the ordinary consequences of selling a defective product.

Nor were those billions paid directly out of pharmaceutical-company profits. They came from an excise tax attached to vaccine sales, embedded in the product's price and spread across the market. And we as taxpayers get to pay for it all. The public underwrites the vaccination program, assumes the risk of injury, finances the compensation system, and then watches government lawyers contest the claims of the injured. Manufacturers have largely walked away from the courtroom exposure that every other major industry regards as a cost of doing business.

The Bargain Congress Broke

The Public Readiness and Emergency Preparedness Act of 2005, enacted within Public Law 109-148, added sections 319F-3 and 319F-4 to the Public Health Service Act. Section 319F-3 grants immunity for covered countermeasures. Section 319F-4 created CICP, which COVID vaccine injuries are adjudicated. On paper, Congress again paired immunity with compensation. In substance, it preserved the part that protects manufacturers and eliminated the part meant to protect injured people.

The 2005 PREP Act shield is broad in every direction. It covers manufacture, distribution, administration, and use, and it protects manufacturers, distributors, program planners, qualified persons who administer the products, and the United States itself. It reaches claims caused by, arising out of, relating to, or resulting from the use of a covered countermeasure. The nominal exception is willful misconduct, pursued through an exclusive federal cause of action in the United States District Court for the District of Columbia under a standard that has never been met in a reported vaccine case.

Congress designed CICP differently from the beginning. The PREP Act, enacted on December 30, 2005, limited compensation to serious physical injury or death, demanded proof of direct causation based on “compelling, reliable, valid, medical and scientific evidence,” excluded pain and suffering, and placed the process under the control of the Secretary of Health and Human Services. When HHS issued the program’s implementing regulations in 2010, it added the one-year filing deadline, detailed the internal reconsideration process, and made clear that the agency’s final eligibility decision was not subject to judicial review. CICP was not a generous program that later became restrictive. It was built that way.

Again, the deadline is one year from administration rather than VICP’s three years from the first manifestation of injury. That alone can defeat someone who spends months seeking a diagnosis, being passed from specialist to specialist, or being told that the symptoms are coincidental, psychological, or unrelated

Most physicians were discouraged or even reprimanded by employers and health-care institutions during the pandemic from documenting a possible vaccine injury in the medical record, leaving patients without the very evidence the government would later demand as proof.

There is no comparable injury table for those who are COVID-19 vaccine-injured. The claimant must produce compelling, reliable, valid medical and scientific evidence without the ability to pay attorney fees from the award. Compensation is secondary to other payment sources, lost employment income is capped, and pain and suffering are excluded altogether.

Then comes the feature that should offend people across the political spectrum: a denial cannot be reviewed by a court. An unsuccessful VICP claimant may seek review by the Court of Federal Claims and then the Federal Circuit. A CICP claimant may not appeal. Government has closed the courthouse door, designed the replacement process, set the evidentiary burden, decided the claim, and insulated its denial from independent judicial scrutiny. Calling that compensation for a vaccine injury is a farce.

In 2005, Congress kept the protection for manufacturers and abandoned the promise it had made to the injured.It preserved the half of the 1986 bargain that benefits manufacturers and reduced the half intended for injured people to an administrative gesture. The consequences remained obscure until COVID-19 transformed a little-known countermeasures program into the only legal recourse for injuries arising from an unprecedented national vaccination campaign.

And Then COVID Happened

During the pandemic, Americans were repeatedly told that vaccination was not merely a private medical choice but a civic obligation. Federal agencies promoted the products. State and local governments imposed requirements in many settings. The federal government, universities, hospitals, corporations, and the military made vaccination a condition of education, participation, or employment. People lost their jobs, careers, schooling, military status, or ability to enter public life if they refused.

When serious injuries occurred, those same people discovered that the system for justice was non-existent. The authorities that had insisted upon vaccination did not build a transparent, accessible, adequately funded system prepared to care for the injured. Manufacturers were protected under the PREP Act. Employers and administrators operated under emergency policy. The injured person was left with CICP, its one-year deadline, its formidable causation standard, no paid lawyer, no damages for pain and suffering, and no judge empowered to review a denial.

This is why the fewer-than-1-percent figure matters. A program rejecting more than 99 percent of decided claims while concealing much of its case-specific reasoning should have to demonstrate that it is separating valid claims from invalid ones rather than merely separating injured people from compensation. The government demands evidence from claimants while providing remarkably little evidence that its own adjudication is fair.

For people who lost their health after doing what government, employers, physicians, and public campaigns urged or required, the abandonment is especially bitter. Many did not oppose vaccination. They trusted the system. Some participated precisely because they believed they were protecting others. Once injured, they encountered a bureaucracy far more certain of its reasons for denial than many physicians had ever been about diagnosis or treatment. They became inconvenient witnesses to a policy success story that had no room for damaged lives.

Who Really Pays for the Liability?

A liability shield does not eliminate the cost of injury. It transfers it. A manufacturer avoids unpredictable tort exposure while the injured person absorbs medical expenses, lost income, disability, and pain the compensation system does not cover. Families provide unpaid nursing care. Private insurance, Medicare, Medicaid, disability programs, charities, churches, and local communities absorb other portions. The injury disappears from the manufacturer’s balance sheet and reappears everywhere else in society.

Litigation also does more than transfer money. Discovery can reveal what a manufacturer knew, when it knew it, how safety signals were evaluated, what was excluded from a trial, and how risks were described internally compared with how they were described to the public. Jury trials place those facts in an adversarial forum outside the agencies that approved, purchased, recommended, or promoted the product. When immunity prevents litigation, it suppresses not only damages but one of the principal mechanisms by which information about defective products reaches the public.

The excise tax in the 1986 system was intended to place some of the financial burden back on the product. Yet that tax has remained $0.75 per antigen since 1986 and has lost roughly two-thirds of its real value. Congress has increased the rate vaccine manufacturers must pay. Manufacturer protection remained extraordinarily valuable while the monies allocated to help the vaccine injured steadily shrank.

No Immunity Without a Real Remedy

Reform should begin with a principle simple enough for any member of Congress to understand: no immunity without meaningful compensation. If government eliminates the ordinary right to sue, the replacement cannot be an office that rejects more than 99 percent of decided claims, pays no attorney fees, excludes pain and suffering, and answers to no court. An administrative system may be faster and more predictable than tort litigation, but only if it is designed to deliver justice rather than simulate it. Only if the process and discovery is as transparent as the court system it is designed to replace.

Every PREP Act declaration should therefore be conditioned on a compensation system meeting minimum standards: at least three years to file, reasonable attorney fees for claims brought in good faith with a reasonable basis, a transparent presumption mechanism comparable to an injury table, compensation reflecting actual economic loss and catastrophic disability, and independent judicial review. CICP denials should be reviewable in the Court of Federal Claims under ordinary administrative-law standards. The program should also publish deidentified decisions explaining the evidence, reasoning, and standard applied to every denial. If its work is fair, transparency will vindicate it. If it is not, injured people should not have to guess how the government failed them.

VICP also requires repair. The excise tax should be adjusted and indexed to inflation, and compensation caps should reflect present-day losses rather than the value Congress assigned to death and suffering four decades ago. Department of Justice attorneys should stop treating the program as though its principal purpose were to defeat claimants. Congress promised a humane alternative to civil litigation, not years of publicly financed warfare against sick people and exhausted families.

The PREP Act contains one supposed escape from manufacturer immunity: “willful misconduct.” But this does not mean ordinary negligence, inadequate testing, failure to investigate a safety signal, or even reckless conduct. A claimant must prove an intentional act or omission taken knowingly, without legal or factual justification, and in disregard of a known or obvious risk so great that the harm was highly probable.

The claim may be filed only in the United States District Court for the District of Columbia, must satisfy special pleading requirements, and is subject to an unusually demanding evidentiary standard. The exception has never succeeded in a reported vaccine case.

Congress should either create a standard that a person with credible evidence of deliberately concealed harm can realistically meet or admit that manufacturer immunity is effectively absolute. What we have now is public relations written into law. If credible evidence shows that a manufacturer deliberately concealed material harm, an injured person should have a practical route to court.

None of these reforms requires Americans to agree about vaccine efficacy, autism, COVID policy, mandates, or the wisdom of any particular recommendation. One can believe vaccines have prevented disease and still insist that a society benefiting from them owes a profound debt to those seriously injured by them.

Indeed, the stronger one’s belief in vaccination as a public good, the stronger the obligation should be. A public-health program cannot ask individuals to accept a small risk for everyone else and then treat the unlucky few as collateral damage when that risk becomes their life or death.

The Government Cannot Take Away a Right and Call It Justice

State legislatures require certain vaccines for school attendance and sometimes for employment. Federal agencies recommend and promote them; they even buy them in bulk for free distribution for school-aged children. During emergencies, public officials may pressure nearly every institution in the country to require them.

Congress then closes the courthouse door to people injured by vaccines and substitutes a remedy of its own design. Each of these acts is an exercise of public power over a private person, and each creates an obligation on the part of the state toward the sovereign individual.

That obligation does not vanish because causation is scientifically complicated, because acknowledging an injury is politically inconvenient, or because a claimant has received a diagnosis the compensation program does not recognize as vaccine-caused. A child with lifelong disability still needs lifelong care. A previously healthy adult who can no longer work still has a mortgage, a family, and a body that no longer functions as it did. People who did what authorities asked of them should not have to beg those same authorities to acknowledge their existence and their injuries.

The government cannot have this both ways. It cannot describe vaccination as a collective duty when demanding compliance and then redefine injury as a purely private misfortune when compensation is due. It cannot tell the public that serious injuries are exceedingly rare while constructing a claims process seemingly designed to ensure that exceedingly few are recognized. It cannot shield manufacturers from discovery, jury verdicts, reputational damage, and punitive exposure while pretending that a minuscule administrative payment leaves an injured family made whole, when it isn’t.

The 1986 Congress decided that big pharma needed protecting from lawsuits. It protected vaccine supply, protected manufacturers, and created a funded remedy for those who paid the human price. The 2005 Congress's program to address vaccine damage was flawed. COVID exposed what that choice meant on a national scale: manufacturers received extraordinary protection while injured Americans encountered a program that compensated fewer than one in a hundred decided claims.

These are not anti-vaccine statistics or pro-vaccine statistics. They measure whether the United States keeps faith with its own citizens. Big Pharma received something worth incalculable sums: protection from the civil-justice system and from verdicts that might have forced companies to pay the full price of the injuries their products caused.

The victims received deadlines, caps, closed records, hostile burdens of proof, and far too often, rejection. Many were damaged for life. Some have spent years without meaningful compensation. Families caring for profoundly disabled children, including families who believed a vaccine contributed to an autism diagnosis or regression, were left to carry costs that no government ruling could make disappear. And frankly, people died from their vaccine injuries.

The bargain is not complicated. If the state takes away a citizen’s right to seek justice, the state assumes the duty to provide it. If industry wants immunity, it must bear the full and honest cost of the injuries caused by its products. Anything less is not public health. It is the privatization of profit and the profiteering of human suffering. Congress did not merely forget the injured. It built a system in which their losses could be denied, discounted, and ultimately ignored while one of the most powerful industries in America walked away protected.

This failure did not end with COVID. The liability shield now intersects with a federal strategy built around adaptable medical platforms, raising a new question: what happens when a manufacturer changes the genetic payload but retains much of the regulatory and legal architecture of the original product?

That is the subject of the next essay.

RWM/JGM

Audio Version:

Three hundred sixty-six of the 435 seats in the House of Representatives are not seriously in play this November. The Cook Political Report, which handicaps congressional races for a living, rates them solid for one party or the other. Just twenty-one are genuine toss-ups.

Think about what that means. Five out of every six House races are effectively over before the first ballot is printed. In most of America, the decisive political act was not Election Day. It was drawing the district lines.

If you live in one of those 366 districts, your representative is extraordinarily unlikely to lose in November. The real danger comes in the primary, where turnout is a fraction of what it is in the general election. That changes who matters. Your representative answers first to the relatively small number of voters who actually show up for the party primary. If you belong to the other party, or simply skip the primary, you barely enter the political calculation.

You still get to vote in November, of course. You get the sticker. You get the ritual. But in practical terms, you have very little ability to fire the person supposedly representing you.

And here is where the system gets even more interesting.

Two private political organizations largely control whose names appear on those primary ballots and which candidates receive the money, infrastructure, endorsements, consultants, voter data, and institutional support necessary to win. Between them, they have divided nearly the entire American political marketplace. Neither organization appears anywhere in the Constitution.

Economists have a word for competitors who fight furiously over market share while sharing an enormous interest in keeping everyone else out of the market. They call it a cartel.

That is the part of American elections nobody is fighting about, and it is the part that matters the most.

What the fight is about instead

The election integrity argument concerns the count. Voter identification, mail ballots, voter rolls, machine certification, audits after the fact. Those are real questions, and most of them belong to the states, where they have always belonged.

None of them touches the lines. Every congressional map in this year’s fight is constitutional. Districts come out equal in population to within a person or two, because the software that draws them is built to hit that number exactly. That is the entire federal standard. Count the people, make the numbers match. It says nothing about who drew the line, or why, or whether the winner was known a year in advance.

A clean count in a district drawn to produce a known result is a clean count of a question already settled.

Missouri’s eight days in court.

In September 2025, the Missouri legislature redrew the state’s congressional map in the middle of the decade, which is unusual, to eliminate the seat held by Democrat Emanuel Cleaver. The new map would move Missouri from six Republicans and two Democrats to seven Republicans and one Democrat. It was not a general rule about how districts should look. The voters had not changed. The census had not changed. The politicians simply changed the electorate around the seat.

More than 305,000 Missourians signed petitions to force a public vote on the redistricting. Under the Missouri constitution, enough signatures suspend a law until voters decide. Secretary of State Denny Hoskins refused to certify the petitions and ran the August primary under the new map anyway.

On September 3, the Missouri Supreme Court unanimously ruled against Hoskins. The petitions had suspended the new map back in December, meaning it had never taken effect. On September 8, the deadline for printing ballots, Justice Kavanaugh rejected the state’s emergency appeal. Minutes later, however, a federal judge in St. Louis ordered Missouri to use the new map. An appeals court left that order in place. Two days later, on September 10, the Supreme Court blocked it in an unsigned order, with no explanation and no recorded dissents.

Missourians will vote in November under the lines drawn in 2022. But the August primary was conducted under the now-discarded 2025 map. That means some voters will vote in a different congressional district in November than they did in August, and some nominees will face voters who had no opportunity to participate in the primary that selected them.

Republicans did it in Missouri. Democrats did it in California.

Last year, California voters approved Proposition 50, setting aside the congressional maps drawn by the state’s independent redistricting commission and replacing them with maps drawn by the Democratic-controlled legislature for elections through 2030.

Democrats defended the move as a response to Republican redistricting in Texas. Republicans defended Missouri’s redraw as part of the same national fight.

The pattern is difficult to miss. Neither party consistently opposes partisan redistricting. Each condemns it when the other party draws the lines and defends it when the advantage is its own.

Congress had the constitutional authority to intervene. It did not.

Instead, Missouri spent eight days bouncing between four courts while voters did not know which congressional districts would ultimately govern the election. The institution the Constitution expressly empowered to make or alter the rules never entered the fight.

Congress already has the power.

The Constitution gives Congress the power to regulate congressional elections. Article I, Section 4 says the states set the times, places, and manner of those elections, but the very same sentence says Congress may “make or alter” those rules.

This is not a loophole or a modern interpretation. It is written into the Constitution itself, and the founders debated the provision publicly. Hamilton defended it in the Federalist Papers because leaving Congress entirely dependent upon rules made by the states could ultimately leave the federal government at their mercy.

Political parties, by contrast, have no constitutional role at all. The Constitution does not even mention them. The men who wrote it were deeply concerned about factions and the emergence of permanent political parties. Madison devoted Federalist 10 to the problem of faction. Washington later used his Farewell Address to warn against the “spirit of party.” The constitutional structure tied representation to states and congressional districts, not to two national party organizations dividing the country between them.

And Congress used the power the Constitution gave it. Beginning in 1842, Congress imposed federal requirements on how House members were elected. In 1872, it required congressional districts to contain roughly equal populations. In 1901, it required them to be compact, a restraint on the sprawling and contorted districts we now associate with gerrymandering. The 1911 law carried the single-member, contiguity, population, and compactness requirements forward.

Then Congress let those rules disappear. The Permanent Apportionment Act of 1929 made the allocation of House seats after each census automatic, ending the need for Congress to pass a new apportionment law every decade. But it included none of the previous federal standards governing how congressional districts were drawn. The 1941 law restored none of them. Congress finally restored the single-member-district requirement in 1967, amid concern about states using at-large elections.

The federal compactness requirement disappeared nearly a century ago. Congress has never restored it.

So yes, states normally administer elections. But congressional elections are constitutionally different. The Constitution gives states the first move and Congress the final authority to “make or alter” their rules. Congress once used that authority to set national standards for congressional districts.

It could do so again.

The Court tried to fix the problem and then quit halfway.

The federal courts eventually imposed the rule Congress had abandoned: equal population. By the early 1960s, some legislative districts contained vastly more people than others, while the legislators who benefited from those disparities had little incentive to change them.

In 1962, the Supreme Court opened the federal courthouse door to reapportionment challenges. Two years later, it required congressional districts within a state to contain roughly equal populations. The principle became known as “one person, one vote.”

But equal population governs the numbers, not the shape of the district. Every map in this year’s fight can satisfy the population rule while still being drawn for partisan advantage.

For decades, litigants tried to persuade federal courts to police partisan gerrymandering as well. In 2019, the Supreme Court said no. In Rucho v. Common Cause, Chief Justice Roberts wrote for the majority that federal courts had no neutral, manageable constitutional standard for deciding how much partisan advantage was too much. Partisan-gerrymandering claims, the Court held, were beyond the reach of federal courts.

But Roberts did not say there was no remedy. He pointed directly back to Congress. The Framers had given Congress the power to address partisan gerrymandering through the Elections Clause, he wrote, and Congress had used that power before. “That avenue for reform,” the Court concluded, “remains open.”

Seven years later, Congress has not even attempted reform.

Why Congress prefers the system it has

The people with the power to fix congressional redistricting are often the people who benefit from leaving it alone.

Call your representative’s office and ask Congress to take up redistricting. You may be told that redistricting is a matter for the courts. But in 2019, the Supreme Court said federal courts would not police partisan gerrymandering. The responsibility circles back to Congress, and the circle closes.

Economist James M. Buchanan won the Nobel Prize in 1986 for developing public-choice theory, which applied economic analysis to political decision-making. His central insight was simple: politicians respond to incentives just as everyone else does. The people who write the rules have interests of their own.

That is the problem here. Neutral rules are easiest to adopt before anyone knows who will benefit from them.

Eighty-four percent of House seats are rated safe. A member sitting in one of them has little political incentive to make that seat more competitive. And the system protecting that member is the very system Congress would have to regulate.

The same incentive exists at the party level. The two parties would be writing rules that could take away districts they currently draw to their own advantage.

Congress has done this before.

The Authorization for Use of Military Force passed on September 14, 2001, three days after the attacks. It turns twenty-five on Monday. Five administrations have stretched it to cover groups that did not exist when it was written. Repealing it would take a majority and an afternoon.

The National Emergencies Act tells a similar story. In 1976, Congress gave itself the power to terminate a presidentially declared emergency by concurrent resolution, without a presidential signature. After the Supreme Court invalidated that kind of legislative veto in 1983, Congress amended the law to require a joint resolution, which the president can veto. Ending an emergency over a president’s objection therefore requires a two-thirds vote in both chambers.

The pattern is the same. Congress has constitutional and statutory powers that it can exercise, amend, reclaim, or restructure. When it declines to act, other institutions make the consequential decisions instead.

Congress holds a power, declines to use it, and somebody else fills the power vacuum. Nobody seized these powers. They were given away by Congress.

What Congress could do tomorrow

Congress does not need a constitutional amendment, a new agency, or new authority. It could write the rules now.

It could require congressional maps to respect county and municipal boundaries. Those lines were not designed for partisan advantage. They accumulated over generations around rivers, market towns, and courthouses, and often reflect communities that already conduct their affairs together. Carl Menger called institutions like these organic: they carry knowledge nobody planned and nobody can reproduce on demand. A county line tells us something about a community. A line drawn to eliminate one congressman tells us who drew it.

Congress could also repeal the 1967 requirement that every congressional district elect only one House member. That is what a “single-member district” means: one district, one seat, one winner. Gerrymandering depends heavily on that structure because mapmakers can manipulate who wins each seat by packing opponents into some districts and spreading them thinly across others. That becomes much harder when a larger district elects several representatives instead of just one.

Most important, Congress could make any new rules take effect years from now. A law governing maps drawn after the 2030 census would bind a future Congress whose partisan composition and political geography cannot be known with certainty today. A delayed start recreates some of the uncertainty that makes neutral rules easier to accept. It also addresses the immediate objection from either party: neither side has to change the rules solely for the other's benefit.

Why Congress has not fixed it.

Congress can change these rules. The political incentives run in the opposite direction.

The people who would have to change the system include those who benefit from it. Buchanan’s objection was fundamental: ordinary legislation is a weak restraint when the people being restrained retain the power to rewrite it. Durable rules are easiest to establish when nobody knows in advance who will benefit.

That is the problem with redistricting. Of the 435 House races this November, 366 are currently rated Solid for one party or the other. In most districts, the decisive political competition occurs before the general election. A perfectly accurate count does not change the boundaries within which those votes are cast. Voter ID does not change them. An audit does not change them. The argument over election integrity largely begins after the political geography has already been decided.

Missouri and California are not opposing principles. Republican officials pursued a mid-decade redraw in Missouri; Democratic officials pursued one in California. Each case used different procedures and arose in different circumstances, but both demonstrate the same structural fact: political parties have powerful incentives to seek favorable district lines when the law allows them to do so.

That brings us back to where we started. The parties compete fiercely over who controls the lines. The harder question is who writes the rules governing the lines themselves. The Constitution gives Congress that power. For nearly a century, Congress has largely declined to use it.

Free subscribers get every post. Paid subscribers pay for the work behind them.

References

Apportionment Act of 1842, 5 Stat. 491.

Apportionment Act of 1872, 17 Stat. 28.

Apportionment Act of 1901, 31 Stat. 733.

Apportionment Act of 1911, 37 Stat. 13.

Authorization for Use of Military Force, Pub. L. 107-40 (2001).

Baker v. Carr, 369 U.S. 186 (1962).

Brennan Center for Justice. 2026. “Declared National Emergencies Under the National Emergencies Act.” Last updated September 2.

Buchanan, James M. 1979. “Politics Without Romance.” Institute for Advanced Studies, Vienna.

Buchanan, James M., and Roger D. Congleton. 1998. Politics by Principle, Not Interest. Cambridge University Press.

Buchanan, James M., and Gordon Tullock. 1962. The Calculus of Consent. University of Michigan Press.

Colegrove v. Green, 328 U.S. 549 (1946).

Cook Political Report. 2026. “2026 House Race Ratings.” Last updated August 25.

Hamilton, Alexander. 1788. Federalist No. 59.

INS v. Chadha, 462 U.S. 919 (1983).

Madison, James. 1787. Federalist No. 10.

Menger, Carl. 1883. Investigations into the Method of the Social Sciences.

National Emergencies Act, 50 U.S.C. §§ 1601 et seq.

People Not Politicians and Richard von Glahn v. Denny Hoskins, Supreme Court of Missouri, September 3, 2026.

Permanent Apportionment Act of 1929, 46 Stat. 21.

Reynolds v. Sims, 377 U.S. 533 (1964).

Rucho v. Common Cause, 588 U.S. 684 (2019).

Uniform Congressional District Act, 2 U.S.C. § 2c (1967).

Washington, George. 1796. Farewell Address.

Wood v. Broom, 287 U.S. 1 (1932).

My liberal facebook friends are all into puppies (and horses), but children… not so much.

I think that speaks volumes…

The American experiment was never supposed to be an experiment in how much responsibility could be transferred from citizens to government. It rested on almost the opposite proposition: that a free people could govern themselves, care for their communities, solve problems locally, and voluntarily cooperate without waiting for direction from a distant bureaucracy. Alexis de Tocqueville recognized this as one of the most remarkable characteristics of America when he traveled the young nation in the 1830s. Americans, he observed, were extraordinary association-builders. When they wanted a church, school, hospital, road, charitable organization, civic improvement, or political reform, they formed a group and did something about it.

