Nobody calls the fire department socialism

When your house is burning down, nobody comes up to you and has you pay a bill first. A truck shows up that the government owns, from a fire station that the government owns, using a hydrant that the government owns, worked by people who are employed and paid by the government through your taxes. There are something like 29,000 fire departments in this country, more than 360,000 career firefighters, and nearly 680,000 volunteers.

You send your kids to schools, in buildings that the government owns, to be taught by teachers employed by the government, funded in large part by property taxes you pay whether or not you have children in them. Nearly 50 million students attend public schools, taught by more than 3 million teachers.

The largest integrated health system in the country is owned by the government. The VA health system owns the hospitals and the doctors are federal employees. It serves 9.1 million enrolled veterans, and it is a textbook definition of socialized medicine.

All of it fits the standard definition of socialism, where the public owns the assets and employs the workers, and no one (no one reasonable) calls it socialism.

But now you’re skeptical. You’re skeptical of socialism and you’re skeptical of everything I’ve just said, and you’re probably saying to yourself, “Lots of people called the VA socialized medicine.” Or you’re saying to yourself, “My fire hydrants belong to the water utility, a utility that is investor-owned. That’s not socialism.” Or perhaps you’re saying, “That’s not what socialism means.”

In America, socialism only becomes a slur if the government uses its collective resources to give more security or power to people who have less of it. When the government assumes costs or risks for those who already have power, we tend to call it something else. We don’t consistently object to socialization in this country. We object to who receives it.

Subscribe now

What the word means

Let’s start with the definition.

The standard definition is social ownership of the means of production. The “means of production” are the things used to make other things. Your toothbrush is not one. A factory is. A power line is. A hospital building is. And social ownership has more than one form: direct public ownership, where the government owns the asset. Common ownership, where a whole society holds it without state involvement. Cooperative ownership, where the thing is owned by the people who work there or the people it serves. A volunteer fire company owned by its members and funded by the town it serves is cooperative ownership.

Historically, socialism was seen as the midpoint between capitalism and communism, so picture a line with those two at the ends.

Capitalism places the means of production, the thing that makes other things, into private hands. The downside of capitalism is that when it’s unregulated it can lead to the exploitation of labor and the concentration of wealth in the hands of very few.

Communism, particularly twentieth-century communism, handed the means of production to a one-party state that set prices and prohibited opposition. Originally the idea of communism was to create a classless society with absolute equality, where the government dictates the price, supply, and distribution of goods. The downside of that is that the only practical way to redistribute wealth with absolute equality tends to be force. So these governments usually end up becoming authoritarian regimes.

Socialism comes in right down the middle. The public owns the society’s major productive assets, through the government, cooperatives, or the workers, while some market systems still determine supply and demand. The downside of socialism is that it is very vulnerable to corruption. If the people running that system are corrupt, it skews toward authoritarianism too.

Democratic socialism then squeezes in between capitalism and socialism. It is socialism with a commitment to political democracy: elections, political pluralism, civil liberties, unions, and popular participation. It rejects the thing we actually fear from twentieth-century communism, the one-party rule that sets price, supply, and distribution.

The last term you need is social democracy. Social democracy leaves capitalism and private ownership intact and uses taxes, regulation, unions, and public programs to spread the benefits and risks more broadly. Universal health care, subsidized public college, strong labor protections, and a generous welfare state do not, by themselves, make an economy socialist.

I promise we are not hair-splitting here. People assign the word socialism to Denmark, and the Danish prime minister has said: “Denmark is far from a socialist planned economy. Denmark is a market economy.” People assign the word to Bernie Sanders, and Sanders has defined democratic socialism twice in formal speeches, at Georgetown in 2015 and George Washington in 2019, both times as Franklin Roosevelt’s unfinished Economic Bill of Rights: guaranteed health care, education, housing, and a secure retirement. Ownership stays where it is. Even the Democratic Socialists of America’s own history describes his program as social democratic and notes that he opposed state ownership of most corporations.

So the definitions are unambiguous, but the feelings are not. Gallup finds 39% of Americans view socialism positively: 66% of Democrats, 14% of Republicans. Two sides hear one word and see two different pictures, and neither picture is the definition. Which means the definition can’t tell you what the word does. For that, you have to watch it work.

Share

The word, aimed down

In 1961, the AMA got a fading actor, Ronald Reagan, to record a 10-minute LP called “Ronald Reagan Speaks Out Against Socialized Medicine.” The point was to target a bill called King-Anderson, or H.R. 4222, which would supply hospital coverage for people over 65. You might know this as Medicare. The campaign was called Operation Coffee Cup. Doctors’ wives hosted coffee gatherings, played the record, and wrote letters to Congress. The bill died and it wasn’t until 1965 that Medicare passed with a Republican bill inside it.

That Republican bill belonged to John Byrnes of Wisconsin, the ranking Republican on Ways and Means, who proposed voluntary coverage of doctor visits, paid partly by the people who signed up and partly out of general tax revenue. Republicans called it Bettercare. In 1965, instead of picking one plan, Congress stacked them on top of each other. King-Anderson became Part A, which covers the hospital. Byrnes’s plan became Part B, which covers the doctor. An expansion of an older program for low-income families became Medicaid.