This dense network of families, churches, voluntary associations, businesses, towns, and local institutions created a civil society strong enough that government did not have to occupy every corner of life. The American experiment was always about self-rule, otherwise known as self-sovereignty.

Volunteerism builds citizenship because people retain both responsibility and agency. Big government progressively replaces both. The American experiment depends not merely upon elections and constitutional language, but upon preserving a people sufficiently independent, prosperous, cohesive, and self-reliant to govern themselves. Once citizens become clients of the state and communities surrender their functions to centralized bureaucracies, we may retain the outward machinery of a republic while losing much of what made the republic possible in the first place.

But a culture must first understand the principles of volunteerism and the importance of governing itself. Self-government is not simply a constitutional arrangement. It is a learned behavior, passed from one generation to the next.

Ronald Reagan may have articulated the fragility of that inheritance best when he warned, “Freedom is never more than one generation away from extinction.” Liberty does not perpetuate itself. A nation is only as capable of self-government as its people are willing and able to assume the responsibilities that self-government requires. Citizens must understand that freedom comes with obligations: to family, neighbors, community, local institutions, and ultimately the nation itself. When those habits disappear, government (and in particular, the courts) inevitably moves into the vacuum.

This is also why immigration and immigration policy cannot be separated from culture. America has historically absorbed enormous numbers of immigrants, but successful immigration required something more than crossing a border and finding employment. It required assimilation into the American experiment.

Orderly immigration gives a nation the ability to decide how many newcomers it can successfully absorb and provides newcomers the time and institutional structure necessary to learn the language, laws, history, customs, and civic expectations of their new country. That means they must want to become Americans and participate in the American experiment.

Flooding a nation with newcomers faster than assimilation can reasonably occur is something fundamentally different. Many immigrants may be admirable, industrious people seeking better lives, but they cannot reasonably be expected to arrive already understanding Tocqueville’s America: voluntary associations, limited government, local responsibility, constitutional restraint, private property, individual liberty, and the expectation that citizens govern themselves rather than look automatically to the state. Those ideas are culturally transmitted.

If immigration occurs faster than the receiving society can transmit them, America risks importing populations without successfully creating Americans in the civic sense. The issue, therefore, is that the immigration wave became so large and rapid over the past fifty years that the American experiment itself has been diluted before newcomers have even had the opportunity to understand it.

In the terrible days after September 11, 2001, Americans were reminded of something we had begun to forget: what it meant to be American. For a brief time, the labels that divide us mattered less than the country that united us. Flags appeared on porches, strangers helped strangers, churches filled, and firefighters, police officers, soldiers, and ordinary citizens became our heroes.

We remembered that America is more than a government or a piece of land. It is a shared inheritance of liberty, courage, sacrifice, and responsibility to one another. Perhaps the lesson of 9/11 worth remembering today is that we should not need another tragedy to remember who we are.

“And I proud to be an American…" God bless the USA”

Let’s see who figures it out first!

JGM

Audio Version:

When we talk about respiratory infections, we tend to talk about where the disease eventually causes trouble. Influenza can cause pneumonia. RSV can infect the lower respiratory tract. COVID-19 can involve the lungs. Bacterial infections can cause pneumonia and serious systemic illness. But this way of thinking skips one of the most important parts of the story. For many respiratory pathogens, the lungs are not where the infection begins. The first encounter happens considerably farther north, in the nose, nasal passages, throat, and connected sinuses.

The nose is the primary portal of entry into the respiratory tract, but it is also one of its most important gatekeepers. Every day we inhale thousands upon thousands of particles: dust, pollen, fungal spores, bacteria, viruses and all sorts of microscopic debris. Yet relatively few of these succeed in causing disease. That is because the nose is not simply an open pipe leading to the lungs. It is a remarkably sophisticated biological barrier designed to trap, disable and remove potential invaders before they can establish themselves.

Imagine standing in a room with someone who has influenza. They cough, laugh or simply breathe, releasing virus-containing droplets and tiny aerosols into the air. You inhale some of them. The virus does not magically appear in your lungs or bloodstream. Much of what you inhale first encounters the warm, moist surfaces lining your nose and upper airway. Immediately, a contest begins. The outcome of that contest may determine whether you ever know that encounter occurred.

The first obstacle is surprisingly simple: mucus. We tend to notice mucus only when we have too much of it, but a thin layer covers the nasal passages all the time. It is sticky for a reason. Viruses, bacteria, and particles entering the nose can become trapped in it rather than continuing deeper into the respiratory tract. Underneath this mucus are specialized cells equipped with microscopic hair-like structures called cilia. The cilia beat in coordinated waves, continuously moving mucus and the material trapped within it toward the throat, where it can be swallowed and ultimately destroyed or eliminated. This is called mucociliary clearance, but I prefer to think of it as the respiratory tract’s conveyor belt.

The system is considerably more sophisticated than a sticky filter and a conveyor belt. The cells lining the nose are themselves active participants in immunity. They can recognize molecular signs that a virus or bacterium has arrived and respond by producing interferons and other chemical signals. These signals warn neighboring cells, activate antiviral defenses, and recruit immune cells to the site. Antimicrobial substances are present in nasal secretions, and antibodies, particularly secretory IgA, patrol the mucosal surface where respiratory pathogens first arrive.

This is mucosal immunity, and it is critically important. During COVID-19, enormous attention was focused on antibodies circulating in the blood. Those antibodies matter, particularly in preventing severe systemic disease, but they are not the whole immune system. A respiratory virus generally does not begin by jumping directly into the bloodstream. It lands on a mucosal surface. If we want to understand whether infection becomes established, we have to pay attention to the immunity waiting at the actual front door.

The sinuses are part of this same system. They are air-filled spaces within the bones of the face and skull, lined with respiratory mucosa and connected to the nasal passages through relatively small openings. They produce mucus, have ciliated cells, and participate in the same system of filtration, clearance, and immune surveillance. Under normal circumstances, mucus continuously moves through these spaces and drains into the nasal cavity. The nose and sinuses therefore should not really be thought of as separate organs. They form an interconnected mucosal environment guarding the entrance to the respiratory tract.

But every gatekeeper has vulnerabilities. Suppose a virus succeeds in reaching a susceptible nasal cell and entering it. That one infected cell can become a miniature virus factory, producing new viral particles that infect neighboring cells. The infection begins spreading across the nasal mucosa. Inflammation follows. Blood vessels dilate, tissues swell, and mucus production increases. Suddenly that elegant drainage system does not work quite as efficiently. The small openings connecting the sinuses to the nose can narrow or become obstructed, mucus becomes trapped, and the environment changes.

This is one reason a seemingly ordinary viral cold can sometimes turn into something more complicated. The original infection may be viral, but inflammation and impaired drainage can create conditions in which bacteria that normally inhabit the nose and throat gain an advantage. The viral infection has damaged the gate and disrupted the guards. A secondary bacterial infection can then follow. The same general principle applies farther down the respiratory tract. Influenza, for example, can damage respiratory defenses and make subsequent bacterial pneumonia considerably more likely.

Temperature and humidity also affect these defenses. The nasal passages are cooler than the body's core, which can favor replication of some respiratory viruses while also weakening local antiviral responses. Dry air can further impair the mucus and cilia that normally trap and sweep pathogens away. This may help explain why respiratory infections flourish during winter, when cold outdoor air and indoor heating create particularly dry conditions. Grandma may not have understood interferons or mucociliary clearance, but telling you to keep warm and avoid getting dried out wasn't such bad advice after all.

What is remarkable is how often this system succeeds. We constantly encounter microorganisms that never cause recognizable illness. A virus may become trapped in mucus and swept away. An antibody may bind it before it enters a cell. An infected cell may trigger a rapid interferon response that prevents the infection from spreading. We never count these events because there is nothing to count. Nobody goes to the doctor to report the influenza infection they almost had.

This changes how we should think about respiratory disease. There is an enormous difference between exposure, infection, and disease. Exposure means the pathogen reached you. Infection means it successfully entered cells and began replicating. Disease means that process progressed far enough to produce symptoms or tissue damage. The nose and sinuses sit at the critical boundary between those stages. They are where exposure can either be stopped or allowed to become established infection.

Modern medicine understandably becomes most interested when things go badly. We study pneumonia, respiratory failure, systemic inflammation, and hospitalization because these are the events that kill people. But by concentrating on the end stages of disease, we can lose sight of the biology that determines whether many respiratory infections ever get that far. Before there is pneumonia, there is often a battle in the upper airway. Before a virus reaches the lungs, it frequently must establish itself in the nose or throat. Before the adaptive immune system mobilizes its full response, mucus, cilia, epithelial cells, interferons, antimicrobial molecules and local antibodies are already at work.

The nose and sinuses therefore occupy a rather extraordinary position. They are simultaneously the respiratory system’s most important portal of entry and one of its most important defensive barriers. The same doorway that gives viruses and bacteria access to us is guarded by an elaborate system evolved to keep them out. For most of our lives, that gatekeeping system quietly does its job and we never give it a thought. We tend to notice the nose only when it fails.

Perhaps we should pay considerably more attention to what keeps it working.

Can We Wash the Invaders Away?

Once we understand the nose and sinuses as both portal and gatekeeper, an obvious question follows: can we help that defensive system do its job?

One of the oldest and simplest approaches is nasal lavage. The idea is hardly sophisticated: put saline into the nasal passages and physically wash away mucus, particles, allergens, and at least some of the microorganisms sitting on the mucosal surface. At the same time, saline can help maintain the physical properties of mucus and support the cilia that continuously move material out of the respiratory tract.

For years, nasal irrigation was generally relegated to the category of “probably makes you feel better.” But the evidence has become more interesting. Recent randomized trials suggest that saline nasal sprays or irrigation can modestly shorten respiratory illnesses, reduce nasal symptoms and, in some studies, accelerate the decline in viral load. One very large randomized trial involving almost 14,000 people found that those assigned to use saline nasal spray at the first sign of a respiratory infection or after likely exposure experienced substantially fewer days of illness over the following six months.

This does not mean that you can wash your nose and magically prevent influenza, COVID or RSV. The studies are heterogeneous, not all trials have been positive, and once a virus has entered large numbers of cells, simply washing the surface cannot remove an intracellular infection. But timing may matter enormously. There is a considerable biological difference between trying to reduce the number of viral particles sitting on the nasal surface shortly after exposure and attempting the same thing three days later, after the virus has already replicated millions of times.

Think of it as washing mud off the porch before everyone has tracked it through the house.

There is also growing interest in adding antimicrobial substances to saline. One of the more intriguing is performing nasal washing using a very dilute mixture of saline and hypochlorous acid, or HOCl. Despite the chemical-sounding name, HOCl is not foreign to human biology. Our own neutrophils manufacture it as part of the innate immune response to microorganisms. It is one of the chemical weapons our immune system itself uses against invading organisms. At appropriately low concentrations, HOCl has broad antimicrobial activity against bacteria, fungi, and viruses.

Laboratory experiments using human nasal epithelial cells have found that low concentrations of HOCl can kill or inactivate multiple microorganisms, including influenza A virus, without detectable toxicity to the nasal cells under the conditions tested.

HOCl Has Moved Well Beyond the Laboratory

Hypochlorous acid is not some obscure experimental compound supported by a few laboratory studies. It has accumulated a substantial scientific and clinical literature involving antimicrobial activity, antisepsis, wound care, mucosal use and direct application to human tissues. In 2025, the World Health Organization added an explicit listing for hypochlorous acid solution as an environmental disinfectant to its Model Lists of Essential Medicines for adults and children, importantly distinguishing HOCl from hypochlorite, or bleach.

The clinical literature is substantial. The WHO submission itself summarized 25 publications, including 13 randomized controlled trials, involving applications ranging from infected wounds and diabetic ulcers to surgical infections and direct application to human tissues. The larger literature cited in that submission includes thousands of publications addressing HOCl antimicrobial activity, safety and wound healing.

There is also a US regulatory history that brings this much closer to home. Curativa Bay has an FDA-listed medical device for wound and burn care, intended for both over-the-counter and prescription use. Its indications include minor cuts, lacerations, abrasions and minor burns, including sunburn, and the federal device record classifies it as an “antimicrobial wound irrigation solution.”

In other words, we are not talking about a chemical whose established use is limited to disinfecting countertops. Properly formulated HOCl products are already regulated for application directly to injured human tissue.

But perhaps most relevant to our discussion, HOCl has also been put directly into human noses and studied in randomized clinical trials. A randomized placebo-controlled study of adults with chronic rhinosinusitis that had failed conventional medical therapy compared low-concentration HOCl nasal irrigation with saline for eight weeks. HOCl produced significantly greater improvement in sinus symptoms than saline at several early time points.

Researchers have also studied children with chronic sinusitis. In a prospective randomized trial, both saline and low-concentration HOCl irrigation improved symptoms, but improvement on sinus X-rays was significantly greater with HOCl. The authors concluded that HOCl nasal irrigation was an effective adjunctive treatment for pediatric chronic sinusitis. A separate multicenter, randomized, double-blind trial enrolled 139 patients with perennial allergic rhinitis. Both saline and HOCl irrigation significantly improved symptoms, although HOCl did not outperform saline. Just as importantly for the question of putting HOCl into the nose, investigators reported no clinical side effects attributable to the HOCl irrigation. HOCl nasal spray has also been investigated as an adjunct after functional endoscopic sinus surgery in patients with chronic rhinosinusitis.

So the question is no longer whether dilute HOCl can be used in the human nose. It has been. Nor are we relying solely upon laboratory experiments to argue that nasal exposure is tolerable. We have controlled human studies involving direct nasal irrigation, including adults, children, and patients with chronic inflammatory sinonasal disease.

What we do not yet have is something much more specific: a large randomized trial demonstrating that routine use of dilute HOCl nasal cleansing prevents influenza, COVID-19, RSV or the common cold better than saline alone.

But the data we have also changes the question. We know that the nose is a major portal of entry for respiratory pathogens. We know that saline nasal cleansing can improve mucociliary clearance and has clinical evidence in respiratory illness. We know that HOCl is produced naturally by our own neutrophils as part of innate immunity. We know that properly formulated HOCl has broad antimicrobial activity against bacteria, fungi and viruses. And we now know from randomized human studies that dilute HOCl can be used directly in the nasal passages and has shown clinical benefit in some chronic sinonasal conditions.

So we are no longer starting at square one.

The unanswered question is whether adding HOCl to routine nasal cleansing can help the gatekeepers do something even more important: stop or reduce an acute respiratory infection near the place where it begins.

That study should be relatively easy and inexpensive to do. Which raises an interesting question: given the enormous amounts of money spent developing drugs and vaccines against respiratory viruses, why has comparatively little serious research been devoted to inexpensive interventions aimed directly at the portal where many of these infections first become established?

Why did the government not fund or develop such studies during COVID?

Of course, HOCL is made by our own neutrophils. It is not a drug. It can not be patented. Big pharma has no big interest in the product. So, for some reason, the government has also not been interested in funding such a study. Coincidence? But given what we now understand about the nasal mucosa's role in establishing respiratory infection, I think conducting such a study should be a research priority.

Give the Gatekeepers a Hand

This brings me to something that has become part of our own routine. We brush our teeth because we understand that the mouth is constantly exposed to bacteria. We wash our hands because we understand that our hands pick up microorganisms from the environment. Yet we inhale thousands of liters of air every day, carrying dust, pollen, bacteria, viruses and countless other particles across the mucosal surfaces of our noses, and most of us never give those surfaces another thought.

Given what we now know about nasal cleansing, mucociliary clearance and HOCl, the next question seems rather obvious: why not combine the mechanical benefits of saline nasal cleansing with a very dilute HOCl solution specifically formulated for the nose?

That is the idea behind the Gentle Nose Cleanser made by Curativa Bay, which we use and keep around ourselves. The nasal formulation contains saline and HOCl at approximately 40 ppm, substantially lower than the company’s 200 ppm HOCl skin formulation. It is designed to cleanse and hydrate the nasal passages while adding a low concentration of an antimicrobial molecule that our own immune system produces.

Saline nasal cleansing has clinical evidence behind it. HOCl has an extensive antimicrobial and human safety literature. HOCl nasal irrigation itself has been tested in randomized human trials, including patients with chronic sinusitis and allergic rhinitis. And laboratory studies demonstrate antimicrobial activity against respiratory viruses.

Put those pieces together, and I think HOCL deserves considerably more attention than it receives.

This is also where I part company with the idea that every useful intervention in medicine must begin with an expensive pharmaceutical product. Sometimes biology points us toward much simpler interventions. Supporting the physical and immunological barrier at the place where infection often begins is not particularly glamorous. There is no billion-dollar drug involved. Nobody needs an infusion center. You are simply helping the gatekeeper at the gate.

When we travel, spend time in crowded airports and airplanes, attend large events, or know that respiratory infections are circulating, nasal cleansing has become one of the simple measures we use. I cannot tell you that doing so guarantees that you will not become infected. It doesn’t. Nor should anyone selling a nasal cleanser make that promise.

What I can say is that the scientific case is considerably more developed than “it seems like a good idea.” The nose is a major portal of respiratory infection. Nasal cleansing has a clinical literature behind it. Dilute HOCl has actually been studied in the human nose, including in randomized trials. And HOCl itself has broad antimicrobial activity and an extensive history of human medical use.

That is enough evidence for me to use it routinely, and I do.

The nose is the front door to the respiratory tract. We spend enormous amounts of money dealing with what happens after an invader gets through that door.

Perhaps we should spend a little more effort taking care of the door.

For those interested, the product we use is Curativa Bay’s Gentle Nose Cleanser.

By: RWM

Full disclosure: I am the Chief Medical Officer at Curativa Bay. No, they did not pay me to write this article. I wrote it because I believe in the product. Sometimes a conflict of interest is simply an interest.

References

1. Little P, Vennik J, Rumsby K, et al. Nasal sprays and behavioural interventions compared with usual care for acute respiratory illness in primary care: a randomised, controlled, open-label, parallel-group trial. Lancet Respir Med. 2024;12(8):619-632. doi:10.1016/S2213-2600(24)00140-1. This large trial provides particularly useful evidence for the saline-nasal-spray portion of the article.

2. Ramalingam S, Graham C, Dove J, Morrice L, Sheikh A. A pilot, open labelled, randomised controlled trial of hypertonic saline nasal irrigation and gargling for the common cold (ELVIS). Sci Rep. 2019;9:1015. doi:10.1038/s41598-018-37703-3.

3. Kim HJ, Lee JG, Kang JW, Cho HJ, Kim HS, Byeon HK, Yoon JH. Effects of a low concentration hypochlorous acid nasal irrigation solution on bacteria, fungi, and virus. Laryngoscope. 2008;118(10):1862-1867. doi:10.1097/MLG.0b013e31817f4d34. PMID: 18677274. This in-vitro study used primary human nasal epithelial cells and found no detectable cytotoxicity after exposure to 3.5 ppm HOCl under the experimental conditions. HOCl demonstrated greater than 99% bactericidal or fungicidal activity against most organisms tested and produced a 3.2-log10 reduction in influenza A virus.

4. Cho HJ, Min HJ, Chung HJ, et al. Improved outcomes after low-concentration hypochlorous acid nasal irrigation in pediatric chronic sinusitis. Laryngoscope. 2016;126(4):791-795. doi:10.1002/lary.25605. This was a prospective randomized trial; both saline and HOCl improved symptoms, while radiographic improvement was greater with HOCl.

5. Yu MS, Kim BH, Kang SH, Lim DJ. Low-concentration hypochlorous acid nasal irrigation for chronic sinonasal symptoms: a prospective randomized placebo-controlled study. Eur Arch Otorhinolaryngol. 2017;274(3):1527-1533. doi:10.1007/s00405-016-4387-5. PMID: 27853946. Forty-three adults with chronic rhinosinusitis refractory to medical therapy were randomized to HOCl or saline irrigation for eight weeks. The HOCl group showed greater improvement in CRS symptoms than saline, although endoscopic scores and bacterial culture results did not differ significantly.

6. Kim HC, Kim DK, Kim JS, et al. Hypochlorous Acid Versus Saline Nasal Irrigation in Allergic Rhinitis: A Multicenter, Randomized, Double-Blind, Placebo-controlled Study. Am J Rhinol Allergy. 2022;36(1):129-134. doi:10.1177/19458924211029428. Both groups improved; HOCl was well tolerated but did not significantly outperform saline.

7. Gutiérrez-García R, De La Cerda-Ángeles JC, Cabrera-Licona A, et al. Nasopharyngeal and oropharyngeal rinses with neutral electrolyzed water prevents COVID-19 in front-line health professionals: A randomized, open-label, controlled trial in a general hospital in Mexico City. Biomed Rep. 2022;16(2):11. doi:10.3892/br.2021.1494. The trial randomized 170 frontline health-care workers; COVID-19 occurred in 1.2% of the rinse group versus 12.7% of controls. The study was open-label and several authors were employees of the product manufacturer, so those limitations should accompany the striking result.

8. Wang L, Bassiri M, Najafi R, et al. Hypochlorous acid as a potential wound care agent: Part I. Stabilized hypochlorous acid: a component of the inorganic armamentarium of innate immunity. J Burns Wounds. 2007;6:e5. This is a useful foundational paper for explaining that HOCl is part of our own innate antimicrobial defenses rather than simply an industrial disinfectant.

9. Weiss SJ, Klein R, Slivka A, Wei M. Chlorination of taurine by human neutrophils: evidence for hypochlorous acid generation. J Clin Invest. 1982;70(3):598-607. doi:10.1172/JCI110652. This is one of the classic mechanistic papers establishing production of HOCl by human neutrophils.

10. Guan H, Nuth M, Weiss SR, et al. HOCl rapidly kills Corona, Flu, and Herpes to prevent aerosol spread. J Dent Res. 2023;102(9):1031-1037. doi:10.1177/00220345231169434.

11. Dianty R, Hirano J, Anzai I, et al. Electrolyzed hypochlorous acid water exhibits potent disinfectant activity against various viruses through irreversible protein aggregation. Front Microbiol. 2023;14:1284274. doi:10.3389/fmicb.2023.1284274.

12. Block MS, Rowan BG. Hypochlorous acid: a review. J Oral Maxillofac Surg. 2020;78(9):1461-1466. doi:10.1016/j.joms.2020.06.029.

13. Del Rosso JQ, Bhatia N. Status report on topical hypochlorous acid: clinical relevance of specific formulations, potential modes of action, and study outcomes. J Clin Aesthet Dermatol. 2018;11(11):36-39.

14. Burian EA, Sabah L, Kirketerp-Møller K, Gundersen G, Agren MS. Effect of stabilized hypochlorous acid on re-epithelialization and bacterial bioburden in acute wounds: a randomized controlled trial in healthy volunteers. Acta Derm Venereol. 2022;102:adv00727. doi:10.2340/actadv.v102.1624.

15. Fazli MM, Kirketerp-Møller K, Sonne DP, et al. A first-in-human randomized clinical study investigating the safety and tolerability of stabilized hypochlorous acid in patients with chronic leg ulcers. Adv Wound Care. 2024. doi:10.1089/wound.2024.0040. The broader WHO submission contains numerous additional randomized and prospective human studies of HOCl in wound and tissue applications.

16. World Health Organization. Hypochlorous acid: antisepsis, disinfection, and wound care. Application to the 2025 WHO Model List of Essential Medicines and Model List of Essential Medicines for Children. The application assembled the extensive HOCl antimicrobial, safety, wound-care and clinical literature discussed in this article.

17. World Health Organization Expert Committee on Selection and Use of Essential Medicines. The Selection and Use of Essential Medicines, 2025. WHO added aqueous hypochlorous acid at ≥150 ppm to the core Model Lists of Essential Medicines for adults and children as an environmental disinfectant, explicitly distinguishing HOCl from hypochlorite products.

Audio Version:

OK, this is a bit of a rant today. So bear with me.

Massachusetts has written into law something that most people probably STILL do not understand: a person can kill someone, successfully claim insanity, and never serve a prison sentence for the killing. Ever.

In Massachusetts, if the jury has reasonable doubt about whether the defendant was criminally responsible due to insanity, the verdict is not guilty by reason of lack of criminal responsibility.

The defendant can then be sent to a psychiatric hospital for an initial 40-day period of observation. After that, the state can seek to keep the person committed, initially for six months and then in renewable one-year periods. SO TECHNICALLY LINDSAY CLANCY COULD BE RELEASED 40 DAYS AFTER SENTENCING.

This is psychiatric confinement, not a prison sentence. Once the person no longer meets the requirements for involuntary commitment, they can be released. There is no murder sentence sitting there waiting to begin. There is no requirement that they spend 20 years, 30 years, or life behind bars because they took another person's life.

Think about what Massachusetts has created. Someone can commit a horrific murder, avoid a guilty verdict because there was reasonable doubt about their sanity at the time of the crime, spend some period of time in a psychiatric hospital, and walk out the door in 40 days if they are later judged well enough to be released. The victims remain dead. The killer can go home and murder again.
That is not a minor technicality in Massachusetts law.

There is no mandatory 20-year sentence. There is no mandatory ten-year sentence. There isn’t even a mandatory five-year sentence. Once the state can no longer prove that the killer remains mentally ill and dangerous enough to justify involuntary commitment, the legal basis for holding that person disappears. The murder does not then send them to prison, because legally they were acquitted of it.

Even more remarkable, it is extraordinarily difficult to determine what happens to these people after release. The Boston Globe found that Massachusetts privacy laws and fragmented records make systematic follow-up nearly impossible.

So the Commonwealth has constructed a system capable of releasing people who have killed after only a few years, while making it remarkably difficult for the public to determine how often they subsequently commit crimes. That is not transparency. And it is certainly not reassuring.

So when you listen to the juror implying that the holdout juror must have had ulterior motives, such as simply wanting to see Clancy go to prison, she is putting words in his mouth. She can’t get inside his head.

Being evil, even being possessed by evil, does not make someone insane. It makes them a murderer. There was substantial evidence that Clancy’s crimes were premeditated and not brought on by drugs. The evidence showed that she went on and off medications and was not taking anything particularly likely to induce psychosis at the time.

But even if she was in a drug-induced psychosis, that does not mean she is insane or shouldn’t be held responsible.

A man who gets drunk or takes drugs and then rapes a woman does not get to claim that the alcohol or drugs made him do it. Intoxication may impair judgment, lower inhibitions, or make someone more violent, but it does not magically erase responsibility for the crime, nor does it prove insanity. Being under the influence is not the same thing as being legally insane.

Massachusetts did not arrive at this absurd system overnight. It is the product of more than a century of law, followed by a series of changes beginning in the 1960s that made an already unusual system considerably more consequential.

Read more

Audio Version:

For Preventing the Voters of Eastern Germany from Being a Burden to Their Betters, and for Making Them Beneficial to the Union

It is a melancholy object to those who read the better newspapers of London and Brussels, to see the opinion pages crowded with persons of the first quality, all of them weeping, and all of them weeping about Magdeburg.

On the sixth of September, in the state of Saxony-Anhalt, the Alternative for Germany received 43.8 percent of the vote. The governing Christian Democrats, who supply the Chancellor of the Federal Republic, received 17.2. The offending party took thirty-nine seats, and required forty-two, so that the Republic survives by three chairs and a coalition arithmetic problem.

I do not raise these numbers to alarm anyone. I raise them because they establish the difficulty precisely.

The voters of Saxony-Anhalt were informed in advance. The domestic intelligence service had classified the state branch as a confirmed right-wing extremist effort. The classification was published. The broadcasters carried it. The Chancellor spoke. Museum directors, grandmothers, and the whole civic apparatus of the Federal Republic marched, petitioned, and explained. The voters were told what the party was, and then they went out on a Sunday and voted for it by a margin of more than two to one over the government.

Here we arrive at the true nature of the emergency:

The instrument of democracy has been discovered in the hands of persons not qualified to operate it.

THE FAILURE OF PREVIOUS EXPEDIENTS

I have turned my thoughts upon this subject for many years, and I find the projects of other men wholly mistaken in their computation.

The firewall was tried. The cordon was maintained with great discipline, at considerable cost to every governing coalition in the Republic, and the effect upon the offending vote share was to double it. Surveillance was authorized and telephones were tapped under the lowered legal threshold that the extremist classification so conveniently provides. Documentaries were commissioned. School curricula were revised. Ten thousand articles were written explaining to the population of the eastern states what the eastern states were about to do, and why they must not.

And on the sixth of September they did it, in the largest numbers ever recorded, in a state that Berlin has been explaining things to for thirty-six years.

The lesson is not that the explanations failed. The lesson is that explanation, as a technology, has reached the limit of what it can accomplish against a citizen who has decided that his electricity bill is a political document.

THE CONTAGION

The danger of spread is now acknowledged by every responsible authority. The French finance minister has called the result a very, very worrying signal for Europe. The Prime Minister of Poland has observed that only fools and traitors could welcome it. From Washington, the ambassadorial class has cheered, which settles the question of authorship.

In the autumn of 2022 the Italian Republic elected Giorgia Meloni, and we were assured by the same publications, in the same typeface, that the Duce had returned and that the Constitution of 1948 would be waste paper by Christmas. Four years have passed. The Constitution is intact. The lady sits at every summit, is photographed beside every head of government, and is now described in the serious press as a moderate and a pragmatist.

This is the most frightening feature of the disease. The contagion does not kill its host. It governs the host, competently enough that after a few years nobody can remember what the emergency was.

From Rome the sickness passed to the Netherlands, then to the Nordic countries, then across the Atlantic, where it has taken the form of an American president who cannot be blamed for the weather during most news cycles, but is blamed for everything else. It has reached Canada. It threatens France, where the socialist paradise built by three generations of enlightened administrators may be overturned by capitalism. Even the European Council, that last citadel of governance by persons nobody elected, may fall.

All of it under the false populist banner of freedom, which is the word the disease uses when it wants to be let in.

THE ROMANIAN PRECEDENT

I anticipate the objection that my scheme is novel and untried. It is neither. It has been done, it worked, and the sky remained where it was.

In November of 2024, the Romanian people held the first round of a presidential election. An obscure gentleman named Calin Georgescu came first, having been assembled almost entirely out of TikTok. On the sixth of December, two days before the runoff, with diaspora voting already under way, the Constitutional Court of Romania set the whole round aside. Intelligence findings were declassified concerning coordinated social media manipulation and undeclared campaign funding.

No party was banned that week, and no citizen was turned away from any ballot box. The votes were cast in perfect freedom, counted in public, and then correctly understood. What had appeared on Sunday evening to be a tally became, upon expert inspection, evidence of foreign interference, which is a far more useful thing for a tally to be. A number can only be counted. Evidence can be interpreted, appealed, published, and used.

When the election was held again in May, the labor of a second interpretation was spared by the simpler expedient of removing the leading candidate from the ballot. Therefore, the correct man won with 53.6 percent. Brussels congratulated Bucharest on the resilience of Romanian democracy, and the resilience was undeniable, since Romanian democracy had proved capable of surviving a Romanian election.

The method was even announced in advance, which is the mark of a confident administration. In January of 2025, the former commissioner Thierry Breton observed on French radio that “We did it in Romania,” and added that the same would obviously have to be done in Germany if it became necessary. A written question was duly put to the Commission in the Parliament. Mr. Breton then clarified that the Union possesses no mechanism whatever for annulling any election, which is entirely correct, and is the very reason the mechanism performs so well. An instrument that does not exist requires no legal basis, no right of appeal, and no budget line.

I note in passing, and without any wish to alarm, that the Romanian election was voided over a TikTok operation, and that the gentleman who has just carried Saxony-Anhalt is himself a product of TikTok, thirty-five years old, formerly employed in marketing, and reportedly amiable. The precedent is therefore not merely applicable. It is exact.

THE PROPOSAL

I shall now therefore humbly offer my own thoughts, which I hope will not be liable to the least objection.

I propose that elections in the afflicted states continue exactly as at present. The ballot papers shall be printed. The polling stations shall open at eight. The queues shall be photographed. The exit polls shall be broadcast at six with the customary music. Nothing whatever of the ceremony shall be diminished, for the ceremony is the part that the international community observes.

I propose only that the ballot boxes, upon closing, be forwarded unopened to a Commission of Qualified Interpreters sitting in Brussels, who shall determine what the voters meant. The count itself shall then be annulled, exactly as the Romanian count was annulled, and the finding of the Commission shall stand in its place as the result of the election. I would not have anyone call this an overturning. An election is not overturned by being interpreted. It is completed, since a vote that has not yet been correctly understood cannot be said to have finished happening.

The principle is already conceded in every article written since Sunday. We are told on all sides that a vote for this party is not really a vote for this party. It is a cry of pain. It is a howl of economic anxiety, a symptom of postindustrial dislocation, an artifact of Russian (dis)information operations, a wound left by reunification. Every serious commentator agrees that the ballot means something other than what is marked upon it.

A cry of pain is a clinical finding. Clinical findings require trained interpretation. No one would permit a patient to read his own scan and prescribe his own surgery, and yet we permit forty-three percent of a German state to interpret its own suffering and act upon the diagnosis.

The advantages are considerable, and I shall enumerate them.

First, the offending party is removed from every parliament at a stroke, without any party being banned, any ballot being denied, or any citizen being disenfranchised in the technical sense that matters for the annual reports. Democracy is preserved in its strongest and most durable form, which is the ceremonial form.

Second, employment is secured for the interpretive class: the extremism researchers, the disinformation scholars, the foundation officers, the correspondents who have spent a decade explaining eastern Germany to readers in west London. These persons have accumulated an expertise that no longer has any electoral outlet, and it would be a waste of public investment to strand it.

Third, the voters themselves are relieved of the anxiety of consequence. A man may mark his ballot with perfect freedom, knowing that his mark cannot injure his neighbor, his country, or the Union. He may vote as extravagantly as he wishes. Nothing follows. This is the greatest expansion of political liberty since the secret ballot.

Fourth, the pain, once properly interpreted, can be addressed by the instruments proper to pain: a listening tour, a regional fund, a museum of memory in Halle with a café.

Fifth, and I raise this with some delicacy, the Commission will put an end to the problem of cheerfulness. The lead candidate in Saxony-Anhalt is thirty-five years old, a former marketing executive, and the correspondents keep reporting, with visible unease, that he is upbeat. He smiles on TikTok, packages the thing in positive slogans, and appears at rallies in the manner of a pop singer. This is far more dangerous than anger, because anger can be filmed and captioned. Under my proposal, no cheerfulness will ever again appear on a ballot, since no ballot will ever again be opened.

OBJECTIONS

Some will object that the scheme is undemocratic. This misunderstands the nature of a value. Democracy is a value, and values are not consumed by use; they are, if anything, better preserved by disuse, as with the good china, which survives precisely because it never comes down from the shelf.

Others will object that the population will not stand for it. Bucharest stood for it. The crowds came out in December, thinned by March, and were gone by summer, and the man they had voted for now describes himself as a passive observer of public life. Six months is the full natural lifespan of an annulled electorate.

Others will object that the eastern population will notice. They will not. They have been told for two decades that their votes were the product of Russian interference, economic frustration, and unresolved historical trauma. Being told at last that their votes will be professionally interpreted is a promotion.

Let no man speak to me of other expedients. Of asking why industrial electricity in Germany costs what it costs, and where the chemical plants went, and who signed the papers. Of asking why a man in Halle believes that nobody in Berlin has put a question to him in twenty years and waited for the answer. Of settling immigration by open legislative debate and a recorded vote, rather than by announcement followed by a decade of denouncing the objectors as morally defective. Of a governing party keeping one promise made at one election, as an experiment, to see what happens.

These are wild and visionary schemes. I would desire those politicians who dislike my overture to answer me one question first: whether they have ever, in living memory, asked 43.8 percent of a population what it wanted and then given it to them.

I profess, in the sincerity of my heart, that I have not the least personal interest in promoting this necessary work. I hold no chair in extremism studies. I have no column to fill on Mondays. I sit on the board of no foundation established to defend the information space, and I have received no grant to study democratic resilience. My own country is distant and, I am assured by the same newspapers, already lost.

I offer the proposal from pure public spirit, and from tenderness toward the poor voters of Saxony-Anhalt, who have suffered the misfortune of being asked a question, and the far greater misfortune of answering it.

MALONE.NEWS

No grant, no foundation, no chair in extremism studies. Jill and I read the primary documents ourselves, which is slow, and paid subscribers are what buys the time to do it.

The interpreters will always be paid. Whether we are is up to you. Malone News is a reader-supported publication. To receive new posts and support our work, consider becoming a free or paid subscriber.

Audio Version:

American families have spent decades applying herbicides and pesticides to American lawns. Dogs walk through them barefoot. Children crawl and play on them. The remarkable thing is how little serious research has been done to determine the extent of the biological injury and long-term health effects associated with these exposures.

Every spring, lawn-care trucks begin making their rounds through American neighborhoods. Herbicides kill the broadleaf weeds. Insecticides kill the grubs and other insects. Fungicides may be added where needed. On some occasions, a small sign is planted in the yard advising people and pets to remain off the grass for a period of time. Then the sign disappears, the children and dogs go back outside, and everyone assumes the problem has disappeared with it. Or, more likely, no signage is ever posted, and no one questions whether that lawn has a toxic film lingering on its blades of grass.

There is an important distinction that is easily overlooked in studies of residential pesticide exposure: adults, children, and dogs do not use a lawn in the same way. An adult generally walks across grass wearing shoes, remains upright, and has little direct contact with the ground. A dog walks across it on bare paws, lies in it, rolls in it, carries residues on its feet and coat into the house, and then licks its paws and grooms its body. A young child sits and plays on the grass, repeatedly touches the ground, puts hands and objects into the mouth, and then spends considerable time playing on the floors of the same house.

Those are fundamentally different routes of exposure. And the limited research that has actually measured what happens after a lawn is treated suggests that distinction matters.

There is also the matter of residue. Sunlight breaks down some chemicals more readily than others. Rain may wash residues from treated surfaces, while dew can dilute or redistribute them. How long these chemicals remain on grass, soil, and other surfaces varies considerably depending on the chemical, weather conditions, and how and where it was applied.

The rules governing treated lawns are a patchwork of federal and state law. At the federal level, EPA-approved pesticide labels are legally enforceable under FIFRA, and many lawn products instruct applicators to keep people and pets off treated turf until the spray has dried, or until a granular product has been watered in and the surface is dry. States can impose additional notification requirements, and these vary considerably.

New York, for example, requires commercial lawn applicators to post markers and instructs that they remain for at least 24 hours, while New Jersey requires signs for commercial turf applications that may be removed after 72 hours. Virginia is much less restrictive: its pesticide law does not generally require commercial applicators to notify or post residential lawns unless the pesticide label itself requires it, although HOAs must post conspicuous notice in common areas at least 48 hours before an application.

A national review published in 2000 found that roughly 30 states had no statewide requirement for posting treated lawns. More than two decades later, there appears to be no comprehensive, publicly available 50-state review documenting how those laws have changed, leaving even the current extent of pesticide-notification requirements surprisingly difficult to determine.

Most importantly, “dry” does not mean “gone.” It means that the liquid carrier has dried sufficiently to satisfy the product's reentry instruction. The pesticide and its residues may remain on grass, soil, leaves, shoes, paws and other surfaces well after the lawn is dry.

There is no federal requirement that pesticide-treated public parks be routinely posted for the public, and state requirements vary widely. That means a baby crawling on exposed grass is more likely than not to come into contact with herbicides and pesticides soon after application.

Federal law also does not require warning signs after pesticides are applied in public parks, on school grounds, or at playgrounds.

Lawn Chemicals, Dogs and Cancer

In 2004, researchers studied Scottish Terriers, a breed with an unusually high susceptibility to transitional cell carcinoma of the bladder. Eighty-three dogs with bladder cancer were compared with 83 controls. Dogs exposed to lawns or gardens treated with herbicides had 3.6 times the odds of developing bladder cancer compared with dogs exposed to untreated lawns. Exposure to both herbicides and insecticides was associated with 7.2 times the odds, while exposure specifically to phenoxy herbicides, the family that includes 2,4-D, was associated with approximately 4.4 times the odds.

This was an observational study and cannot establish with 100% certainty that the chemicals caused the cancers; the genetic susceptibility of Scottish Terriers makes them an unusual study population. Nevertheless, the magnitude of the associations raised an obvious additional question: were dogs living around treated lawns actually absorbing these chemicals?

An earlier study had measured the common lawn herbicide 2,4-D in canine urine. Among 44 dogs potentially exposed to 2,4-D-treated lawns, 75 percent had urinary concentrations of at least 10 μg/L. Dogs with access to lawns treated within the preceding seven days had the highest urinary concentrations.

In 2013, Purdue researchers examined this question more directly by measuring common herbicides on residential lawns and in the urine of dogs living at those homes. Herbicides were detected in the animals, including in some dogs whose owners reported that their own lawns had not been treated. The authors raised the possibility that these dogs were being exposed through neighboring properties, walks, or other environmental sources.

There is also evidence involving canine lymphoma. A 1991 case-control study reported an association between owners’ use of 2,4-D herbicides and malignant lymphoma in their dogs, although the finding subsequently became controversial and an industry-funded reanalysis challenged the strength of the association. A larger 2012 study involving more than 700 dogs found that professionally applied lawn pesticides were associated with approximately 70 percent higher odds of canine malignant lymphoma.

Given the widespread use of these products, the findings warranted a much larger and more systematic research effort. That research effort largely failed to materialize.

Residential Transport of 2,4-D Indoors

One of the most informative studies was not a cancer study at all. Researchers applied 2,4-D to residential lawns and then sampled the interiors of the homes before and after application. They tested indoor air, floors, tables, windowsills and household dust.

After the lawn was treated, 2,4-D was detected in indoor air and on surfaces throughout the homes. The investigators identified the homeowner who applied the chemical and an active family dog as the two most important mechanisms carrying 2,4-D indoors. Once inside, contaminated floor dust was resuspended into the air and redistributed onto tables and other surfaces.

This is where the canine and childhood questions converge. The dog is not merely exposed to the treated lawn. It can become a vehicle carrying residues from the lawn into the environment where a child spends much of the day.

The investigators estimated that nondietary 2,4-D exposure for young children after lawn treatment could be approximately ten times higher than before application. Their estimated exposures from contact with floors alone were 1–10 μg per day, with additional exposure estimated from contact with tabletops.

That finding also challenges the comforting assumption that the principal exposure ends once the spray has dried. The lawn is only the beginning of the pathway. Material can move from grass to shoes and paws, from shoes and paws to floors, from floors into dust and air, and ultimately from hands to mouths.

The Childhood Cancer Studies

A systematic review and meta-analysis of 15 studies found that residential herbicide exposure during pregnancy was associated with 61 percent higher odds of childhood leukemia. Residential insecticide exposure during pregnancy showed an even larger association, approximately doubling the odds. Childhood exposure to residential insecticides was also associated with increased leukemia risk, although childhood herbicide exposure itself was not statistically significant in that particular analysis.

A separate meta-analysis published in Pediatrics in 2015 examined 16 studies involving approximately 7,400 childhood cancer cases and 9,400 controls. Childhood exposure to residential herbicides was associated with 26 percent higher odds of leukemia. Indoor residential insecticide exposure was associated with 47 percent higher odds of leukemia and 43 percent higher odds of lymphoma. The authors emphasized that the relatively small number of available studies remained an important limitation.

What these studies generally do not tell us is what happens decades later. We have remarkably little prospective research following children with documented pesticide or herbicide exposure into their thirties, forties, and beyond to determine whether early-life exposure alters their subsequent cancer risk.

That question has become more important as the incidence of several cancers in younger adults has increased. A major review in Nature Reviews Clinical Oncology concluded that early-onset cancers, generally defined as cancers diagnosed before age 50, have been increasing in multiple countries since the 1990s. The increases involve cancers of the breast, colorectum, endometrium, kidney, pancreas, stomach, thyroid and several other organs. The authors specifically identified changes in the early-life exposome, including environmental exposures, as one area requiring much more investigation and called for prospective studies capable of linking childhood exposures with cancers occurring decades later.

The trend is also apparent in U.S. cancer-registry data. An analysis of more than 562,000 Americans diagnosed with cancer before age 50 found that overall early-onset cancer incidence increased between 2010 and 2019, while cancer incidence among people 50 and older declined. The increase was particularly apparent among adults aged 30 to 39, and gastrointestinal cancers showed the fastest increase as a group. Globally, an analysis of the Global Burden of Disease database reported a 79 percent increase in the number of early-onset cancer cases between 1990 and 2019, although population growth and demographic changes account for part of that increase.

We have evidence that children are exposed, epidemiological evidence associating residential pesticide exposure with childhood cancers, and a documented rise in several cancers among younger adults. What is largely missing are the long-term studies needed to determine whether those observations are connected.

What Is MAHA-GOV Doing?

The MAHA initiative has begun to address some of these larger questions. In 2026, the Administration directed HHS, USDA, and EPA to develop a research framework for cumulative chemical exposures and directed HHS to establish an NIH challenge to improve the measurement and evaluation of those exposures. NIH and NIEHS have also begun encouraging new research that combines environmental monitoring, biomonitoring, and biomedical research to understand how environmental chemicals, including pesticides, affect health over a lifetime. EPA, meanwhile, is currently conducting its registration review of 2,4-D, with an interim regulatory decision scheduled for 2026.

These are important developments. But I can find no new federal research program specifically following children and companion animals living in homes where lawns are chemically treated, despite decades of evidence showing that lawn chemicals enter homes, are absorbed by exposed animals, and are associated in epidemiological studies with serious health outcomes.

Studies have reported associations between lawn herbicide and pesticide exposure and bladder cancer and lymphoma in dogs, while human studies and meta-analyses have associated residential pesticide exposure during pregnancy and childhood with increased risks of childhood leukemia and lymphoma.

Nor do I find a new MAHA-era federal requirement for warning signs following pesticide applications to residential lawns, parks or school grounds. EPA’s current public guidance on 2,4-D continues to state that registered products can be used safely when label directions are followed.

MAHA has therefore identified the larger problem of environmental and cumulative chemical exposure, but the residential lawn remains an important gap. A prospective study measuring lawn residues, household contamination, and biological exposure in children, adults, and dogs would align closely with MAHA’s stated commitment to studying environmental contributors to chronic disease. After more than thirty years of preliminary evidence, this would seem an appropriate time to conduct it.

MAHA-Gov has apparently decided that independent “truth-tellers” like me are no longer welcome to communicate directly with government officials. So I am asking readers to do it instead.

Please contact your members of Congress, the EPA and NIH and ask them to support independent, prospective research into the long-term health effects of residential herbicide and pesticide exposure, particularly in children and companion animals.

Ask EPA to reconsider whether current notification, signage and reentry requirements adequately protect children and pets on lawns, school grounds and public parks. Feel free to reference or send them this article.

Thirty Years of Evidence, but No Prospective Answer

This is ultimately what I find most troubling about the literature. The first study reporting an association between lawn use of 2,4-D and canine lymphoma appeared in 1991. Researchers measured 2,4-D in the urine of dogs exposed to treated lawns in 1994. By 2001, researchers had demonstrated that lawn-applied 2,4-D could be transported into homes and estimated that nondietary exposure among young children increased substantially following application. In 2004, the Scottish Terrier study reported strong associations between exposure to treated lawns and bladder cancer.

The federal government participated in some of this research. EPA supported work examining the movement of lawn-applied 2,4-D into homes and children’s potential exposure. NIH, NCI and EPA supported important research examining residential pesticide exposure and childhood leukemia, and National Cancer Institute investigators participated in the original canine lymphoma study. These were not obscure observations occurring entirely outside the federal research establishment.

This early work done decades ago, was followed by… crickets.

What did not follow was a large, prospective research program designed to determine what these exposures mean over time. Much of the subsequent canine research depended instead upon universities, breed organizations, the American Kennel Club Canine Health Foundation, and private funding. Despite decades of concern about environmental exposures and cancer, including the Cancer Moonshot’s emphasis on cancer prevention, I can find no major federal program that followed children and dogs living in households with chemically treated lawns while directly measuring their exposures and subsequent health outcomes.

Such research is entirely feasible. Households could be enrolled before the lawn-treatment season, including families that use professional lawn services and those that do not chemically treat their lawns. Every product and concentration applied could be documented. Grass, soil and household dust could be sampled before and after treatment, while urine or other biological samples could be collected repeatedly from adults, children and dogs living in the same homes. Rather than attempting to reconstruct exposure years later from questionnaires, investigators could measure what actually enters the body, how long residues remain in the home, and how repeated applications affect exposure over an entire season. Long-term follow-up could then begin to address the much more important question of whether these measured exposures are associated with chronic disease. This could have all been done by now, including adult follow-up of pregnancy and childhood exposures.

But after more than thirty years of findings involving measurable canine exposure, transport of lawn chemicals into homes, associations with cancers in dogs, and epidemiological associations between residential pesticide exposure and childhood cancers, the absence of definitive evidence of harm cannot reasonably be treated as evidence of safety.

These are also largely avoidable exposures. We are not discussing a drug needed to treat disease or an unavoidable environmental contaminant. Residential herbicides and insecticides are commonly applied because homeowners want lawns without weeds or insects, to public parks and school grounds. Children and dogs, meanwhile, have substantially more direct contact with grass, soil and household floors than most adults and have behaviors that can increase ingestion of residues.

More than thirty years after the first of these findings, we are still applying these chemicals to lawns, parks, and other places where children and animals play without adequately answering the long-term health question.

A green, weed-free lawn is a trivial benefit against which to accept an avoidable and still poorly characterized exposure. Until we know considerably more, I see little justification for routinely applying these chemicals where children and animals live and play.

References

1. Glickman LT, Raghavan M, Knapp DW, Bonney PL, Dawson MH. Herbicide exposure and the risk of transitional cell carcinoma of the urinary bladder in Scottish Terriers. J Am Vet Med Assoc. 2004;224(8):1290–1297. doi:10.2460/javma.2004.224.1290.

2. Reynolds PM, Reif JS, Ramsdell HS, Tessari JD. Canine exposure to herbicide-treated lawns and urinary excretion of 2,4-dichlorophenoxyacetic acid. Cancer Epidemiol Biomarkers Prev. 1994;3(3):233–237.

3. Knapp DW, Peer WA, Conteh A, et al. Detection of herbicides in the urine of pet dogs following home lawn chemical application. Sci Total Environ. 2013;456–457:34–41. doi:10.1016/j.scitotenv.2013.03.019.

4. Hayes HM, Tarone RE, Cantor KP, Jessen CR, McCurnin DM, Richardson RC. Case-control study of canine malignant lymphoma: positive association with dog owner’s use of 2,4-dichlorophenoxyacetic acid herbicides. J Natl Cancer Inst. 1991;83(17):1226–1231.

5. Takashima-Uebelhoer BB, Barber LG, Zagarins SE, et al. Household chemical exposures and the risk of canine malignant lymphoma, a model for human non-Hodgkin’s lymphoma. Environ Res. 2012;112:171–176.

6. Nishioka MG, Lewis RG, Brinkman MC, Burkholder HM, Hines CE, Menkedick JR. Distribution of 2,4-D in air and on surfaces inside residences after lawn applications: comparing exposure estimates from various media for young children. Environ Health Perspect. 2001;109(11):1185–1191. doi:10.1289/ehp.011091185.

7. Turner MC, Wigle DT, Krewski D. Residential pesticides and childhood leukemia: a systematic review and meta-analysis. Environ Health Perspect. 2010;118(1):33–41.

8. Chen M, Chang CH, Tao L, Lu C. Residential exposure to pesticide during childhood and childhood cancers: a meta-analysis. Pediatrics. 2015;136(4):719–729. doi:10.1542/peds.2015-0006.

   Zahm SH, Weisenburger DD, Babbitt PA, et al. A case-control study of non-Hodgkin’s lymphoma and the herbicide 2,4-dichlorophenoxyacetic acid (2,4-D) in eastern Nebraska. Epidemiology. 1990;1(5):349–356. doi:10.1097/00001648-199009000-00004.

9. Smith AM, Smith MT, La Merrill MA, Liaw J, Steinmaus C. 2,4-Dichlorophenoxyacetic acid (2,4-D) and risk of non-Hodgkin lymphoma: a meta-analysis accounting for exposure levels. Ann Epidemiol. 2017;27(4):281–289.e4. doi:10.1016/j.annepidem.2017.03.003.

10. International Agency for Research on Cancer. 2,4-Dichlorophenoxyacetic acid. In: DDT, Lindane, and 2,4-D. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Vol. 113. Lyon, France: International Agency for Research on Cancer; 2018.

11. Ugai T, Sasamoto N, Lee HY, et al. Is early-onset cancer an emerging global epidemic? Current evidence and future implications. Nat Rev Clin Oncol. 2022;19(10):656–673. doi:10.1038/s41571-022-00672-8.

12. Koh B, Tan DJH, Ng CH, et al. Patterns in cancer incidence among people younger than 50 years in the US, 2010 to 2019. JAMA Netw Open. 2023;6(8). doi:10.1001/jamanetworkopen.2023.28171.

13. Zhao J, Xu L, Sun J, et al. Global trends in incidence, death, burden and risk factors of early-onset cancer from 1990 to 2019. BMJ Oncol. 2023;2. doi:10.1136/bmjonc-2023-000049.

14. U.S. Environmental Protection Agency. 2,4-D. Ingredients Used in Pesticide Products. Updated January 23, 2026. EPA notes that 2,4-D has been used since the 1940s, is used on turf and residential lawns, remains under registration review, and states that registered products can be safely used when label directions are followed.

15. U.S. Environmental Protection Agency. Introduction to Pesticide Labels. EPA. 2026. Describes pesticide labeling requirements under the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) and explains that pesticide labels are legally enforceable.

16. New York Environmental Conservation Law § 33-1003. Visual Notification. Requires commercial lawn applicators to place visible markers on treated property and instructs that the markers not be removed for at least 24 hours.

17. New Jersey Administrative Code § 7:30-9.13. Notification: Turf or Ornamental Applications. Requires posting following commercial turf pesticide applications, including applications at residences, schools, institutions and parks; signs may be removed after 72 hours.

18. Code of Virginia § 55.1-1818. Common Areas; Notice of Pesticide Application. Requires property owners’ associations to post conspicuous notice in common areas at least 48 hours before pesticide application.

19. U.S. Environmental Protection Agency. Pest Control in the School Environment: Implementing Integrated Pest Management (IPM). EPA 735-F-17-004. Washington, DC: EPA; 2017. Discusses pesticide notification and posting at schools and notes that requirements vary by state and locality.

20. The White House. Executive Order 14414: Advancing Regenerative Agriculture and Strengthening American Farm Resilience. June 25, 2026. Directs USDA, HHS and EPA to develop a research and evaluation framework for cumulative chemical exposure and directs HHS, in consultation with USDA and EPA, to establish an NIH challenge addressing cumulative chemical exposures.

21. U.S. Department of Health and Human Services. President Trump Signs Executive Order Advancing Regenerative Agriculture as Part of Make America Healthy Again Agenda. June 25, 2026. Announces an NIH Grand Prize Challenge addressing cumulative chemical exposures and ARPA-H research intended to reduce reliance on conventional chemical crop-protection tools.

22. National Institutes of Health. Integrating Environmental Science and Engineering with Biomedical Research for Effective Exposure Prevention and Disease Intervention. NIH Highlighted Topics. Posted June 18, 2026. Encourages research combining environmental monitoring, personal exposure monitoring, biomonitoring and biomedical research involving environmental chemicals, including pesticides.

23. National Institutes of Health. Characterizing Indoor Exposures to Advance Exposome Research. NIH Highlighted Topics. Posted July 9, 2026. Calls for research identifying sources of indoor environmental exposure, quantifying exposure, and evaluating contributions to health burden.

24. National Institutes of Health. Detecting and Evaluating Contaminants of Emerging Concern: Optimizing Data-driven Exposure Characterization. NIH Highlighted Topics. Posted July 31, 2026. Identifies major gaps in understanding environmental exposure pathways, biological uptake and chronic outcomes including endocrine disruption, neurotoxicity, immune dysfunction, reproductive toxicity and cancer.

What is Labor Day?

Labor Day began in the late nineteenth century as a recognition of American workers at a time when industrial labor was often dangerous, poorly paid, and largely unprotected. Congress made it a federal holiday in 1894, following years of organizing and bitter labor disputes.

Much has changed since then. Federal and state laws now establish protections involving workplace safety, wages and hours, discrimination, and child labor. Labor Unions have largely become outdated, and yet, they persist in strength and number.

The unions, however, did not exactly declare victory and go home. Many of today’s large unions increasingly resemble political NGOs, complete with permanent bureaucracies, lobbying operations, enormous political war chests, and agendas that sometimes seem rather far removed from the guy fixing a power line or pouring concrete.

That does not make Labor Day obsolete. If anything, it provides a reason to reclaim it. Labor Day should be about the dignity of productive work and the Americans who build things, grow food, care for others, run businesses, fix things, and keep the country functioning, rather than the institutions that have appointed themselves to speak for them.

What Americans have historically been good at is hard work, innovation, and problem-solving.

So, Happy Labor Day to the people who actually labor.

… Shortly thereafter, the applicant was offered the CEO position

Malone News is a reader-supported publication. To receive new posts and support our work, consider becoming a free or paid subscriber.

JGM

Audio Version:

Back in 2020, Pew Research Center surveyed more than 11,000 Americans. Political scientist Zach Goldberg subsequently examined the underlying data and broke the results down by age, sex, race, and political ideology.

What he found was rather remarkable.

Among white women ages 18 to 29 who identified as liberal, 56.3% reported that a doctor or other health-care provider had diagnosed them with a mental health condition.

• Not 15%.

• Not 25%.

• Fifty-six percent.

  ---

• For white moderate women of the same age, the number was 28.4%.

• For white conservative women, 27.3%.

In other words, young white liberal women were more than twice as likely as their conservative counterparts to report having received a mental health diagnosis.

Now, correlation is not causation. Perhaps liberals are simply more willing to seek psychiatric care. Perhaps conservatives are less willing to admit that they have a problem. Perhaps political ideology attracts people with particular personality traits rather than creating those traits. All perfectly reasonable possibilities.

But here is where it gets interesting.

Five years later, the numbers have not exactly moved in a reassuring direction.

Gallup found that only 15% of American women ages 18–29 surveyed from 2020 through 2024 described their mental health as excellent. In 2010–2014, that number had been 48%.

That is not a small decline. That is a collapse.

And the ideological divide is still there.

The 2024 American Family Survey found that only 12% of liberal women ages 18–40 described themselves as “completely satisfied” with their lives.

• Among moderate women: 28%.

• Among conservative women: 37%.

Liberal women were also nearly three times as likely as conservative women to report feeling lonely several times a week or more: 29% versus 11%.

So perhaps endlessly telling young women that they live in an oppressive society, that the world is on the verge of climate catastrophe, that words are violence, that disagreement is dangerous, that our President is evil, that the patriarchy is lurking behind every corner, that Western civilization is fundamentally corrupt, and that their personal unhappiness is caused by vast structural forces over which they have virtually no control is not actually a recipe for human flourishing.

Who could have guessed?

There is an important caveat. A 2025 study found that some of the apparent liberal-conservative mental-health divide may reflect differences in how people think and talk about “mental health.” When researchers simply asked people about their mood, rather than their mental health, much of the ideological difference disappeared.

Fair enough.

But that doesn’t explain away the extraordinary collapse in reported well-being among young women generally, nor does it make the life-satisfaction and loneliness data disappear. Maybe the question isn’t whether liberalism causes mental illness.

Maybe the more interesting question is whether a culture built around grievance, victimhood, fear and powerlessness is particularly attractive to unhappy people, makes unhappy people unhappier, or both. That seems like a question worth asking. Even if asking it makes someone very, very upset.

There is, however, an uncomfortable counterpoint to all this discussion of female unhappiness and mental illness. Men are far more likely to actually die by suicide. In the United States, the male suicide rate is nearly four times the female rate, and men account for roughly four out of every five suicide deaths. The disparity exists globally as well. Perhaps that exposes a deeper truth about the modern cult of victimhood. We have created a culture extraordinarily skilled at teaching certain groups to name every grievance, diagnose every distress, identify every oppressor and demand that society recognize their pain, while other forms of suffering remain remarkably easy to ignore. Young women may be reporting extraordinary levels of anxiety, depression and psychiatric diagnosis, but men are quietly killing themselves at vastly higher rates.

That does not diminish women's suffering. It suggests that our entire framework for understanding suffering may be badly distorted. A society obsessed with determining who is the victim and who is the oppressor can become remarkably blind to the human being who doesn't fit the approved narrative, particularly when that human being is male.

There is another possibility here that deserves far more attention: the diagnosis itself has become part of the epidemic. When more than half of young women in a political demographic report having been diagnosed with a mental health condition, the appropriate response is not simply to declare that half the population is mentally ill. It is to ask what the hell the health-care system is doing. Psychiatry has steadily expanded the boundaries of pathology, medicalizing distress, grief, anxiety, loneliness, adolescent turmoil, and ordinary difficulty as diagnosable disease. A system that can convert normal human suffering into a billing code, a prescription, and a lifelong medical identity should not be treated as a neutral observer of this crisis. At some point, an exploding diagnosis rate stops being proof of exploding disease and starts becoming evidence that the diagnostic machinery itself may be broken.

Consider what we are now being asked to accept as normal. The 2026 Lancet Global Burden of Disease study estimates that 1.17 billion human beings, roughly one person in seven on Earth, are living with a mental disorder, and mental disorders now account for 17.3 percent of all years lived with disability worldwide, making them the single largest source of nonfatal disability. Women carry an even greater burden than men, with the female age-standardized mental-disorder burden about 18 percent higher. This is insane, not because hundreds of millions of people do not genuinely suffer from serious psychiatric illness, but because at some point a civilization has to ask what these numbers actually mean. Did humanity suddenly become catastrophically mentally ill, or have we built a medical system that increasingly converts sadness, fear, loneliness, grief, adolescent turmoil and difficulty coping with life into psychiatric pathology? When one in seven people on the planet fits into the modeled category of mental disorder, the diagnostic framework itself deserves scrutiny. A medical system cannot simply keep expanding the definition of sickness and then point to the resulting mountain of sickness as proof that it was right.

Toxic Feminism

There was a feminism that told women they should have choices. Then there is the more modern version of feminism that increasingly seems to tell women which choices an enlightened woman is supposed to make. Career is achievement. Marriage is dependence. Motherhood is unpaid labor. Men are potential oppressors. Traditional families are suspect. Finances should always be kept separate, plan for a divorce, financial independence is a must. Independence is liberation, and needing another human being is somehow weakness. Delay marriage. Delay children. Establish the career. Find yourself. Become independent. Never settle. Put yourself first.

There is just one rather awkward problem with this prescription for female happiness: human beings are profoundly social animals, and the scientific literature keeps rediscovering the importance of precisely the relationships that modern culture has spent decades teaching women not to need.

Femininity is Real

There is also an uncomfortable biological reality that modern ideology would prefer to treat as an inconvenience: women are not simply smaller men with different reproductive organs. Progesterone is central to female reproduction, but its effects extend into the brain and behavior. Human research has associated higher progesterone with affiliation, social bonding, sensitivity to relationships and even greater willingness to sacrifice for another person.

Estrogen, particularly estradiol, does far more than regulate reproduction. It acts throughout the female brain, interacting with oxytocin and other neuroendocrine systems involved in mood, sexuality, emotional processing, social attachment, and stress. Together with progesterone and oxytocin, estrogen contributes to the biological systems underlying bonding, maternal behavior, and caregiving. And that capacity for nurturing does not suddenly switch off when there is no baby in the room.

Women show stronger tendencies toward caregiving, empathy, and relational investment, although there is enormous individual variation. Biology is not destiny, but neither is it imaginary. To dismiss these deeply rooted female tendencies as merely products of the patriarchy, or to teach women that caring for and depending upon others is somehow incompatible with independence, comes at a real psychological cost, and we are witnessing this worldwide. A feminism that requires women to deny important parts of their own biology is a strange definition of female liberation.

The female impulse toward attachment, relationship and nurturing cannot simply be dismissed as a patriarchal social construct. There is biology underneath it. A culture that tells women that caring for others is servitude, that motherhood is an impediment to achievement, and that putting another human being ahead of oneself represents oppression may therefore be fighting something considerably older than the patriarchy. It may be fighting female biology itself.

Psychology even has a validated instrument called the Ambivalence Toward Men Inventory, developed specifically to measure hostile prejudices and stereotypes about men. And the hostility is measurable. In a subsequent study of 488 college students, women scored significantly higher than men on the instrument's Hostility Toward Men scale, 2.76 versus 2.25 on a 0-to-5 scale.

How a broader culture came to normalize the sweeping negative generalizations about men, male sexuality, male power and masculinity that would immediately be recognized as prejudice if the sexes were reversed is an interesting question. Hostility between the sexes is not liberation. It poisons precisely the trust, affection and intimacy upon which successful relationships depend.

The evidence surrounding work, family and mental health should make us reconsider what women have been sold as liberation. In one nationally representative study of 4,714 working women, high work-family conflict was associated with 2.29 times the odds of depressive symptoms, with some of the strongest associations occurring among younger, highly educated, higher-income and never-married women. Studies of female physicians and longitudinal studies of working women tell much the same story: when professional life and family life are placed into sustained conflict, mental health suffers.

The mistake was never telling women that they could become physicians, scientists, executives, farmers, entrepreneurs or anything else they were capable of becoming. Of course they could, and they should have been free to do so.

The mistake was turning independence itself into the objective, as though dependence upon people who love us were some sort of character defect. A husband and wife depend upon one another. Children depend upon parents. Parents eventually depend upon their children. Families are elaborate networks of mutual obligation. That isn’t oppression. It is much of what gives human life meaning.

And perhaps most poisonous of all is the idea that relations between men and women should be understood primarily as a contest for power. Teach a generation of women to approach half of humanity with suspicion, to interpret ordinary disagreement through the language of oppression, and to view compromise as surrender, and then act surprised when relationships become difficult. Whatever one chooses to call that worldview, it is a miserable foundation upon which to build intimacy.

Girls do not generally sit through a classroom lesson entitled Marriage Is Bad. The message is subtler than that. From school through college, the language surrounding female success increasingly emphasizes independence, autonomy, empowerment, career, individual rights and self-realization. What is striking is what receives far less emphasis: interdependence, marriage, sacrifice, obligation, motherhood, building a household, choosing a good husband and creating a stable family.

Dependence itself has acquired an almost pathological meaning, as though needing a husband, wife, children or extended family represents personal failure rather than the normal condition of human beings. A girl is repeatedly asked what career she wants when she grows up. How often is she asked what kind of marriage she hopes to build, what kind of mother she hopes to become, or what sacrifices a lasting relationship will require? The lesson does not have to be written on the blackboard to be learned: success is something you achieve individually; family is something you fit around it later, and you would be stupid not to be financially independent before having children, cause men generally aren’t to be trusted to hang around long enough to raise a family.

There is an enormous difference between telling a woman you do not have to marry or have children and telling her, explicitly or implicitly, that marriage and motherhood represent lesser ambitions than professional status. Biology also imposes a deadline that corporate advancement does not. A promotion can arrive at 42. A first child may not. Women deserve to know that when they are 22, not discover it when they are 39.

In our lives, we know two women in their thirties who are in the process of freezing their eggs; evidently, they believe that when they are forty or fifty, they will be in a better position to have a child? And we have another single friend, who at fifty, has decided she has aged out of ever being a mother. All of these women are college-educated professionals, never married, and are coming to terms with their biological clocks running out.

A culture that encourages women to postpone permanent relationships, prioritize professional status, regard dependence as weakness, and treat men principally as a source of danger or oppression should at least be willing to examine the results. If the promised liberation ends with unprecedented numbers of young women anxious, depressed, medicated, lonely and unable to form the families they later discover they wanted, declaring the experiment an unquestionable success is not feminism.

It is marketing.

The Big Lie

And then there is the part nobody is supposed to mention: this arrangement is extraordinarily good for industry. The intact family is an economic institution as well as a social one. Husbands and wives share labor. Grandparents care for grandchildren. Families cook meals, repair things, care for the sick and elderly, grow food, teach children, share houses, lend money, pass down skills and provide an enormous amount of economic activity that never appears on a corporate balance sheet. Break those bonds apart and suddenly nearly everything has to be purchased. Two adults maintaining two households need two rents or mortgages, two sets of utilities, two kitchens and twice the household goods.

Put both parents into full-time employment and families purchase more childcare, prepared food, transportation and other substitutes for work once performed within the home. Send rural children hundreds of miles away to college and then into metropolitan corporate employment, and another generation is separated from grandparents, land, community, and the informal economy of extended family. None of this requires a conspiracy. It requires only incentives. And corporations benefit enormously from this arrangement.

Corporations benefit when people become workers first, consumers second, and family members somewhere farther down the list. The Big Lie was convincing women (and men) that exchanging dependence upon people who loved them for dependence upon employers, banks, universities, childcare companies, and the consumer economy was somehow independence. The destruction of family and community did not make people independent. It merely changed whom they depended upon, and unlike a family, the corporation collected the dividends.

The Result is “Mental Illness”

Going back to those numbers at the beginning of this essay, perhaps we have been asking the wrong question. When 56.3 percent of young white liberal women report having been diagnosed with a mental health condition, the immediate assumption is that something must be terribly wrong with these women. Maybe instead, something is terribly wrong with the culture they have been raised in.

For decades, young people, and young women in particular, have been sold a remarkably destructive definition of success. Leave home. Go deeply into debt for college. Build the career. Delay marriage. Delay children. Treat dependence upon another person as weakness. Treat compromise as surrender. Regard traditional institutions with suspicion. Learn to identify grievances. Put yourself first. And when the loneliness, anxiety, purposelessness and unhappiness arrive, there is an enormous medical industry waiting to provide the diagnosis and, frequently, the prescription.

Then we marvel at the epidemic of mental illness.

Perhaps a significant portion of what we are calling a mental-health crisis is actually a crisis of meaning, belonging and human connection. A pill cannot provide a husband or wife. Therapy cannot manufacture grandchildren around the Thanksgiving table. A diagnosis cannot replace friendship, faith, family, community, useful work, children who need you, parents who depend upon you, or the knowledge that one’s life matters to people other than oneself. Medicine can treat disease. It is considerably less capable of treating a culture that has systematically stripped away many of the institutions that once gave ordinary people purpose.

And men have hardly escaped this experiment. While women report extraordinary levels of anxiety, depression and psychiatric diagnosis, men are killing themselves at vastly higher rates. That should be the flashing red warning light. The ideology of victimhood may teach women to describe their suffering endlessly while feminist teachings overlook the suffering of those assigned the role of “oppressor.”

None of this means returning women to some imaginary 1950s existence or denying anyone education, independence or a career. It means admitting something considerably more radical in 2026: human beings need one another. Men need women. Women need men. Children need parents. Parents need children. Families need communities, and communities need people who remain invested in them. Independence was never the highest human good.

That may be the real Big Lie. We were told that liberation meant needing fewer people, carrying fewer obligations and being free to put ourselves first.

Instead, we became more dependent upon employers, corporations, universities, government programs, therapists, pharmaceutical companies and the marketplace to provide things that families and communities once provided for one another.

And now 1.17 billion people are estimated to be living with a mental disorder.

Perhaps the strangest response imaginable is to look at that number and conclude that the problem is simply that humanity needs more mental-health treatment.

Maybe people need more life.

More family. More marriage. More children. More friendship. More community. More responsibility. More people depending upon them, and more people upon whom they can depend.

That isn’t oppression.

It is called being human.

By JGM/RWM

International diplomacy has reached the point where the Prime Minister of Canada is formally insisting that Americans stop making memes about Canada and himself.

On September 1, Canadian Prime Minister Mark Carney declared that trade talks could resume only when:

“when the Americans stop doing memes, stop throwing shade and stop trying to be tough, and start being serious about having those discussions.” He then added, “It’s not constructive, but that’s their democracy.”

Another News Video:

Remember when changing the weather was this easy!

By: JGM

Stolen Money, Missing Approvals

Federal grant capture and the consequences that never came.

Robert W. Malone, MD, MS

Audio Version:

Executive Summary

Poul Thorsen stole more than $1 million from a CDC-funded Danish research program and has now pleaded guilty.

Thorsen did not persuade someone at CDC to approve fraudulent invoices. He manufactured CDC approval himself. The indictment describes two separate streams of forgery. In May 2007 and March 2008, he sent the Danish Agency for Science, Technology and Innovation letters bearing forged signatures of CDC grant-management officers, falsely representing that federal funds had been awarded and were available. Between February 2004 and June 2008, he sent Aarhus University and Odense University Hospital more than a dozen invoices bearing the forged signature of a laboratory section chief at CDC’s National Center on Birth Defects and Developmental Disabilities (United States v. Thorsen, 2011, ¶¶ 7–8).

Aarhus then paid those invoices into accounts at the CDC Federal Credit Union in Atlanta. The university apparently believed it was sending money to CDC.

It was sending the money to Poul Thorsen.

He used it to buy a house on Briarlake Road, a Harley-Davidson motorcycle, an Audi, and a Honda. He converted still more into dozens of cashier’s checks. By the time it was over, more than $1 million had been diverted.

But the theft is the simplest part of this story.

The larger scandal is what the documents reveal about the federal research program surrounding him.

CDC scientist Diana Schendel helped create a sole-source funding opportunity for a Danish research program whose principal investigator was a man with whom she was having an undisclosed romantic relationship. She argued for sole sourcing, helped shape the program announcement, participated in the review process, accepted gifts from him, and co-authored research produced through the award. Those studies became part of the scientific evidence used by the United States government to reject claims that vaccines caused autism.

The required Danish human-subjects approvals were never obtained. CDC discovered the problem in 2009, after two of the papers had already been published. Rather than disclose the failure to the journals or to the special masters then adjudicating thousands of vaccine-autism claims, CDC personnel helped obtain retroactive ethical coverage by extending an unrelated approval from 2000.

Schendel was reprimanded. The reprimand remained in her personnel file for two years. She later left CDC and became a professor at Aarhus University, one of the institutions that had received the federal research money.

Much of the advocacy reporting about this case has nevertheless gotten an important part of the story wrong. Schendel did not approve Thorsen’s fraudulent invoices, and the evidence I have found does not establish that she participated in his theft. Thorsen forged CDC signatures precisely because he did not have legitimate authorization for the money.

The distinction matters because the documented case is more important than the exaggerated one.

This was a cooperative agreement, not a procurement contract. That distinction placed the award outside many of the competition, protest, conflict, recovery, and revolving-door safeguards that people assume govern federal spending. Scientific authority sat in one office. Financial authority sat in another. Ethics enforcement sat somewhere else. Each piece of the transaction could pass through its own bureaucratic channel without anyone being responsible for the whole.

This essay does not claim that the Danish autism data were fabricated. The record I have reviewed does not establish that - one way or the other. Of course, the CDC, doesn’t appear to have ever gone back and re-analyzed the study design, results, statistics, or papers - despite the fact that Poul Thorsen had stolen a million dollars and was the principal investigator responsible for the research program itself. One might think that further investigation is in order of the research.

This case establishes something else: a federal research system capable of putting more than $11 million behind politically consequential science despite an undisclosed intimate relationship, gifts, a sole-source award, missing human-subjects approvals, and a principal investigator who was stealing from the program, while producing almost no lasting institutional accountability.

That was the crime for which Poul Thorsen was eventually prosecuted. But it is not the institutional scandal this essay is about.

The question now is whether CDC will ever be required to account for everything else…

What Children’s Health Defense Got Wrong

Children’s Health Defense had the documents. It paid to obtain them. It published them. And then it repeatedly got some of the most basic facts of the case wrong.

That matters because the actual record is bad enough. The documented facts are damning enough. Getting them wrong only makes it easier for CDC, as well as public health pro-vaccine advocates, to dismiss the entire story.

In August 2026, Brian Hooker told The Defender that Diana Schendel “was Thorsen’s direct grant supervisor and lover and approved all of his invoices for expenditures from his CDC grant money.” He went further, suggesting that she “could have spent some of the stolen grant money as well” (Children’s Health Defense 2026). CHD’s own 2017 caption identifies Schendel as “supervisor to Poul Thorsen.” The MAHA Report similarly describes her as “the CDC administrator who was supposed to be supervising him” and claims that “she was involved, to some extent, in Thorsen’s financial fraud” (Conte 2026).

Those claims are not supported by the documents. In several important respects, they are flatly contradicted by them.

First, Schendel was not Thorsen’s employment supervisor. Thorsen was not a CDC employee during the grant period. He was the principal investigator for a foreign recipient institution, employed by Aarhus University and working in Denmark. Schendel was a CDC scientist collaborating on the project. Calling her his “direct grant supervisor” creates an authority relationship that the record does not establish.

Second, Schendel did not approve Thorsen’s fraudulent invoices. Nobody at CDC did. That is the entire point of the forgery scheme. The invoices were sent to Aarhus University and Odense University Hospital bearing the forged signature of a CDC laboratory section chief. Thorsen forged that signature precisely because he had no legitimate CDC authorization for the payments. The invoices never went through Schendel for approval.

Third, the evidence I have found does not establish that Schendel controlled the grant money, received the stolen money, or spent any of it. CDC’s own personnel records identify her as Lead Scientific Collaborator, while a November 2009 call sheet identifies her as Epidemiology Team Lead in the Developmental Disabilities Branch (Clay 2017, Exhibit 6). Those positions may raise serious questions about her scientific role, her influence over the cooperative agreement, and her conflicts of interest. They do not make her the financial officer who authorized Thorsen’s theft.

And that distinction is important. There is plenty to investigate about Schendel without accusing her of participating in a million-dollar theft for which the available evidence does not implicate her. She helped construct the sole-source opportunity. She argued for sole sourcing. She was romantically involved with the principal investigator. She accepted gifts from him. She co-authored the resulting research. Those are documented issues. Claiming she approved forged invoices only distracts from them.

Fourth, even the criminal charges have been repeatedly misstated. Thorsen was indicted on twenty-two counts total: thirteen counts of wire fraud and nine counts of money laundering. The MAHA Report turns that into twenty-two counts of wire fraud. It also says CDC officials indicted Thorsen. They did not. A federal grand jury indicted him. CDC employees do not have the power to indict anyone.

These are not obscure discrepancies buried in inaccessible government archives. Every one of them could have been checked against documents Children’s Health Defense itself obtained and published.

That is what makes this particularly frustrating. The FOIA work was valuable. The documents are important. They expose a story that deserved far more scrutiny than it received. But the reporting layered on top of those documents became careless, and in places speculative.

And that carelessness has consequences.

If you accuse Schendel of approving invoices she never approved, CDC can prove you wrong. If you call Thorsen her employee when he was not, CDC can prove you wrong. If you say he faced twenty-two wire-fraud counts when the indictment says thirteen, CDC can prove you wrong. If you suggest she spent stolen money without evidence that she did, CDC can dismiss the allegation as speculation.

Then something very convenient happens: the demonstrably false claims become the defense against the demonstrably true ones.

The undisclosed romantic relationship gets buried. The gifts get buried. The sole-source grant process gets buried. The missing ethics approvals get buried. The retroactive effort to cover them gets buried. And the absence of meaningful institutional consequences gets buried.

The underlying record does not need exaggeration. It needs accuracy. Because accurately told, the story is worse.

Schendel Did Not Control the Money. She Helped Control the Science.

To understand what Diana Schendel actually did, you have to understand how federal money moves. This distinction has been badly muddled in almost every retelling of the Thorsen story.

Federal money generally leaves an agency in one of two ways: acquisition or assistance. Different laws govern them, different officials control them, and confusing the two has led people to accuse Schendel of something the evidence does not show while missing what the evidence does show.

Acquisition means contracts. The government buys a good or service for its own use. The Federal Acquisition Regulation governs the process, and only a warranted Contracting Officer can bind the United States. That authority is specific and limited. Other agency employees may oversee technical work, but they cannot simply obligate federal money because they want to.

Assistance means grants and cooperative agreements. Here the government is funding someone else to carry out a public purpose. The Federal Grant and Cooperative Agreement Act of 1977 draws the basic distinction, and what separates a cooperative agreement from an ordinary grant is the expectation of substantial federal involvement in carrying out the work (31 U.S.C. § 6301 et seq.).

That distinction is crucial here.

The Denmark project was funded as assistance, specifically as a cooperative agreement. CDC was not simply writing a check and waiting for the results. The structure contemplated continuing involvement by CDC scientists in the work. Program Announcement 02006, later Cooperative Agreement U10 DD000230, was awarded first to the Danish Medical Research Council and then to DASTI, with more than $11 million flowing into the project from 2000 through 2009.

Three very different federal roles operated inside that system, and conflating them is one reason this story has been told so badly.

The Grants Management Officer held the business authority and signed the Notice of Award. At CDC during this period, that function sat in the Procurement and Grants Office. Aarhus University’s January 2010 statement identifies the official whose funding-commitment letter Thorsen used to obtain advances, and whose signature CDC subsequently suspected had been forged: Randolph B. Williams of CDC’s Procurement and Grants Office (Aarhus University 2010).

The Grants Management Specialist handled day-to-day grant administration. Nosrat Irannejad appears in that traffic in 2009.

Then there was the scientific side: the Project Officer, or in Schendel’s case, the Lead Scientific Collaborator. That person did not possess the authority to sign away federal money.

But do not mistake the absence of financial authority for the absence of power.

The scientific side helped shape the program announcement. It argued for the scientific concept. It participated in the justification for the award. It could suggest reviewers. It remained involved in calls and scientific decisions. And because this was a cooperative agreement rather than an ordinary grant, continuing federal scientific involvement was built into the relationship.

That was Schendel’s role.

She did not have the authority to approve Thorsen’s fraudulent invoices. She did not need to have it for the documented conflict to be serious.

The issue is not that Diana Schendel controlled the federal checkbook. The issue is that she helped shape a sole-source federal research program that sent millions of dollars to a project led by the man with whom she was having an undisclosed romantic relationship, and then became a co-author of the science that project produced.

That is a different allegation from participating in Thorsen’s theft.

And unlike the allegation that she approved his fraudulent invoices, it is supported by the documentary record.

There Was Only One applicant

Federal research money is normally competed. An agency publishes a funding opportunity, institutions apply, and a panel of reviewers scores the applications. Competition is the control that keeps a program scientist from choosing the recipient. She can argue for a subject, and she can shape what the agency asks for, but she cannot pick the winner, because people she does not control rank the applicants against each other.

A sole-source award removes that control. The agency declares that one entity alone can perform the work and writes the announcement so that only that entity qualifies. Nothing about this is illegal, and the justification here had substance. Denmark held population registers covering every birth in the country, and Danish privacy law barred American researchers from touching the data directly. If you want a Danish registry study, you fund Danes.

The justification is written by the person who wants the award to issue, and nobody outside her chain reviews whether the restriction is real or engineered to fit a chosen recipient. Once the field is narrowed to one, the review panel has nothing to compare the project to - there is no competition. The December 7, 2001 panel had a single applicant before it, met in the afternoon, and reported approval before the day ended.

Across the life of this award, Schendel held every one of those levers of control. She helped write the announcement, argued the sole-source case twice, and rewrote the concept in 2006 to make that case stronger. She obtained a five-year project period instead of the standard three, which stretches the interval between competitive renewals. She nominated names for the panel that would review the second competition. The recipient was a man she was involved with and had an intimate relationship with, from whom she accepted gifts, and with whom she published.

Taken together, they remove every external check on who received eleven million dollars. That is what sole-source contracting does when the person writing the justification has an undisclosed interest in the outcome. It converts one official’s preference into a federal award without requiring anyone else’s agreement. It allows conflicts of interest to creep into the process, in ways that are usually deemed unacceptable.

What she actually did

In November 2001 Diana Schendel, then a CDC employee, worked with colleagues to prepare the program announcement as a sole-source solicitation for the Danish Medical Research Council. She cleared it through the Office of Management and Budget and obtained an extension to a five-year project period rather than the standard three. The objective review panel met on the afternoon of December 7, 2001. There was one applicant. Before the day was out, the chair sent a memo to the center director reporting approval (Clay 2017).

In May 2006 Schendel argued internally for a second sole-source stream and rewrote the concept to strengthen the sole-source justification. In October 2006 she sent forward her preferred names for the Special Emphasis Panel that would review it. In November 2008 she asked colleagues to reissue a 2005 letter to Thorsen with a fresher date, because the institute leader at Aarhus wanted the reassurance in the first paragraph.

Under 5 C.F.R. 2635.202, an employee may not accept a gift from a prohibited source, and a current or prospective grantee is a prohibited source. Under 5 C.F.R. 2635.502, an employee whose impartiality would reasonably be questioned because of a covered relationship must not participate in the particular matter without authorization. Joanne Wojcik’s internal search reported roughly 330 responsive emails and at least four apparent gifts, including wine carafes, earrings, a sword, and a mountain retreat (Children’s Health Defense 2017).

Title 18 U.S.C. § 208, the criminal conflict-of-interest provision, applies to financial interests. A romantic relationship with a grantee’s principal investigator, standing alone, does not trigger it. That gap is why this ended as a personnel matter rather than a referral.

What the money bought

Three papers came out of this award, and all three are load-bearing have Thorsen name on them and the contracting officer’s name is one two of those papers.

Madsen and colleagues published a population-based study of measles, mumps, and rubella vaccination and autism in the New England Journal of Medicine in November 2002, following 537,303 Danish children. An ecological analysis of thimerosal and autism incidence was published in Pediatrics in 2003, again with Madsen as first author. Lauritsen and colleagues then validated the childhood autism diagnoses in the Danish Psychiatric Central Register, e-published in the Journal of Autism and Developmental Disorders in September 2009. Thorsen is an author on all three. Diana Schendel is an author on the first and third, while serving as the CDC lead on the award that funded them. Authorship like this requires that the author make substantial contributions to the manuscript. While serving as a CDC employee, Diana Schendel was clearly deeply involved in initial planning, award capture, research execution, and interpretation of the findings.

Government counsel relied on this body of work in the Omnibus Autism Proceeding, in which more than 5,000 claims for vaccine injury against the government were dismissed.

In the fall of 2009, the CDC and Danish team discovered that the ethics committee permissions required by Danish law had never been applied for. Erik Mortensen checked with the committee and confirmed it. Thorsen had been asked repeatedly to produce the approvals and had not, because they did not exist (Clay 2017, Exhibits 18 and 40). Two of the three papers were already in print.

On the November 30, 2009 grantee call, Coleen Boyle, Marshalyn Yeargin-Allsopp, and Diana Schendel discussed the problem with the Danish investigators. The call notes record the working conclusion that permission could probably be obtained for ongoing and future work but not for research already finalized. The immediate concern was ethical coverage for Schendel’s own forthcoming paper on autism and inflammatory markers. The solution they settled on was to extend a separate 2000 approval, granted to Rikke Maimburg for a study of obstetric factors and autism, to cover Schendel’s study, which the committee had never reviewed (Clay 2017, Exhibit 7).

No retraction followed, and no notification went to the journals or to the special masters who were then relying on the work. To this day, these papers have not been retracted.

Nothing in this record shows fabricated data, and the advocacy literature blurs that line. The 2003 protocol change that dropped the 2001 birth cohort is a documented design criticism, and Irva Hertz-Picciotto told a 2007 Institute of Medicine workshop that the Danish registry shifted from inpatient-only to inpatient-and-outpatient coverage across the study window, which makes the before-and-after comparison unsound. Those are methodological objections, and they were available to any reader in 2007.

That said, anyone who would outright steal money from the government through fraud is certainly capable of tweaking a study's design and results to obtain more government funds. Of that I have no doubt.

The documents add three facts that can not be denied. The principal investigator stole money from the award and has now pleaded guilty. The CDC scientist who created the sole-source solicitation was romantically involved with him, accepted gifts from him, and co-authored the output. The legally required human subjects approvals were never obtained, and when CDC learned that, it papered over the gap instead of reporting it.

CDC’s public position has been that Thorsen’s conduct was a fiscal matter that did not influence the science. Interesting perspective…

Who decided on a letter, and why that was the ceiling

Schendel’s reprimand is dated May 14, 2009. It was signed by Marshalyn Yeargin-Allsopp, MD, Chief of the Developmental Disabilities Branch (Yeargin-Allsopp 2009) at the CDC.

She was Schendel’s immediate supervisor. She was also, by the letter’s own account, the official to whom others had reported the conduct. The same person received the complaints, also set the penalty. No independent deciding official was inserted anywhere in the process.

The charge is improper use of official authority or credentials. The specification recites the romantic relationship with the principal investigator, the sharing of sensitive budgetary and programmatic information with the grantee, and advocacy that others witnessed as a loss of impartiality and objectivity. Schendel is directed to suspend all activities as Lead Scientific Collaborator, retaining only analyses already begun. The letter invokes the standard expected of a GS-15 senior scientist. The reprimand stays in her Official Personnel Folder for two years.

The indictment came down in April 2011, by which time the reprimand was scheduled to be removed from her file.

The choice of instrument did the rest of the work. Under HHS Instruction 752, a written reprimand is discipline. A suspension of more than fourteen days, a demotion, or a removal is an adverse action under 5 U.S.C. 7512, and an adverse action requires a written proposal, a thirty-day notice period, a reply opportunity, a separate deciding official, and an appeal right to the Merit Systems Protection Board. Every one of those steps generates a record that outlives the branch. A reprimand generates a letter that a branch chief signs alone and that expires on schedule.

The letter itself lists the alternatives it declined to use, running through one-day, five-day, fourteen-day, and thirty-day suspensions and removal. Yeargin-Allsopp had the full range in front of her and took the option at the bottom. A simple reprimand.

No document I have seen establishes whether she chose it because it kept the matter inside her branch. That question is what the adverse-action file would answer, and the adverse-action file has never been released. In this case, the lack of transparency and accountability at CDC is self-evident.

The bottom line is that Diana Schendel received no disciplinary action, other than a temporary letter placed in her internal personnel file.

No Forum

The reprimand was a remedy against the employee. A captured solicitation calls for a remedy against the award, and HHS has several.

Under 45 C.F.R. 75.371, an agency that finds noncompliance may withhold payments, disallow costs, suspend or terminate the award in whole or in part, or begin suspension and debarment proceedings. Nonprocurement debarment runs through 2 C.F.R. part 180 and reaches individuals as well as institutions. All of it was available to CDC from January 2009, when the Danish investigators told Diana Schendel and her colleagues that the money was gone.

CDC used none of it, so far as the public record shows. Thorsen kept joining monthly calls with CDC staff through December 2009. The award continued. Whether he was ever excluded from federal awards can be checked against the SAM.gov exclusions record, and I have not confirmed an entry there.

When a competition is tainted, the fix is to cancel the solicitation and run it again. That remedy never happened.

Under 4 C.F.R. 21.5(m), the Government Accountability Office will not review a protest of the award of a grant, a cooperative agreement, or any other non-procurement instrument, because none of them is a contract under the Competition in Contracting Act. One narrow exception survives, for a protest that an agency is using an assistance instrument where a procurement contract is legally required, and it must be filed before award within ten days of learning which instrument the agency picked. The Court of Federal Claims is bounded the same way, to protests connected to a procurement.

Before standing comes the prior problem: no rule was broken. The Competition in Contracting Act requires full and open competition for contracts, treats sole source as one of seven enumerated exceptions, and makes each use of an exception depend on a written justification approved by the person above the requester. CICA does not apply to assistance. No statute imposes a competition requirement on grants or cooperative agreements, and GAO’s own appropriations law treatise notes that agencies hold broad discretion over grant competitions and that assistance instruments can be used to escape the rules governing contracts. In a contract, Schendel would have needed a signature above her own on a document any interested firm could attack. In assistance, her preference was the standard.

The gap runs to recovery as well. A compromised acquisition can be unwound, and the money pulled back. FAR subpart 3.7 implements 18 U.S.C. 218, delegated to agency heads by Executive Order 12448, and lets an agency head declare a contract void, rescind it, and recover the amounts expended and property transferred, less the fair value of any tangible benefit the agency kept. A demand under that authority is not a claim under the Contract Disputes Act, so the contractor does not get the ordinary disputes route. The Procurement Integrity Act adds a second track at 41 U.S.C. 2105, requiring an agency to consider rescission and permitting cancellation of the procurement and recovery of amounts paid. FAR 3.700 preserves the government’s common law right of avoidance on top of both, the line running through United States v. Mississippi Valley Generating Co., where a contract tainted by a conflict of interest was held unenforceable.

Three separate things keep this award from applying. The FAR governs acquisitions, and this was assistance. The conviction trigger in subpart 3.7 points to Chapter 11 of Title 18, sections 201 through 224, where the criminal conflict-of-interest provision sits, and that provision was unavailable because the conflict was romantic rather than financial. Thorsen’s conviction is wire fraud under 18 U.S.C. 1343, which is nowhere near Chapter 11.

The assistance-side tools answer a different question. Cost disallowance and the remedies at 45 C.F.R. 75.371 reach unallowable costs. Nothing in the Uniform Guidance permits HHS to recover eleven million dollars on the ground that a conflicted official wrote the solicitation. The restitution DOJ is seeking covers the stolen million.

One question remains open, and nobody appears to have asked it. The statute behind FAR subpart 3.7 reaches contracts, “and other transactions,” and the FAR implements it only for procurement contracts. Whether that phrase extends to a cooperative agreement is unsettled. If it does, 18 U.S.C. 218 is the only authority capable of voiding this award, and it would be available to the Secretary of Health and Human Services rather than to a court.

An entity could still have had standing. Three profiles fit, and none of them is the disappointed applicant.

The first is the instrument challenger, which is the one exception GAO preserved. The Federal Grant and Cooperative Agreement Act assigns contracts to acquisitions for the government’s own benefit and assistance to transfers serving a public purpose (31 U.S.C. 6303 and 6305). A protester arguing that CDC was buying studies rather than assisting Danish science would be heard. That entity has to be an actual or prospective offeror able to perform the work, has to read the announcement while it is open, and has to file within ten days of learning which instrument CDC chose. On these facts the merits were arguable, since CDC specified the studies it wanted, co-designed them, sat on the analysis calls, co-authored the papers, and used the output to defend its own vaccine program.

The second is an APA plaintiff. A non-awardee has no contract with the government, so the Tucker Act channeling that now governs grant-termination suits does not push it into the Court of Federal Claims. It would sue in district court under 5 U.S.C. 702, and because no statute required competition, its claim has to be that CDC departed from procedures CDC had bound itself to follow. That entity must have applied, or be able to show it stood ready to apply, and was excluded by eligibility terms written to exclude it.

The third is a relator. The False Claims Act reaches grant funds and requires no competitive injury at all. The qualifying characteristic is original-source knowledge of the false invoices. The Aarhus finance staff who found the forged CDC letters in the winter of 2008 and 2009 fit that description precisely, and for several years they were the only people who did.

Every one of those profiles requires knowing something before the award closes, or knowing something held inside a circle of a few people. The sole-source structure governs both. Restricting eligibility to one Danish government body meant no American research organization had reason to read the announcement, let alone to challenge the instrument within ten days. Keeping the conflict inside CDC email meant it surfaced in 2009, when CDC searched its own servers for another reason entirely.

The entity with standing exists on paper. It is an American research organization that read program announcement 02006 in November 2001, understood that CDC was acquiring studies rather than funding Danish public health, and protested inside ten days. Nobody occupied that position, and the announcement was written so that nobody would.

Acquisition is policed in part by the parties who lose. Assistance produces no losers, so it is policed by the officials who made the award, in whatever forum they convene and at whatever severity they select. Here the forum was a branch chief’s office and the severity was a letter that expired in twenty-four months.

The Forgery Nobody Chased

CDC did not detect this theft. Aarhus University did, in the winter of 2008 and 2009, when it went looking for research money that had not arrived.

Aarhus and DASTI found three letters purporting to acknowledge CDC grants. Falsified signatures appeared on documents totaling close to two million dollars. DASTI filed a police report in May 2009 that named no person. CDC investigated and, per Aarhus, a suspicion arose that the documents were forgeries (Aarhus University 2010).

The forged CDC funding letters entered the criminal case as a predicate for wire fraud. They were never charged as forgery of a federal document under 18 U.S.C. 495, and no separate false-statement count appears. Thorsen’s plea on September 1, 2026 was in regard to the wire fraud, not the forgeries.

Search the public record for what CDC did about the fact that letters bearing its grant management officers’ forged signatures circulated in Denmark for at least four years and moved millions of dollars. I have found no HHS Office of Inspector General report on the administration of the Denmark cooperative agreement, and no published corrective action governing how a foreign recipient verifies an award letter. Nothing indicates that the officials whose signatures were forged were ever notified, or that the Procurement and Grants Office asked how counterfeit award correspondence moved money undetected for four years.

An agency that learns its own award letters can be counterfeited has a control problem independent of whoever counterfeited them. Aarhus responded to that problem in five weeks by cutting Thorsen loose and going to the police. The public record shows no equivalent response at CDC.

From Funder to Faculty

Schendel has been a professor of psychiatric epidemiology at the National Centre for Register-based Research at Aarhus University since 2013. In July 2020, she also joined the A.J. Drexel Autism Institute as professor and leader of its Modifiable Risk Factors Research Program. She holds both positions today, continues to publish actively, and continued to co-author papers with Thorsen after his federal indictment.

Consider the sequence.

As a CDC scientist, Schendel helped shape the program announcement that directed millions of dollars into the Danish research program. She argued for sole-source funding. She participated in the scientific administration of the resulting cooperative agreement. She was in an undisclosed romantic relationship with its principal investigator, accepted gifts from him, and co-authored research produced through the collaboration. CDC eventually reprimanded her over conduct arising from that relationship and her advocacy on behalf of the Danish program.

Then she left CDC and became a professor at Aarhus University, one of the institutions at the center of the research collaboration she had helped fund from inside the federal government.

And federal post-employment law did not stop her.

Nothing applicable to her position required a cooling-off period before taking the Aarhus job. Nothing required public disclosure of the move. Nothing required CDC to make a public determination that the employment was permissible. The federal ethics system that was supposed to police conflicts while she was at CDC imposed essentially no barrier when she crossed from the agency that supplied the money to the university that had received it.

The revolving door did not malfunction. For this kind of federal assistance relationship, there was barely a door at all.

The Door Congress Left Open

The cooling-off period is the restriction most people assume covers this. It is found at 18 U.S.C. § 207(c), and it bars a former official from communicating back to her old agency for one year.

It did not apply to Diana Schendel because it applies only to senior employees. A senior employee is one whose basic rate of pay equals or exceeds 86.5 percent of Executive Schedule Level II (5 C.F.R. § 2641.104). In 2012, that threshold was roughly $155,000. GS-15 step 10 basic pay, excluding locality, was $129,517. A GS-15 program scientist came nowhere near the threshold. She could leave federal service with no one-year cooling-off obligation.

For the officials the restriction does reach, Congress wrote an exemption that swallows much of the rule in the world of biomedical research. Under 18 U.S.C. § 207(j)(2), the restrictions in subsections (c), (d), and (e) do not apply to acts performed as an employee of, and on behalf of, an accredited degree-granting institution of higher education, a hospital, or a medical research organization. The permanent bar at § 207(a)(1), covering particular matters the official handled personally and substantially, survives the exception. The NIH ethics office states the consequence plainly on its own website: a former NIH senior employee may represent a new employer back to NIH when that employer is one of those three (NIH Ethics Program 2021). A former Senior Executive Service official can leave NIH on Friday and telephone NIH on Monday on behalf of a university, lawfully.

The exemption appears to reach only domestic institutions. Section 101 of the Higher Education Act defines an institution of higher education as an educational institution in a State. Foreign schools are addressed separately through section 102, under which the Secretary of Education approves institutions outside the United States by finding them comparable to a section 101 institution (Public Law 108-98). Section 207(j)(2) incorporates section 101, not section 102, so a Danish university does not appear to fall within the exemption.

Nothing turns on that distinction here because the cooling-off period never reached Schendel as a GS-15 in the first place. But it reveals something important about how these controls work. Grade and funding mechanism determine whether the safeguards apply. The seriousness of the underlying conflict does not.

The third gap becomes obvious when the two kinds of federal awards are placed side by side. Under 41 U.S.C. § 2103 and FAR 3.104-5, a contracting official who is merely contacted about employment by a competing offeror must report the contact in writing and disqualify herself from the procurement. The duty is affirmative, documented, and triggered before any employment agreement exists.

There is no equivalent requirement for assistance awards.

A program officer negotiating employment with a grantee is instead governed by the seeking-employment rules at 5 C.F.R. §§ 2635.601–2635.606. Those rules depend largely on the employee recognizing the conflict and acting on it. Unlike the procurement regime, they do not create the same automatic written record of an employment contact.

That difference matters years later. In procurement, there may be a document showing that the contact occurred, when it occurred, and what the official did about it. In an assistance case like this, there may be nothing comparable to request. A FOIA request for a post-employment ethics determination can therefore return nothing responsive, not necessarily because records disappeared, but because the system never required such a record to be created.

And then there is the statutory carve-out itself.

Congress expressly exempted universities, hospitals, and medical research organizations from important parts of the cooling-off regime. Those are also among the institutions receiving enormous amounts of federal research money. The result is a revolving-door rule that is weakest in one of the places where the federal government spends most heavily and where scientific officials can exercise enormous influence without ever signing a check.

An FDA reviewer leaving for a drug company looks like a revolving door and attracts headlines. A CDC or NIH scientist leaving for the university whose federally funded work she helped shape can look like an ordinary academic promotion.

Sometimes the difference is not the conflict. It is the law governing the money.

This Is the Ordinary Case, Not the Exotic One

Strip the autism politics out of this case, and what remains is a template that can repeat across HHS.

A program scientist helps write the announcement for the work she wants done. Substantial federal involvement is built into the cooperative agreement, so extraordinary closeness to the recipient can look like the system working exactly as designed rather than a warning sign. The person most committed to the award can help make the case for sole sourcing. Objective review can run a single applicant through in an afternoon. The business authority sits in another office, staffed by people who do not evaluate the underlying science and cannot independently judge its merits. The scientists understand the science but do not control the money. The grants officers control the money but do not control the science. Nobody owns the whole conflict.

And when something goes wrong, the gaps become even more important.

Section 208 may not reach the conflict because the interest is personal rather than financial. The agency can handle the conduct as an internal personnel matter. A reprimand goes into a file. The reprimand expires. The employee leaves for the recipient institution. The award continues.

None of this requires a conspiracy to conceal an autism epidemic. It requires a federal assistance system capable of allowing a sufficiently motivated scientist to help select the recipient, shape the solicitation, influence the review, remain deeply involved in the resulting research, and eventually follow the money out the door. If the conflict falls between the statutory categories, the institutional response may amount to little more than a letter signed by her own supervisor and kept in her personnel file for twenty-four months.

A conspiracy would require people to organize it.

This is more troubling. This is what the machinery can produce when everyone simply follows the machinery.

What to Request Under FOIA

The criminal case against Poul Thorsen is finally moving toward its conclusion. The institutional case is not.

Anyone continuing this work should request the following records. CDC records should go to the CDC/ATSDR FOIA Office. Personnel and ethics records should go to the HHS Office of the Assistant Secretary for Administration.

1. Notice of Award and all amendments for Cooperative Agreement U10 DD000230 and its predecessor under Program Announcement 02006 or A02006, fiscal years 2001 through 2010. The award face should identify the Grants Management Officer and designated Project Officer. This should settle the authority question permanently.

2. The sole-source justification memoranda for both the 2001 and 2007 competitions, including OMB clearance correspondence.

3. The complete objective review files for the December 7, 2001 panel and the 2007 Special Emphasis Panel, including reviewer nominations and all conflict-of-interest certifications.

4. Diana Schendel’s OGE Form 450 filings for fiscal years 2001 through 2012, together with any CDC ethics-office record concerning gifts, disqualification, or authorization under 5 C.F.R. § 2635.502.

5. The complete disciplinary file supporting the May 14, 2009 reprimand, including any proposal memorandum, the deciding official’s rationale, and any grievance filed under the administrative grievance procedure.

6. All CDC Procurement and Grants Office records concerning the suspected forged award letters, including referrals to HHS-OIG, notifications to Randolph B. Williams or any other official whose signature was forged, and any resulting changes to award-letter verification procedures for foreign recipients.

7. HHS-OIG investigative and management records concerning administration of the Denmark cooperative agreement, separate from the criminal investigative file. Expect a 7(A) exemption claim while the criminal proceeding remains pending and refile after Thorsen’s sentencing on December 1, 2026.

8. Any post-employment ethics determination or advisory opinion concerning Schendel’s transition from CDC to Aarhus University.

9. All CDC correspondence from 2009 and 2010 concerning ethics-committee permissions for the Danish autism registry studies, including records of the November 10 and November 30, 2009 grantee calls.

10. Any HHS Office of the General Counsel opinion, memorandum, or referral addressing whether 18 U.S.C. § 218 or FAR Subpart 3.7 could reach a cooperative agreement, together with any consideration given to voiding, terminating, or recovering funds from the Denmark award.

11. Any suspension or debarment referral, notice, or determination concerning Poul Thorsen, the Danish Agency for Science, Technology and Innovation, Aarhus University, or Odense University Hospital, together with any enforcement action considered or taken on the award under 45 C.F.R. § 75.371.

12. Any CDC or HHS communication with the New England Journal of Medicine, Pediatrics, or the Journal of Autism and Developmental Disorders concerning the ethics-approval status of the Danish papers, and any communication with the Department of Justice or the Office of Special Masters concerning the same.

Two important documents do not require FOIA. The plea agreement and any statement of facts filed in United States v. Thorsen, 1:11-cr-00194, Northern District of Georgia, are available through PACER. A statement of facts may identify additional participants, describe what federal officials knew, or state explicitly that no agency employee participated in the theft. Any of those answers would matter.

But the unanswered questions are now larger than Poul Thorsen.

We know he stole the money. He has pleaded guilty to doing it. What remains unanswered is how a federal research program of this importance was administered before, during, and after the theft.

Who knew about the relationship between Schendel and Thorsen, and when? Who approved the sole-source arrangement? What conflicts were disclosed to the reviewers? What did CDC do when it learned that the required Danish ethics approvals had never been obtained? Who decided that retroactive coverage was an acceptable solution? Why were the journals not notified? Why were the vaccine-court special masters not notified? Why was Thorsen still participating in CDC calls after the problems were known? Why was the institutional response to Schendel apparently limited to a reprimand that disappeared from her personnel file two years later? And what, if anything, did HHS-OIG investigate about the administration of the award itself?

These are questions about whether the federal government followed its own rules while producing science that it subsequently held up as authoritative.

For years, the Thorsen scandal has been reduced to two competing stories. One side points to a convicted thief and concludes that the Danish autism research must therefore be fraudulent. The other points out, that stealing grant money does not prove scientific fraud and treats the matter as closed.

Both stories miss the institutional question.

The theft does not prove the science was fabricated. But the absence of fabricated data does not erase the conflicts of interest, the sole-source funding, the missing ethics approvals, the retroactive effort to repair them, or the remarkable lack of institutional consequences.

Furthermore, the research produced by Paul Thorsen needs to be scrutinized. Someone in charge of a huge program that produced the research used to determine that 5,000 petitions for vaccine injury be denied defrauded the government of a million dollars. His financial malfeasance makes the research produced suspect.

Poul Thorsen has finally been made to answer for the money he stole.

CDC has yet to answer for what happened on its watch.

References

Aarhus University. 2010. “Statement Regarding Dr. Poul Thorsen’s Involvement in Aarhus University Projects.” Ref. AU-2010-179-012, January 22. Signed Jørgen Jørgensen, Managing Director. Reproduced as Exhibit 3 in Clay 2017.

Children’s Health Defense. 2017. “The ‘LOVE’ Emails Between CDC’s Diana Schendel and Autism Researcher Poul Thorsen.” September 29. https://childrenshealthdefense.org/news/love-emails-cdcs-diana-schendel-autism-researcher-poul-thorsen/.

Children’s Health Defense. 2026. “Former CDC Scientist Whose Studies Were Used to ‘Debunk’ Vaccine-Autism Link Will Plead Guilty.” The Defender, August. https://childrenshealthdefense.org/defender/poul-thorsen-former-cdc-scientist-studies-debunk-vaccine-autism-link-plead-guilty/.

Clay, Beth. 2017. Poul Thorsen: Fugitive Researcher, Update August 2017. World Mercury Project. https://childrenshealthdefense.org/wp-content/uploads/world-mercury-project-report-on-poul-thorsen-9-7-17.pdf.

Conte, Louis. 2026. “Why Poul Thorsen Matters.” The MAHA Report, August 31. https://www.themahareport.com/p/why-poul-thorsen-matters.

Competition in Contracting Act of 1984. Full and open competition, 41 U.S.C. 3301, and exceptions, 41 U.S.C. 3304.

False Claims Act. 31 U.S.C. 3729 through 3733.

Federal Acquisition Regulation. Disqualification. 48 C.F.R. 3.104-5. Implementing 41 U.S.C. 2103.

Federal Acquisition Regulation. Voiding and Rescinding Contracts. 48 C.F.R. subpart 3.7. Implementing 18 U.S.C. 218 and Executive Order 12448 of November 4, 1983. Procurement Integrity Act remedies, 41 U.S.C. 2105.

Administrative Procedure Act. Right of review, 5 U.S.C. 702.

Government Accountability Office. Bid Protest Regulations. 4 C.F.R. 21.5(m).

Federal Grant and Cooperative Agreement Act of 1977. 31 U.S.C. 6301 through 6308.

Hertz-Picciotto, Irva. 2007. Remarks at Autism and the Environment: Challenges and Opportunities for Research, Institute of Medicine workshop, April 18 and 19.

Government Accountability Office. 2006. Principles of Federal Appropriations Law, 3rd ed., vol. II, chapter 10, Federal Assistance: Grants and Cooperative Agreements. GAO-06-382SP.

Health and Human Services. Uniform Administrative Requirements, Cost Principles, and Audit Requirements for HHS Awards. Remedies for noncompliance, 45 C.F.R. 75.371. Nonprocurement debarment and suspension, 2 C.F.R. part 180.

Higher Education Act of 1965. Institution of higher education. 20 U.S.C. 1001, section 101, and 20 U.S.C. 1002, section 102.

Lauritsen, Marlene B., Mette Jorgensen, Kreesten M. Madsen, Sanne Lemcke, Sonja Toft, Jakob Grove, Diana E. Schendel, and Poul Thorsen. 2010. “Validity of Childhood Autism in the Danish Psychiatric Central Register: Findings from a Cohort Sample Born 1990-1999.” Journal of Autism and Developmental Disorders 40 (2): 139-48. E-published September 1, 2009.

Madsen, Kreesten M., Anders Hviid, Mogens Vestergaard, Diana Schendel, Jan Wohlfahrt, Poul Thorsen, Jorn Olsen, and Mads Melbye. 2002. “A Population-Based Study of Measles, Mumps, and Rubella Vaccination and Autism.” New England Journal of Medicine 347 (19): 1477-82.

Madsen, Kreesten M., Marlene B. Lauritsen, Carsten B. Pedersen, Poul Thorsen, Anne-Marie Plesner, Peter H. Andersen, and Preben B. Mortensen. 2003. “Thimerosal and the Occurrence of Autism: Negative Ecological Evidence from Danish Population-Based Data.” Pediatrics 112 (3): 604-6.

NIH Ethics Program. 2021. “Additional Post-Employment Restrictions and Exception for Government Senior Employees.” National Institutes of Health. https://ethics.od.nih.gov/PostEmp-Senior.

Office of Government Ethics. Post-Employment Conflict of Interest Restrictions. 5 C.F.R. Part 2641, sections 2641.104 and 2641.204.

Office of Government Ethics. Standards of Ethical Conduct for Employees of the Executive Branch. 5 C.F.R. Part 2635, subparts B and E, and sections 2635.601 through 2635.606.

Public Law 108-98. 2003. Foreign school eligibility, amending section 102(a)(2)(A) of the Higher Education Act of 1965. 117 Stat. 1174, October 10.

Restrictions on Former Officers, Employees, and Elected Officials of the Executive and Legislative Branches. 18 U.S.C. 207, subsections (a)(1), (c), and (j)(2).

United States v. Mississippi Valley Generating Co. 1961. 364 U.S. 520.

United States v. Poul Thorsen. 2011. Criminal Indictment No. 1:11-CR-194, U.S. District Court for the Northern District of Georgia, filed April 13. Reproduced as Exhibit 8 in Clay 2017.

U.S. Attorney’s Office, Northern District of Georgia. 2011. “Autism Researcher Indicted for Stealing Grant Money.” April 13.

U.S. Attorney’s Office, Northern District of Georgia. 2026a. “Autism Researcher Extradited from Germany to Face Federal Charges of Stealing CDC Grant Money.” May 12. https://www.justice.gov/usao-ndga/pr/autism-researcher-extradited-germany-face-federal-charges-stealing-cdc-grant-money.

U.S. Attorney’s Office, Northern District of Georgia. 2026b. “Danish Autism Researcher Extradited After 14 Years on the Run Pleads Guilty to Stealing CDC Grant Money.” September 1. https://www.justice.gov/usao-ndga/pr/danish-autism-researcher-extradited-after-14-years-run-pleads-guilty-stealing-cdc.

Yeargin-Allsopp, Marshalyn. 2009. “Reprimand for Improper Conduct.” Memorandum to Diana Schendel, Department of Health and Human Services, Centers for Disease Control and Prevention, May 14. Reproduced as Exhibit 43 in Clay 2017. https://childrenshealthdefense.org/wp-content/uploads/poul-thorsen-criminal-conduct-exhibits-43-schendel-reprimand-for-improper-conduct.pdf.

Audio Version:

Executive Summary

Alan Cassels has written an important essay for the Brownstone Institute about something modern medicine rarely stops to question: our growing obsession with numbers.

Blood pressure. LDL cholesterol. Hemoglobin A1C. BMI.

We are increasingly taught that health can be reduced to whether these four numbers fall on the correct side of an officially sanctioned line. Cross that line and, almost overnight, a healthy person can acquire a diagnosis, a prescription, regular laboratory testing, and perhaps a lifetime relationship with the medical system.

Cassels calls this “numericide.” It is a provocative term, but the underlying argument deserves serious consideration.

His central point is not that blood pressure, blood glucose, cholesterol, or body composition are meaningless. They clearly are not. These measurements can provide useful information, particularly in people with established disease or substantial cardiovascular or metabolic risk. His argument is that medicine has increasingly confused surrogate markers with actual health outcomes.

That distinction matters.

The question should not simply be, Can we lower the number? Modern pharmaceuticals are often very good at doing that. The more important questions are: Does lowering that number meaningfully reduce heart attacks, strokes, disability, or death? By how much? In which patients? And what are the harms and tradeoffs required to achieve it?

Cassels walks through four of the most familiar medical targets.

With blood pressure, he challenges the idea that 120/80 should function as a universal dividing line between health and disease. Blood pressure varies considerably with age, circumstance, measurement technique, stress, and underlying cardiovascular risk. Aggressively lowering it may benefit selected high-risk patients, but treatment also carries risks, particularly in older people, including hypotension, falls, kidney injury, and adverse drug effects.

With LDL cholesterol, he makes a similarly important distinction between primary and secondary prevention. Someone who has already suffered a myocardial infarction is not the same patient as an otherwise healthy person whose LDL happens to exceed 100 mg/dL. Yet public messaging often collapses these very different risk categories into a single mantra: lower is better. Cassels points to the relatively small absolute mortality benefit found in primary-prevention statin trials and argues that patients deserve to hear those absolute numbers rather than simply being told that their cholesterol is “too high.”

His discussion of HbA1C and “prediabetes” raises another problem: the medicalization of risk itself. There is an enormous difference between uncontrolled diabetes capable of causing blindness, kidney disease, neuropathy, and vascular injury and a metabolically healthy person whose laboratory value has drifted slightly across an administrative threshold. Cassels is right to remind readers that aggressive glucose lowering has not invariably produced better outcomes and, in some circumstances, can produce real harm through hypoglycemia and medication effects.

Finally, there is BMI, perhaps the crudest of the four measurements. BMI cannot distinguish muscle from fat, tells us little about fat distribution, and can classify two people with profoundly different metabolic health as essentially identical. Cassels points to large observational studies showing that the relationship between BMI and mortality is considerably more complicated than the familiar “25 equals overweight” threshold suggests.

There is much here that I agree with.

Medicine has become remarkably good at turning healthy people into patients. Change a guideline, move a laboratory threshold, redefine “normal,” and millions of people can acquire a disease or pre-disease without anything about their physiology having changed overnight. And when the committees establishing those thresholds have financial relationships with industries that benefit from expanding treatment populations, skepticism is not anti-science. It is precisely what science requires.

But I would add some caution to Cassels’ argument.

These numbers are not meaningless, nor are the risks associated with hypertension, diabetes, visceral obesity, or dyslipidemia imaginary. There are patients for whom aggressive management of these risk factors is clearly appropriate. Cardiovascular risk is cumulative, and an individual measurement can be useful when interpreted alongside age, family history, smoking, insulin resistance, inflammation, kidney function, coronary disease, body composition, fitness, and many other factors.

The mistake is not measuring the numbers.

The mistake is worshipping them.

A laboratory value is information. It is not a diagnosis in isolation, and it certainly is not a patient. Medicine works best when physicians use these measurements as pieces of a much larger biological picture rather than treating every person who crosses an arbitrary threshold as another candidate for pharmaceutical intervention.

That is ultimately the most valuable lesson in Cassels’ essay.

We should know our numbers. But we should also know what they actually mean, what our absolute risk is, how much a proposed intervention is likely to change that risk, and what that intervention may cost us in adverse effects and quality of life.

Those are considerably more important numbers.

Read the Brownstone article for yourselves, draw your own conclusions and let’s discuss in the comments!

Four Forms of Numericide

By: Alan Cassels

A recent article in the New York Times on the increasing rates of heart disease in young women (Why Are More Young Women Developing Heart Disease?) was amusingly frivolous and about as fact-free as an article about health advice can be.

The key thrust of the article is that any woman who wants to preserve and improve her health should become informed of her “numbers.” Which is to say, she should know her blood pressure, cholesterol, BMI, and hemoglobin A1C (a measure of average blood glucose levels) which are claimed to be “important for assessing risk.”

The Times quotes Dr Priya Freaney, director of the Women’s Heart Care program at Northwestern Medicine: “Knowing your numbers is your most powerful tool,” [she said] “And it’s never too early to start.”

What numbers are we talking about? Well, apparently doctors recommend “all adults strive for these levels:”

• BP (Blood Pressure) under 120 mm Hg / 80 mm Hg

• LDL (Low-density lipoprotein) under 100 mg/dL

• HbA1C (Hemoglobin A1C ) of 5.6 or lower

• BMI (Body-mass index) under 25

Now the four Horsemen of the Apocalypse, blood pressure, blood sugar, cholesterol, and BMI, threatening to strike you down in your healthy middle age unless you submit to the testing, treating (ie: medications and lifestyle changes), and a retesting paradigm which can lead many to become obsessed by the numbers.

By the way, these four measures have one thing in common: They are typically treated with a prescribing pen that has been carefully guided in your doctor’s expert hands by the helpful people in the pharmaceutical industry. Just saying…

Let’s approach these four with a bit of basic epidemiology, examining the wisdom of telling people of “average risk” to aim for these targets. Obviously, people with pre-existing conditions or those with inherently much higher or multiple risks (such as levels that are in the red flashing light category) involve a different calculus.

But for the sake of evaluating the drive to push people toward these health targets, what do the highest quality studies ultimately say about the value in chasing these numeric targets in healthy people?

Numericide Number One: Blood Pressure at 120/80

I will always remember that old quip from an aging physician: “High blood pressure is better than no blood pressure at all.” But the question one should ask is this: If you lower your blood pressure does it lead to a reduction in your chance of a cardiovascular event or dying?

First off, despite what many guidelines and health care providers say about knowing your numbers, what they should have said was know your cardiovascular risk numbers (your percentage chance of having a heart attack or stroke over the next 10-30 years) not your numbers of cholesterol, blood pressure, or glucose numbers. Those are what we call ‘surrogate markers’ and the only reason to lower these numbers is to possibly reduce your risk of heart attacks and stroke. It’s not that these surrogate numbers are useless — rather they are the wrong numbers to worry about. Their only use is to help us give a more accurate ballpark assessment of your cardiovascular risk.

For starters, the target of 120 is not an evidence-based but an “expert-recommended” target, and in case you wondered, having a blood pressure of 130/80 or 140/90 is not the death sentence that your doctor might warn you about. Your blood pressure often jumps all over the map and how and when you measure it can significantly impact that number. It climbs as we get older, naturally. Like grey hair and wrinkles. Chasing this number is one of the biggest uses of a doctor’s time and causes considerable (and frequently unnecessary) worry in patients.

Me? I wish our doctors would have spent time learning more about evidence appraisal and how to properly do shared decision-making with their patients because the decision around the treatment of surrogate numbers is very dependent on a person’s values and preferences. Cholesterol or blood pressure “management” makes a ton of money for the pharmaceutical industry and seems like a good use of our doctors’ time, but often causes a lot of harm in a large part of the otherwise healthy patient population.

The magical “120” comes from the SPRINT trial, a particularly odious piece of work that tortured the data until it showed some benefit of a BP <120 comparted to divergent targets such as <140 in a high-risk population. What most physicians aren’t going to learn is that any apparent benefit evaporated once intensive treatment stopped.

The SPRINT team was behind the 2017 guidelines recommending the 120/80 BP target, but you need to know these guidelines have been discredited for good reason. For instance, the process to produce them was riven with documented conflicts of interest. Again, Big Pharma’s interests are more than well-represented on the guideline production team and their participation automatically broadens the category of people who should take their drug. No — surely that can’t be true? It is true, and to quote a refrain from the movie Airplane, stop calling me Shirley.

The AAFP, American Association of Family Practice— the largest primary care organization in the US — refused to endorse that target. Why? Because, in their words, there were “no significant benefit for all-cause mortality, cardiovascular mortality, myocardial infarction, or renal events.” Let me translate: People don’t live longer or better because you guilt-tripped and drug-equipped them into getting their blood pressure down to 120/80.

There’s also another reason and that has to do with how unsafe it can be to push people’s blood pressure to ridiculously low targets. Not only does this drive people crazy (because many of them have to take two or three or four medications to lower their BP numbers) but it can also lead to falls, fractures, kidney injury, and so on, especially in older adults.

One study in JAMA Internal Medicine found that taking antihypertensive medications was associated with an increased risk of serious fall injuries. SPRINT itself showed nearly double the rate of serious adverse events possibly or definitely related to the antihypertensive drugs.

Like any two-edged sword, drugs to lower your blood pressure can also cause injury to you in ways you haven’t yet foreseen and which could make your life miserable.

Numericide’s Number Two: Low-Density Lipoprotein, or LDL, under 100 mg/dL

This is one of the most aggressively, and irrationally pursued targets: LDL cholesterol. I went back into the data to ask this question: Is there any substantive evidence that an otherwise healthy person with an LDL under 100 mg/dL lives any longer than a person who has an LDL above 100mg/dl?

Basically, should healthy people (what we call ‘primary prevention’) who don’t have a history of a heart attack or heart disease be pushed to get their LDL from, say, 120 to 80 mg/dL? Will this make them live longer?

Sorry to be the bad news bear here, but in a nutshell: No.

The evidence is weak and the absolute benefit is very small. When I mean very small, it’s way smaller than most people expect.

The most comprehensive recent review was published in JAMA in 2022. It was huge, pooling 18 trials of statins for primary prevention including over 85,000 people. The rate of all-cause mortality was statistically significant yet the absolute difference between those who drove their cholesterol that low and those who didn’t was a meagre 0.35%.

The NNT (number needed to treat to prevent one death) was 286.

Another way to say this is that your chance of a normal healthy person dying if they take a statin to drive their LDL below the 100mg/dl target is almost one in 300. Others have strongly contested this number, saying that driving cholesterol down to those levels makes absolutely no difference to the length and quality of a person’s life.

What about older people? You’re in for a surprise here—because in the elderly the situation reverses. A study published in the BMJ in 2016 identified LDL-related studies in the elderly. It found that in a total of over 68,000 elderly people, there was an “inverse association between all-cause mortality and LDL.” Translation: older people with higher LDL levels lived longer! This makes it even more egregious that much of the population taking cholesterol-lowering drugs are over 65 and in the category where they actually need higher cholesterol to stay healthy.

And then there are the harms of cholesterol lowering drugs, like statins which we know cause muscle weakening, elevated liver enzymes, and diabetes, effects that are consistent across many studies.

Doctors know this from real life experience: most patients (up to 75%) stop taking their statins within 2 years — whether it’s from the adverse effects, the costs or the hassle.

Numericide’s Number Three: HbA1C (Hemoglobin A1C ) of 5.6 or Lower

The crazy low levels at which they tell people to drive their HbA1C down to make me bonkers. What is most nutty is that this nefarious hunt for “prediabetes” which is the granddaddy of all disease-mongering tactics is sold under the logic of ”Why wait until you have full blown diabetes when you can find the ‘pre’ signs early?”

To be clear, the 5.6% is not a treatment target — it’s the “prediabetes diagnostic threshold.” In the ever-reductive world of diabetes “normal” keeps shifting. “Normal” is now considered below 5.7%, prediabetes is 5.7–6.4%, and diabetes is diagnosed at ≥6.5%.

The actual treatment target for most adults with diabetes is <7.0%, a level that sadly labels tens of millions of healthy people as diseased. If you have an HbA1C slightly over 7 and your doctor is pushing a metformin prescription at you, it’s time to push back!

Here’s the core point: the evidence does not support that driving HbA1C to 6.5% or lower extends life. The best available data from two big studies — the ACCORD trial and a meta-analysis from the BMJ — show two very surprising things: either no mortality benefit or actual harm from intensive glucose control.

Just to repeat: You do not live longer if you drive your HbA1C into the basement.

Furthermore, there is the opposite thing that could harm you, hypoglycemia — which is often caused by aggressive drug treatment is, in itself, a potent killer.

Just remember that these thresholds are always changed in a uni-directional way, always made lower to expand the pool of patients who allegedly need to be on drugs. This is a process that takes place under the well-documented influence of pharma money, research, and experts, expanding the market for diabetes drugs at each step.

The major push to label people as “prediabetic” — a condition that, incidentally the WHO says shouldn’t be medicalized — is flat-out, brazen disease-mongering.

Having said that, extremely high blood sugar (HbA1C consistently above 9–10%) can cause real harm — kidney failure, blindness, neuropathy, infections. The point isn’t that glucose doesn’t matter at all; it’s that the marginal benefit of pushing from 7.0 to 6.5% or lower does not save lives, and the harm, both in terms of money, medical busy-ness of checking and rechecking your blood sugars, and the harms of the drugs are all real and quantifiable.

Eating a higher quality diet and getting sufficient exercise is the foundation for anyone worried about their blood sugars. Those who are told they have “pre-diabetes” should ignore that advice and refuse to be pasted with a sticky label that causes harm.

Numericide’s Number Four: BMI under 25

Body Mass Index (BMI) is a measure of body weight relative to height, often used to screen whether someone may be underweight, at a healthy weight, overweight, or obese. We need to screen because overweight people live shorter lives, right?

Not necessarily.

We are told that the aim for everyone, concerning this fourth Horseman of the Apocalypse, BMI, is 25. Not only is this absurdly low, but it is essentially meaningless as a measure of health on its own. Anyone with any good musculature whatsoever is going to have a BMI over 25. Why is that? Because muscle weighs more than fat.

It comes down to where the weight sits. A muscular person with BMI of 30 has a different risk profile than a guy with a BMI of 30 who is sedentary with a big beer gut — but the BMI treats them identically.

Where did this “25” come from?

The BMI 25 cutoff comes from a 1995 WHO technical report establishing four categories of body weight (underweight, normal, overweight, obese) and was loosely based on actuarial life insurance data. This threshold was not derived from outcome-based evidence and a BMI of 25 was not linked to any mortality inflection point.

What about someone who has, say, a BMI of 30. Should they get that lower?

The claim that someone at BMI of 30 has a meaningfully shorter life expectancy than someone at BMI of 25 is not supported by the best available observational evidence.

The largest and most influential meta-analysis on this question was published in JAMA in 2013 and it pooled 97 studies with over 2.88 million participants and more than 270,000 deaths. The findings were mind-blowing: Being overweight was associated with significantly lower all-cause mortality than normal weight. In fact, what is called “Grade 1 obesity” (BMI ~30) showed no significant increase in mortality. Only at BMI ≥35 did mortality clearly rise.

Let me translate: Being “kind of chubby” which would describe almost everyone my age, (mid-60s) is not the death sentence they say it is.

In other words, pushing people to have lower and lower BMIs has very little to do with their overall health.

As I have written before, the enthusiasm for high-priced weight loss drugs (see my past articles on GLP-1 drugs here, here, or here) is skewing medicine in a decidedly dark way. The financial machinery behind the current obesity treatment paradigm is putting whole generations of people on drugs that will deliver almost no meaningful impact on the length or quality of their lives. And the conflicted interests in obesity research and advocacy is documented to a degree that makes the cholesterol and blood pressure guideline conflicts look modest by comparison.

So there we have it, four targets — blood pressure, blood sugar, cholesterol, and BMI — the Four Horsemen of the Apocalypse, four ways to medicalize you, four ways to apply a disease-mongered label to you, and four avenues for you to start living under a dark cloud.

I say, reject all that.

Is “knowing your numbers” a powerful tool, as the experts suggest?

I guess it is, only if you want to fast-track yourself to being a patient. If your doctor is interested in your opinion and subscribes to the school known as “shared decision-making” then let your opinions be known.

At the end of the day don’t let anyone make you feel like you’re living under a dark cloud of impending doom because your numbers are slightly higher than what the experts recommend.

This was first published at the Brownstone Institute.

The author’s bio can be found here.

An excerpt from our book Homesteading for Health

Audio Version:

The Great Unlearning

At the turn of the twentieth century, scientific interest in soil biology was near its peak. There were volumes written about earthworms, their life cycle, how they transform soil into a living system, and how to harness their quiet labor. This was not fringe science. Charles Darwin himself published a major work on earthworms in 1881, helping to spark broader scientific study of soil organisms. One could walk into a library and find entire shelves devoted to the subject.

Much of that knowledge has since faded from common use, like the topsoil that once blew across our Great Plains. This is a kind of modern unknowing. Knowledge not entirely lost but set aside. Forgotten in practice, if not in record. Left to gather dust alongside the hardbound journals and books of another era. Some of those old manuscripts have been lost over time, as libraries deaccessioned collections in the march toward modernization.

Likewise, the study of zoology, once broad and immersive, has narrowed in our institutions. It once encompassed the full sweep of animal life, from vertebrates down to the smallest visible creatures. Today, much of that hands-on, whole-organism study has faded from view.

No longer do rows of specimens in jars line the shelves of the biology classroom. The old, tactile way of learning the natural world has largely been replaced. And those sweeping, popular works on the life cycle of the earth, once a staple of science education, have mostly slipped from everyday reading.

Now, bagged mulch and compost line the shelves of home improvement stores, marketed as cure-alls for soils long stripped of life in suburban backyards across America. Then come the supplements, algae extracts, fish emulsions, and more, sold as quick fixes for soil regeneration.

Meanwhile, the tried-and-true methods of the past are often wrapped in layers of jargon and complexity. Carbon sequestration, microbial amendments, branded systems, and organic pesticides, which in the end are still pesticides. Enough to turn away all but the most devoted.

The no-nonsense ways of the past have been crowded out by a mix of commercialism, input-driven agriculture, and a growing class of paid experts. Too often, what gets lost is the simple truth. Restoring soil is not a one-and-done proposition. It is something gardeners and farmers must tend year after year.

There is no cure-all. This is a long game. A commitment for life…

Share

Putting This Together

Soil is not an input. It is not something you buy in a bag, fix with a supplement, or correct with a single season of effort. It is a living system that responds to what you do, and just as importantly, to what you stop doing.

The old farmers also understood this, even if they did not use the language we use today. They planted to protect the ground. They rotated crops. They let roots do the work. They paid attention to water, slope, and season. And they stayed with it

We have, in many ways, unlearned those habits. We have replaced them with products, prescriptions, and the expectation of quick results. But the soil does not care about any of that. It responds to time, biology, and stewardship.

If there is a way forward, it is not new. It is a return. A return to keeping the ground covered. To feeding the soil, not just the plant. To using living roots as the primary tool. And to accepting that this is not a project with an end date. It is a practice. And it is stewardship of our (and our children’s) future. Season after season. Year after year.

By: JGM/RWM

To read more or listen to the audio book, go to:

Homesteading for Health: One Farm Family’s Guide to Food, Land, and Self-Sufficiency is available from:

• Amazon

• Barnes & Noble

• Target

• Walmart

• Books-A-Million

• BookPal

• Skyhorse Publishing

• Independent booksellers

Digital editions can be purchased through Apple Books, Barnes & Noble, Google Play, and Kobo.

The audiobook, personally narrated by Robert Malone, is available through Amazon and Audible

Photos and engravings from the past:

The “chilled plow” by James Oliver, South Bend, Ind., was designed to replace wooden plows with a metal mouldboard that would scour in any soil and last a lifetime. Oliver spent years in experimenting with chilled metal (drenched in cold water after casting to increase strength). Otherwise known as tempering.

Ploughing the Fields by Jean-Baptiste Huet, 18th century

‘The settlement/westward expansion of the American West”, 19th century engraving.

Robert and our team of Percherons, circa 2004

When the Grocery Store Is Gone, Who Vouches for Your Beef?

Direct-to-consumer food is coming. Nobody has built the thing that makes quality claims believable.

Joel Salatin is right that the grocery store is dying. He is wrong about what replaces it. And the problem nobody in the food technology business wants to discuss was solved three thousand years ago by a man standing in a slaughterhouse.

Joel Salatin gave the dinner address at the Brownstone gathering at Polyface last week. Jill and I had spoken earlier that day, out on the farm. His argument was that the grocery store is on its way out, particularly in the cities, and that within a generation urban households will have much of their food delivered directly to their homes. If that happens, he believes it could create a significant new market for homesteaders and small farmers who can ship their products through parcel carriers.

He may be right about the larger trend. But the system he described is unlikely to work as proposed. The reasons why are worth examining, because they point to a much larger problem with how we produce, regulate, transport and sell food in America.

Why Joel Is Right

USDA’s own accounting puts the farm share of the retail food dollar in the mid-teens (USDA Economic Research Service, Food Dollar Series). Everything above that fifteen cents belongs to somebody who did not raise the animal. Processing takes a cut. Distribution takes a cut. Shelf space takes the largest cut of all, and the store extracts it for the service of being conveniently located.

Sell direct and you capture most of what those middlemen were taking. That arithmetic is the whole case for direct marketing, and it is a strong one. Every farmer who has sold a quarter beef to a neighbor already understands it better than any economist can explain it.

Joel’s stronger argument had nothing to do with margin. The cost of buying groceries in a city is no longer mostly the price on the shelf.

Count what a household in a large city now spends to fill a cart. Time sitting in traffic. Parking, where parking exists. A transit trip that is tolerable going and miserable coming back with four bags. A reasonable calculation about the walk to the car after dark. Stores that have cut their hours, locked the razor blades and the baby formula behind plastic, or closed the location outright and left a neighborhood with no full grocery inside of a mile. None of that appears in the price of ground beef. The shopper pays every bit of it anyway.

Carl Menger, the nineteenth-century Austrian economist and one of the founders of the Austrian School of economics, settled this in 1871. Value is not a property sitting inside a good. It is a judgment made by a particular person in a particular situation. The real price of a chuck roast includes the ninety minutes and the low-grade dread it costs to go get one. Raise those enough and delivery wins outright, without the delivered food ever becoming cheaper.

This is not a forecast. It has already happened. DoorDash, Uber Eats, Grubhub, Instacart, and Amazon’s Whole Foods operation are now an ordinary way for urban households to acquire food rather than an occasional convenience, and the shift has held every year since the lockdowns ended.

Joel drew the natural conclusion. If the grocery store is being cut out, the farm can step into the space it leaves. But look at who is already standing in that space. These companies are intermediaries. They take a commission from the seller, charge fees to the buyer and, in many cases, mark up individual grocery items above the price on the store shelf. The city shopper may bypass the grocery store, but not the middleman. Instead, a new middleman controls the transaction and, more importantly, the relationship with the customer, including the customer’s address, purchases and preferences. Amazon did not compete with Whole Foods. It bought it.

The demand half of Joel’s thesis is correct and visible from your kitchen table. The supply half, where a farm in Madison County fills that demand directly, is where the trouble starts.

The Two Hundred and Fifty Acres We Did Not Buy

Jill and I spent time this summer looking at two hundred and fifty acres nearby in Madison County and running the numbers on a fifty-head cow-calf operation. We decided against it.

The opportunity seemed obvious, which is why we took it seriously, and a great many of you are looking at the same opportunity. The national cattle herd sits near a seventy-five-year low. Retail beef has set records. Four companies, two of them foreign-owned, buy roughly eighty-five percent of the steers and heifers sold in this country, which in much of the country leaves a rancher with two or three bidders and very little leverage. A supply chain under that much strain is precisely where a small producer selling direct ought to be able to find room, and every conference in the country is telling him so.

Some of you may find that odd coming from people who write about homesteading. It is the reason you should keep reading. We are not selling you a course or a piece of software. We looked hard at the model being promoted to small producers right now, and we concluded that we are more useful thinking about where this is going than we would be running fifty pairs on Piedmont grass. What follows is what we found.

The Physics of the Last Mile

Congress can repeal a statute. A governor can sign an agreement. Nobody repeals thermodynamics, and that is why refrigeration is the hardest part of this and the part least discussed by the people promoting it.

Frozen beef holds its quality at zero degrees Fahrenheit and starts losing it above that. To keep a box there for two days in transit, you need insulation, and you need either gel packs, which are heavy and not very cold, or dry ice, which is cold and does not stay. Dry ice sublimates. The working rule is five to ten pounds gone every twenty-four hours, more in an under-insulated cooler and more again in August. You are buying refrigeration by the pound and watching it evaporate.

Everything about parcel pricing then works against you. Carriers bill by weight and by zone. Your cooler, your coolant, and your packaging are billable weight that nobody eats. Zone charges climb with distance, so Arlington is affordable from Madison County and Denver is a different business entirely. Two-day ground reaches a few hundred miles. Past that, you are buying air freight to move a frozen box, which is as expensive as it sounds. You also cannot ship into a weekend, because no frozen carton survives a Sunday in a sorting facility. That collapses your shipping week to a few days and hands you an inventory problem on top of a freight problem.

Do not take my figures or anyone else’s. Call your carrier, price a ten-pound insulated box to a city you want to sell into, add the dry ice and the cooler, and set that number beside your margin per pound. Then price the same box in July.

The size of that number is the least of it.

It is charged per box rather than per pound. A shipment costs roughly the same whether it holds eight pounds of beef or thirty, which means your cost per pound falls as the order grows. That is why every company in this business pushes large minimum orders and subscriptions, and why nobody ships a single ribeye. The economics do not permit the small transaction that direct marketing is supposed to make possible.

And the failure mode is total. A dented can of tomatoes is still worth something. A box that arrives at forty degrees is a full refund, a lost customer, and an animal you cannot sell twice. There is no partial credit anywhere in this, and the losses land on the producer.

Put those two together and you have the reason the aggregator keeps reappearing no matter how many times somebody announces its death. A fixed cost per shipment selects for whoever ships the most, every time, with no help from lobbyists and no conspiracy required. The grocery store was never only a building. It was the answer to the problem of moving cold food to a lot of people at once, and it was a good answer. Kill it and the same physics rebuilds something in its place, and that something will be owned by whoever can absorb the fixed cost. Remember this when you reach the fourth of the seven tests at the end of this essay, because it is the same test.

That does not leave the small producer with nothing. It leaves two ways around the problem rather than through it, and Joel found the first one himself long before he described the second one at dinner. Polyface built metro drop points and buying clubs, where one truck serves many households at a single stop on a fixed day. That consolidates the last mile, which is what the grocery store did, while the farm keeps the customer list, which is what the grocery store never allowed. It is regional rather than national, and it works right now.

The second route is to stop shipping cold things. Tallow, jerky, summer sausage, bone broth, cured products, and rendered fat can move at room temperature in a light box at ordinary parcel rates. They also happen to be among the highest-margin uses of the parts of the animal that are hardest to sell as cuts. A producer who cannot economically mail a chuck roast to Denver can mail tallow to Denver all day long. The direct-to-consumer opportunity that survives contact with the freight bill is largely a value-added opportunity, and almost nobody selling you software on this subject will mention it.

The Walls We Built Ourselves

Physics is not the only obstacle. The next ones were created by government and by the structure of the market, which means that, unlike thermodynamics, at least some of them can be changed.

Federal statute is the most obvious, and Joel has been fighting it for forty years. Under the Federal Meat Inspection Act, as amended by the Wholesome Meat Act of 1967, you may slaughter and process your own animals for your household, your nonpaying guests, and your employees. You may use a custom-exempt facility. The meat comes back marked Not for Sale, and that marking is not a suggestion. Commercial sale requires inspection.

The PRIME Act would loosen this, but its full name tells you how far: Processing Revival and Intrastate Meat Exemption. Read it carefully before you celebrate. Both the House bill and the Senate companion address distribution within the state. Neither one lets you ship a lamb to New Jersey. The pilot folded into the House-passed farm bill is also intrastate, and it requires the label to disclose that the meat was not federally inspected. Whatever your view of that disclosure, the nationwide shipping vision Joel described is not on the table in any bill currently moving.

The other wall does not look like a wall at all. It is ownership of the customer.

The grocery store traditionally controlled the point of sale. Direct marketing was supposed to put that relationship back in the farmer's hands. But if the farmer reaches the customer through a digital platform, the platform can simply take the grocery store’s place. Sell through Amazon and Amazon controls the search ranking, collects the referral fee, sets many of the terms of the transaction, and learns an enormous amount about the buyer. The farmer may recover some of the margin once taken by the grocer, only to surrender part of it to the platform.

That distinction matters. Selling directly to a consumer is not the same thing as owning the relationship with that consumer. If someone else controls how the customer finds you, what it costs you to reach him, and whether you can reach him again tomorrow, you have not eliminated the intermediary. You have changed intermediaries.

Richmond, Virginia

There is a federal program almost nobody in Virginia agriculture talks about, and it has been sitting there since the Bush administration.

Congress created the Cooperative Interstate Shipment program in the 2008 Farm Bill. The idea is remarkably simple. A qualifying small meat processor operating under a state inspection system can meet federal standards and then sell its products across state lines under the USDA mark. The federal government even pays sixty percent of eligible inspection costs.

Virginia already has most of what it needs. The Commonwealth has its own meat inspection program, its own inspectors, and state-inspected slaughter and processing plants. What it does not have is an agreement with USDA allowing those qualifying plants to participate.

Virginia has never signed one.

Georgia joined in July and became only the eleventh participating state. Twenty-nine states operate their own qualifying meat inspection programs, yet after fourteen years only eleven have joined CIS. There is not a single participating state in the mid-Atlantic.

Consider what that means from a farm in Madison County. The richest concentration of potential food customers in the country runs almost continuously from Richmond through Washington and Philadelphia to New York. A Virginia producer using a state-inspected plant can sell within Virginia but cannot simply follow those customers across the state line. Joel’s urban delivery future can reach Northern Virginia and then stop at the Potomac.

That boundary is not imposed by thermodynamics, and Congress has already provided a way around it. Virginia simply has not used it.

The contrast with the way government normally approaches agricultural development is striking. Governor Abigail Spanberger recently announced grants for animal processing, a creamery, a cannery, livestock marketing, and other agricultural infrastructure. Earlier this year, Virginia announced millions more for food-system infrastructure, with the stated goal of creating new markets and revenue for small and mid-sized producers.

There is nothing necessarily wrong with those investments. But we are spending public money helping producers increase capacity while leaving in place a government barrier that prevents some of those same producers from reaching customers outside Virginia.

And this is not really a Virginia story.

Twenty-nine states have inspection programs that potentially put them in a position to participate in CIS. Only eleven have done it. That means most eligible states have left a federally authorized route to interstate commerce largely unused while politicians across the country continue announcing grants, subsidies, loan programs, processing initiatives, and rural-development programs intended to help small farmers compete.

Before anyone reaches for a partisan explanation, look at the states that have joined and those that have not. Vermont and Maine participate. So do Montana and the Dakotas. Virginia has declined to participate under Kaine, McAuliffe, Northam, Youngkin, and now Spanberger. Five administrations, both parties, no action.

That may be the more important lesson. We keep trying to subsidize our way around barriers that government has the power to remove.

What Washington Promised on Friday

On Friday, August 28, the President posted on Truth Social that ranchers and farmers had always been a priority for him, that the four dominant meat processors amounted to a monopoly, and that he was authorizing legal documents to give farmers and ranchers the right to process their own food. He said it should move quickly.

The idea had surfaced two days earlier in an interview with Glenn Beck. Beck, who owns a ranch, told the President that the packers and the regulations were killing producers. The President called the proposal a very good idea and said he would look at it that day.

Agriculture Secretary Brooke Rollins followed Friday afternoon with promises of action beginning Monday. USDA would cut processing red tape, expand ranchers’ ability to sell across state lines, help small processors, address consolidation, improve labeling, and remove outdated guidance.

Representative Thomas Massie, who wrote the PRIME Act, responded within hours that the announcement was good, but that it needed to become law rather than an executive order. He is right, and he has been fighting for that law since 2015. The prohibition on commercial sale of uninspected meat is written into federal statute. A President can simplify regulations, rescind guidance, redirect money, and make better use of existing inspection programs. He cannot repeal an act of Congress with an executive order.

There was another problem with the timing. The week before the President’s post, the administration opened a three-hundred-thousand-metric-ton tariff-rate quota for imported lean beef trimmings, divided into three monthly installments beginning September 1. Cattlemen and a number of congressional Republicans objected because additional discounted imported beef puts pressure on the prices received by American producers.

So American ranchers heard two messages within days of each other. More foreign beef was coming into the country immediately, while Washington promised that relief for domestic producers was coming soon. The imported beef can clear customs this week. New slaughter and processing capacity takes years to build.

Then Monday came.

USDA called its package the Ranchers First Initiative. Strip away the acronyms and there are several potentially useful ideas in it. Ranchers will have a new way to insure some of the economic risk of retaining heifers to rebuild their herds. Some conservation money can now be used more flexibly to rebuild fences and water systems after disasters. USDA will support financing for independent processing capacity, try to direct more federal purchasing toward locally processed American beef, and create additional support for beginning and veteran farmers.

Some of that is worthwhile. The heifer provision addresses a real problem. A rancher who keeps a heifer gives up her slaughter value today in exchange for calves that will not arrive for years. With the national herd near historic lows, reducing some of that risk makes sense.

But now read Monday’s announcement for what is missing.

There is nothing that expands interstate sales.

On Friday, the Secretary specifically promised to expand ranchers’ ability to sell across state lines. On Monday, there was no expansion of the Cooperative Interstate Shipment program, no new participating states, no accelerated approval process, and no concrete action that allows a small state-inspected processor to reach customers across a state border.

There is also no identifiable removal of the processing regulations the administration described as red tape. The announcement says the administration is cutting red tape, but does not identify the rule being eliminated. The promised rescission of outdated guidance is not there. Neither is the promised action on faster food-safety data. Truth in labeling largely amounts to a reference to work already done.

Most importantly, there is nothing allowing farmers and ranchers to process their own animals for commercial sale, which was the specific proposal the President announced on Friday.

Massie said that would require Congress. Three days later, USDA effectively confirmed his point. The administration announced everything it could do administratively and left untouched the statutory prohibition that Congress would have to change.

There is also considerably less new money here than the announcement initially suggests. Most of the major spending programs cited by USDA had already been announced during the previous year. The significant new financing tool is a loan guarantee. That may help a processor borrow money, but it is not new processing capacity, and it certainly does not put a new slaughter plant on the ground Monday morning. That is the bait and switch that the language hides, and the farmer is the mark.

One sentence in USDA’s announcement deserves more attention than it will probably receive. The Department acknowledges recent processing-plant closures and argues that the resulting capacity could create an opportunity for American-owned independent processors and cooperatives.

Perhaps. But a closed processing plant is not independent processing capacity. Someone has to buy it, finance it, staff it, obtain the necessary approvals, and reopen it. Calling a closure an opportunity does not accomplish any of those things.

And while all of this was being announced, the first tranche of the new tariff-rate quota for imported lean beef opened on September 1. The administration’s announcement about putting American ranchers first does not mention it.

There is another political problem. Major cattlemen’s organizations have opposed the President’s proposal to loosen on-farm processing requirements, arguing that food-safety inspection should not be treated as unnecessary red tape. Whatever you think of that argument, it means the President’s most consequential proposal already faces opposition from organizations claiming to represent the producers it is supposed to help.

So the tally is fairly simple. The heifer insurance provision is real and potentially useful. Greater federal purchasing of locally processed American beef could matter if it is actually reflected in contracts. There are some useful changes involving conservation and financing. But much of the rest consists of programs already announced, a new office, and promises that have not yet become policy.

The central promises made Friday remain promises.

And for the small producer we have been following through this essay, remarkably little has changed. The freight bill is the same. The federal inspection requirement is the same. The platform still owns the customer. And a farm in Madison County still cannot sell state-inspected meat across the Potomac because Virginia has not joined an interstate program that has been available for fourteen years.

What You Cannot Taste

Suppose every barrier to farm sales of beef were removed tomorrow. Suppose Richmond signs, Congress passes PRIME, and the carriers cut their rates. You still face the problem that makes direct marketing hard, and it is older than any of this.

You can taste tenderness. You cannot taste whether that calf got antibiotics in March.

Economists have a name for this distinction. Some qualities you can judge before you buy. Others, like flavor and tenderness, you discover when you eat the product. But there is a third category that matters enormously to small farmers: qualities the customer cannot verify even after the food has been eaten. Economists call these credence attributes, because ultimately the customer has to believe you.

Now sort the claims on a package of premium beef. Flavor and tenderness can eventually be judged by the customer. Grass-fed, no antibiotics, pasture-raised, truly organic, rotationally grazed, particular feeding practices, and most regenerative claims cannot. Nobody can taste them in a steak.

And those are precisely the claims for which the producer expects to be paid a premium.

A customer in Arlington does not pay eighteen dollars a pound simply because the package contains beef. She pays it because she believes something about how that animal lived, what it ate, what drugs it did or did not receive, and how the land beneath it was managed. Unless someone verifies those claims for her, she is buying the farmer’s word along with the steak.

That creates a problem for the honest farmer.

Economist George Akerlof described the basic mechanism in his famous 1970 paper on the market for “lemons,” work that later helped earn him a Nobel Prize. When buyers cannot reliably distinguish the good product from the bad one, they become unwilling to pay the full price of the good one. The dishonest seller can make the same claims without bearing the cost of keeping them. The honest producer either accepts a lower return, leaves the market, or eventually starts cutting corners himself.

The problem for the honest producer is simple: doing it right costs more, while claiming you did it right costs nothing.

The Reviews Will Not Save You

The standard answer to the trust problem is customer ratings. They fail for two different reasons.

The first is obvious once you think about it. A reviewer knows no more about how your cattle were raised than the person reading the review. A five-star review saying that you can really taste that the beef is grass-fed tells you that the customer liked the steak. It does not verify that the animal was grass-fed. Ten thousand reviews cannot establish a fact that none of the reviewers was in a position to observe.

The second problem is that the reviews themselves cannot necessarily be trusted.

Researchers who gained access to private Facebook groups where Amazon sellers purchased fake reviews were able to follow what happened afterward. Buying reviews worked. Ratings rose, review counts increased, and sales followed. Amazon eventually detected and removed many of the fraudulent reviews, but often only after the seller had already received the benefit. The researchers also found that the sellers buying reviews tended to be selling worse products. Once the manipulation stopped, ratings fell and one-star reviews increased.

More recent research has identified the larger consequence. Fake reviews do more than mislead someone about one product. Once customers understand that ratings can be manipulated, they begin to distrust the rating system itself. The fraud committed by the dishonest seller therefore damages the honest seller as well.

That brings us straight back to the problem of the farmer selling on trust. Reviews are reasonably good at answering Did you like the steak? They are almost useless at answering Did this animal live the way the farmer says it did?

A platform adds another problem: control over provenance. On large marketplaces, the listing, seller, fulfillment system, and physical product can become separated. Amazon historically commingled identical inventory from multiple sellers in some parts of its fulfillment system, meaning that the physical unit reaching a customer did not necessarily come from the seller whose inventory entered the warehouse. Even where inventory is kept separate, the review generally belongs to the product listing rather than documenting the history of the particular object that arrived at the customer’s door.

For ordinary commodities, that distinction may not matter very much. For food sold at a premium because of its provenance, it matters enormously.

The farmer is not really selling a steak. The farmer is selling a steak and a history of that steak. If the system cannot preserve and verify that history from the pasture to the customer, the thing that justifies the premium disappears.

You Cannot Fix a Lie by Carving It in Stone

Which brings us to the blockchain proposals, and to the conversations Jill and I have been having with people building them.

A distributed ledger can prove that a record has not been altered. It cannot prove that the record was true when someone entered it. If a producer or broker enters grass-fed, no antibiotics, the blockchain can preserve those words perfectly and forever, whether or not the animal ever saw a pasture.

This is known as the oracle problem: at some point, information from the physical world has to enter the digital system, and the computer has no independent way of knowing whether that information is true.

Blockchain can be extremely useful for traceability. Walmart famously demonstrated that food which once took days to trace through a supply chain could be traced in seconds. During an outbreak, that matters enormously. But tracing a claim is not the same thing as verifying it.

Economist Michael Spence gave us a useful way to think about this in 1973. A signal conveys useful information when it is harder or more costly for the dishonest person to produce than for the honest one.

Run that test on blockchain. Is typing grass-fed into a ledger more difficult for the farmer whose cattle ate grass than for the farmer whose cattle did not? No. It costs both of them essentially the same. The cryptography can secure the record, but it cannot secure the truth of the claim.

An honest producer and a dishonest one can enter the same claim into a computer. The difference appears only when someone checks whether the claim is true.

Now run the same test on things that actually verify claims. An unannounced inspection is easy for the farm doing what it says and dangerous for the one that is not. Random laboratory testing imposes little expected cost on the honest producer and potentially enormous cost on the fraud. Both work because they reach outside the information system and back into the physical world.

Laboratory testing can already tell us considerably more than most consumers realize. Stable-isotope analysis can detect chemical signatures left in an animal’s tissues by what it ate and where it lived. In some circumstances, laboratories can distinguish a forage-heavy diet from a corn-heavy one and identify geographic patterns that would be extremely difficult to fake with a label. Similar techniques are already used in food-authenticity work.

But the technology has limits. Biology is messy. A conventionally raised animal eating mostly grass can resemble an organic animal, while an organic animal fed substantial grain can produce a different signature. Newer methods have improved the ability to distinguish production systems, but they still do not provide perfect reconstruction of an animal’s life.

They do not need to.

Verification does not require testing every steak. It requires auditing enough of the claims to make fraud expensive.

Random testing that detects most violations, combined with unannounced inspections and penalties large enough to matter, changes the economics of lying. The honest producer has little to fear from an unexpected inspection or laboratory test. The dishonest producer has to operate knowing that any animal, shipment, or claim could be checked.

That is the problem that the blockchain proposals miss. The problem is not creating a permanent record of what the farmer said. The problem is making it expensive for the farmer to lie.

An Alternate Solution: The Kosher Certification Model

Here is the part that surprised Jill and me.

The problem described above has already been solved: a quality the buyer cannot personally verify, sold at a premium across great distances to people who may never see where the product was made. No blockchain was required.

Kosher certification has been doing it for generations, and its marks now appear on an enormous share of packaged food sold in American grocery stores.

The system begins with a person.

A mashgiach is a trained supervisor responsible for making sure that a facility complies with the requirements of the certifying organization. Depending on the product and the risk, supervision may be continuous or involve inspections, including unannounced visits. The important point is that the supervisor’s accountability runs to the certifying system, not merely to the producer whose claims are being verified.

Kosher slaughter goes further. The shochet, the person performing the slaughter, is specially trained and accountable for doing the work correctly. Verification is therefore not simply an inspection performed after the fact. It is built into the process itself, and the standing of the person performing the work depends upon maintaining the standard.

The same model already exists in another market built almost entirely on credence claims: dietary supplements. A consumer cannot look at a capsule and determine whether it contains the stated amount of an ingredient, whether it is contaminated with heavy metals, or whether the contents match the label at all. Independent organizations such as USP and NSF test products and manufacturing systems against defined standards and allow products that meet them to carry their mark of certification (third-party testing). The manufacturer pays for certification, but the value of the certification depends on the reputation of the organization whose mark appears on the bottle. The consumer is not being asked simply to trust the supplement company. A third party with its own reputation at stake is standing behind specific claims about what is in the product.

Compare that with the modern regulatory model. The person doing the work is an employee of the processor, while government pays an inspector to verify that the work complies with federal requirements. Kosher certification developed a different system: qualification and accountability are embedded much closer to the work itself.

But the most interesting part has little to do with slaughter.

Kosher certification and dietary supplements certification have many competing certifying organizations, and the producer seeking certification generally pays for it. That arrangement should create an obvious problem. If the producer pays the auditor, and customers cannot distinguish a rigorous auditor from a permissive one, producers have an incentive to choose the easiest certification available. The weakest certifier should win on price, and standards should gradually fall.

That problem is hardly theoretical. The issuer-pays model contributed to spectacular failures among credit-rating agencies before the 2008 financial crisis.

Yet kosher certification developed differently. Different certifying organizations acquired reputations for different levels of rigor, and consumers who care about those differences, learn the names and marks.

The reason is remarkably simple. The certification mark identifies the certifier.

An OU does not merely say kosher. It identifies the Orthodox Union as the organization standing behind the certification. Other certifiers put their own names and reputations behind theirs. The consumer therefore does not have to evaluate an abstract claim made by the food company. The consumer can evaluate the organization willing to put its reputation behind that claim.

That changes the incentives. A certifier that becomes known for approving products it should reject damages the value of its own mark. A rigorous certifier can build a reputation precisely because it refuses certification when the standard is not met.

The product may travel a thousand miles from the person who made it. The buyer may never visit the plant. Neither needs to know the other. Trust travels with the mark because a named institution has put its own reputation behind it.

What Spain Did About a Pig in the Woods

Joel runs hogs in the woods at Polyface, fattening them on what falls from the canopy. Consider what the customer is actually buying with that claim. The pork may taste different, but the person eating it cannot determine from the pork chop whether the pig actually lived in the woods and fattened on what grew there. That part is a credence attribute, and it accounts for much of the value of the product.

Spain has been dealing with almost exactly that problem, with the same animal, for generations. The result is some of the most valuable pork in the world.

Spanish law protects the highest designation, de bellota, for Iberian pigs finished during the montanera, the season when the animals roam the dehesa woodlands feeding primarily on acorns and pasture. The rules govern how the pigs are raised, what they eat, how long they remain in the finishing system, and how they are identified. Pork produced under other systems can still be sold, but it must be sold under different names. The customer can therefore distinguish the animal finished on acorns in the woodland from one raised outdoors with supplemental feed or one raised conventionally.

The important part is how Spain verifies the claim.

Inspection begins with the land. Before pigs are turned out for the season, independent inspectors assess whether a particular piece of woodland can actually support the number of animals the producer intends to finish there. The land itself must qualify for the system. The number of pigs is tied to the carrying capacity of the ground.

That is a very different form of verification from asking the producer afterward whether the pigs ate acorns. The claim is tied to something physical that can be inspected before the product exists.

Spain protects the language as well. Terms associated with the traditional production system are reserved for products that actually qualify for them. Even imagery that could lead a consumer to believe that ordinary pork came from the protected system is restricted. Pata negra, the famous term associated with the highest grade of Iberian ham, cannot simply be attached to whatever product a marketer would like to sell at a premium.

Compare that with the language surrounding American premium food. Regenerative. Pasture raised. All natural. Farm fresh. Some of these terms have limited regulatory definitions in particular contexts; others are vague, inconsistently defined, privately defined, or largely marketing language. And pastoral imagery can do much of the work without making a specific claim at all. A red barn, green pasture, split-rail fence, and grazing animal can imply a production system that the words on the package never actually promise.

Spain approached the problem from the opposite direction. It did not merely create a premium label and ask consumers to trust it. It defined what the premium claim means, tied that claim to observable conditions on the farm, required independent verification, and prevented everyone else from borrowing the language and imagery that give the claim its value.

That is how a pig in the woods becomes a product worth protecting rather than a marketing story anyone can print on a package.

Where the Organic Seal Went Wrong

Now look at the green USDA Organic seal in that light.

One standard. One meaning. The accredited certifying agent who did the actual work is effectively invisible to the shopper. She cannot distinguish a demanding certifier from a permissive one, which means no certifier can earn anything by being demanding.

That is why organic drifted to hydroponic berries and confinement dairy, and it was not an enforcement failure. It was a design error, made at the beginning, and it followed necessarily from hiding the name of the party doing the vouching.

There is a deeper problem underneath this that Friedrich Hayek identified in 1945 and that farmers understand instinctively. Knowledge about a farm is local. Soil, water, pasture, animals, weather, insects, disease pressure, rotations, and a thousand other things vary from one piece of ground to another. Much of what a good farmer knows from living on that ground cannot be reduced neatly to a form. Berries grown in water under a lamp satisfy every box on the organic form, and that is precisely the point.

When someone offers to solve your trust problem with a new standard, a new seal, or a new ledger, ask two questions. Whose name is on it, and what does he lose if he is wrong.

What the Courts Worked Out

States spent most of a century trying to enforce kosher standards directly, and the courts eventually stopped them. New Jersey’s regulations fell in Ran-Dav’s County Kosher in 1992. Baltimore’s ordinance fell in Barghout in 1995. New York’s statutes fell in Commack Self-Service Kosher Meats v. Weiss in 2002. The reasoning was essentially the same. Enforcing the standard required the state to resolve questions on which the certifying authorities themselves disagreed, which the First Amendment forbids.

But look at what the courts said was still permissible. Stopping a vendor from claiming that a particular authority endorsed a product when it did not requires no interpretation of doctrine at all. That is ordinary fraud. New York rebuilt its law on that basis, requiring disclosure of who certifies and under what standard, and when the new regime was challenged, the Second Circuit upheld it in 2012.

Transfer that to beef and the policy almost writes itself. Government does not define grass finished. Government requires the producer to disclose who attests to the claim and under what published standard, and then prosecutes false claims about that attestation. The standards stay private, plural, and competitive. The state supplies the fraud remedy and nothing else.

That is the inverse of what is being built now. The FDA’s Food Traceability Rule specifies the data elements and tracking events, requires records producible within twenty-four hours, and says nothing about whether the claims attached to the food are true. Compliance costs under such a regime are close to fixed per business, which means they fall hardest per pound on the smallest producer. The rule was set for January 2026. FDA proposed a thirty-month extension in August 2025, and Congress made it binding in the November 2025 continuing appropriations act, directing no enforcement before July 20, 2028. Ask who lobbied for that delay and got it. The Food Industry Association and the National Grocers Association, not the Virginia farmer selling pastured broilers and turkeys, woodland-raised hogs, and grass-fed beef by the quarter.

There is until July 2028 to replace a traceability system that burdens small producers with one that actually verifies the claims consumers are paying for.

Seven Tests

We cannot tell anyone exactly what to build or what will ultimately work. Neither can the government. The solution will have to emerge through trial and error, with competing systems tested and selected in the free market. But it is possible to identify what an honest system must do to survive, and every proposal can be measured against these seven tests.

1. Does the record tie to the animal, or only to the paperwork? Hair and tissue keep. A tag falls off in the brush, and a lot number gets retyped by whoever is closest to the keyboard.

2. Whose name is on the claim, and what does he lose if it is false? An anonymous assertion is worth what was paid for it.

3. Can the producer use the shipping system without surrendering the customer list? If the answer is no, the producer has merely changed landlords.

4. Does it cost by the head or by the year? If it is priced by the year, it selects for scale and will bury the small producer. Priced by the head, it treats a twenty-cow operation the same as a twenty-thousand-cow operation.

5. Does anyone ever check, at random, without warning? Continuous surveillance is unaffordable and unnecessary. Occasional unannounced checking is neither.

6. What happens to a cheat when he is caught? If the answer is a stern letter, the system is merely an expensive way of recording what people say about themselves.

7. Can a woman in Arlington understand it in four seconds, standing up, holding a phone? If not, it is another seal nobody reads.

We know of nothing currently fielded or in development that passes all seven. Most of the ventures raising money right now fail the fourth test badly, which is why they end up selling enterprise software to large processors while describing themselves as friends of the family farm.

Let the Market Find It

We are not going to name the solution. That refusal is the point, not a dodge.

Carl Menger explained how money emerged without anyone designing it. Traders individually preferred goods that were easier to trade, and over time those choices produced a medium of exchange that no committee had specified. Hayek later called competition a discovery procedure: what competition discovers cannot be known beforehand, or there would be nothing to discover.

Every blockchain provenance consortium now working starts from the opposite direction. A committee designs the standard, specifies the fields, and expects the market to follow. That is central planning dressed as a market. The knowledge needed to write the specification is scattered among ten thousand people who were never invited to the meeting.

Whatever works will be discovered by producers and buyers doing business, and it will probably look more like a man standing in a plant than a distributed ledger. Our contribution is the seven tests above, and they are falsifiable.

There is one thing that can be done now. In Virginia, producers and consumers should write their delegate and state senator and ask a single question: Why has Virginia never signed a Cooperative Interstate Shipment agreement?

Virginia is not alone. Eighteen states with their own qualifying meat inspection programs have not joined CIS: Alabama, Arizona, Arkansas, Delaware, Illinois, Kansas, Louisiana, Minnesota, Mississippi, North Carolina, Oklahoma, Oregon, South Carolina, Texas, Utah, Virginia, West Virginia, and Wyoming. If one of those is home, the question for the governor, agriculture commissioner, state senator, and delegate or representative is the same: Why has this state not joined the Cooperative Interstate Shipment program? These states already operate inspection programs required to meet standards at least equal to the federal system. The missing step is opening the interstate market to qualifying small processors. When writing, please link to this article or quote from it so that legislators understand both the question and the larger problem behind it.

Residents of other eligible states that have not joined CIS should ask the same question. The program is fourteen years old, the federal government pays sixty percent of eligible inspection costs, and eleven states have already joined. Meanwhile, Virginia hands out thirty-six-thousand-dollar equipment grants to farmers it prevents from selling state-inspected meat across the Potomac.

Ask the question in writing. And ask for the answer in writing.

RWM/JGM

References

Akerlof, George A. 1970. “The Market for ‘Lemons’: Quality Uncertainty and the Market Mechanism.” Quarterly Journal of Economics 84 (3): 488 to 500.

Barghout v. Bureau of Kosher Meat and Food Control, 66 F.3d 1337 (4th Cir. 1995).

Camin, Federica, Luana Bontempo, Matteo Perini, and Edi Piasentier. 2016. “Stable Isotope Ratio Analysis for Assessing the Authenticity of Food of Animal Origin.” Comprehensive Reviews in Food Science and Food Safety 15 (5): 868 to 877.

Commack Self-Service Kosher Meats, Inc. v. Weiss, 294 F.3d 415 (2d Cir. 2002).

Commack Self-Service Kosher Meats, Inc. v. Hooker, 680 F.3d 194 (2d Cir. 2012).

Darby, Michael R., and Edi Karni. 1973. “Free Competition and the Optimal Amount of Fraud.” Journal of Law and Economics 16 (1): 67 to 88.

Gandhi, Ashvin, Brett Hollenbeck, and Zhijian Li. 2025. “Misinformation and Mistrust: The Equilibrium Effects of Fake Reviews on Amazon.com.” NBER Working Paper 34161. National Bureau of Economic Research.

Hayek, F. A. 1945. “The Use of Knowledge in Society.” American Economic Review 35 (4): 519 to 530.

Hayek, F. A. 1968. “Competition as a Discovery Procedure.” Reprinted in The Quarterly Journal of Austrian Economics 5 (3), 2002: 9 to 23.

He, Sherry, Brett Hollenbeck, and Davide Proserpio. 2022. “The Market for Fake Reviews.” Marketing Science 41 (5): 896 to 921.

Klein, Benjamin, and Keith B. Leffler. 1981. “The Role of Market Forces in Assuring Contractual Performance.” Journal of Political Economy 89 (4): 615 to 641.

Menger, Carl. 1871. Principles of Economics. Translated by James Dingwall and Bert F. Hoselitz. Reprint, Auburn: Ludwig von Mises Institute, 2007.

Menger, Carl. 1892. “On the Origin of Money.” Economic Journal 2 (6): 239 to 255.

Nelson, Phillip. 1970. “Information and Consumer Behavior.” Journal of Political Economy 78 (2): 311 to 329.

Ostrom, Elinor. 1990. Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge: Cambridge University Press.

Ran-Dav’s County Kosher, Inc. v. State, 608 A.2d 1353 (N.J. 1992).

Real Decreto 4/2014, de 10 de enero, por el que se aprueba la norma de calidad para la carne, el jamón, la paleta y la caña de lomo ibérico. Boletín Oficial del Estado núm. 10, January 11, 2014.

Spence, Michael. 1973. “Job Market Signaling.” Quarterly Journal of Economics 87 (3): 355 to 374.

U.S. Department of Agriculture, Economic Research Service. Food Dollar Series.

U.S. Department of Agriculture, Food Safety and Inspection Service. Cooperative Interstate Shipment Program. Accessed August 30, 2026.

U.S. Food and Drug Administration. Requirements for Additional Traceability Records for Certain Foods, 21 CFR Part 1, Subpart S. Compliance date extension proposed August 7, 2025.