Somehow we went from the “dangers of socialized medicine” to the same concept with a new name written by a senior Republican.

So we just watched the word kill a bill for old people and then go quiet when it came back under a different name. Does it work the other way?

The same word, aimed up

On August 22, 2025, the United States bought 433 million shares of Intel at $8.9 billion, giving taxpayers a 10% stake in a private company. Rand Paul called it a step toward socialism: “If socialism is government owning the means of production, wouldn’t the government owning part of Intel be a step toward socialism? Terrible idea.” And yes, hell has frozen over, because Rand Paul and I agree on something (even if that something is the definition of socialism). The Wall Street Journal editorial board called the deal a de facto nationalization. National Review said the government should stay out of the chips business.

Trump pushed back and said he’d make deals like this all day long. And he has. The government has now put roughly $27 billion into 37 deals involving direct federal ownership. On July 29, 2026, the Commerce Department announced $874 million in federal incentives for seven semiconductor companies, on the condition that the U.S. government receive a minority ownership stake in every company. The Intel stake alone has quadrupled in value, and you will not find the gain recorded in any budget document. When Fortune went looking for the whole portfolio this summer, its headline was “good luck finding them.”

None of this is popular. In CNBC’s July poll, about half of voters said it is inappropriate for the federal government to own a share of a private company. The government did it anyway.

In June, Bernie Sanders introduced the American AI Sovereign Wealth Fund Act, a one-time 50% tax on the largest AI companies, paid in stock, with a public commission holding the voting shares. Within days The Wall Street Journal published an article titled “The Road to AI State Socialism.” The law remains an active proposal working its way through the legislative process. But ask yourself, how do you think this will turn out?

Look at the scorecard. Aimed down, at hospital coverage for old people, the word killed a bill. Aimed up, at a president buying pieces of private companies over the objection of half the country, the word produced a week of columns and 37 more deals. For some reason the word “socialism” only has a bite to it when its aimed at certain policies and people. Party doesn’t explain it, because the AMA killed a bill under a Democratic president with Democratic majorities. Public opinion doesn’t explain it, because the public opposed the stakes and the portfolio grew.

So if the word is thrown by whoever needs it, and it only lands with force half the time, then who is the word for?

Share

What the word can’t do

In 2022, three political scientists published a paper called “Viewed from Different Engels?” They ran two experiments. In the first, 790 people from a census-matched sample rated three policies: Social Security, welfare, and free college. Half were told the programs were popular in socialist countries. Half were told the same programs were popular in capitalist countries. Support barely moved. The effect was statistically null. Even the people who said they hated socialism kept supporting the policies.

The second experiment showed people a fabricated news story attacking the government’s COVID response. One version blamed socialism, the other blamed capitalism. The label changed which economic system people blamed. It barely changed their support for the response itself.

The same study asked people to name socialist countries. People who liked socialism named Denmark and Norway. People who hated it named Russia, Venezuela, Cuba, North Korea.

Now, that result should bother you, because a few paragraphs ago I told you a record album killed a health care bill. But if the word can’t change anyone’s mind about a policy, a record album full of the word can’t either. So what really happened with the Medicare story?

The LP was just one piece of a much larger, much older plan. In December 1948, facing Harry Truman’s plan for national health insurance, the AMA charged each of its 140,000 members a special $25 fee and raised $3.5 million. It hired Whitaker & Baxter, the first political consulting firm in the country, and ran what was then the most expensive lobbying campaign in American history. Their name for Truman’s plan was “socialized medicine.” When the campaign started, mail arriving in Congress ran four and a half to one in favor of the plan. Nine months later it ran four to one against, and the plan died in committee.

There were no daily tracking polls in 1949. A congressman read his district through his mail, and a congressman cannot tell purchased mail from spontaneous mail. So the campaign never had to change millions of minds. It had to change what letters landed on a few hundred desks, and that is a much cheaper product. Operation Coffee Cup was that same product with a celebrity attached. The coffee gatherings existed to make doctors’ wives write letters that looked like they came from nowhere.

That is the short game. The long game is what the money does to the word itself.

Since 1984, the General Social Survey has asked half of (its sample of) Americans whether we spend too much or too little on “welfare,” and the other half whether we spend too much or too little on “assistance to the poor.” About two-thirds say we spend too little on assistance to the poor. But, call it welfare and support drops by roughly thirty points. In 1996 the political scientist Martin Gilens found the reason. He found that racial attitudes were the strongest predictor of white opposition to welfare, stronger than views about taxes, government, or the poor themselves. Decades of campaigns had attached a face to the word, and racism did the rest.

This pattern starts showing up everywhere once you see it. In 2013, CNBC polled the same law under two names, and Obamacare polled nine points worse than the Affordable Care Act. Gallup finds 81% of Americans view “free enterprise” positively and 54% say the same of “capitalism.” Same system, twenty-seven points apart.

Animosity toward socialism came easy because two generations of Americans heard the word next to the Soviet Union during a time when the Soviet Union was framed as America’s greatest threat. The rest of the pattern the Cold War cannot explain. “Welfare” had no Moscow. “Capitalism” against “free enterprise” had no USSR. Those gaps were built at home, by campaigns, one American dollar at a time.

So the word on its own doesn’t do much, and that is exactly why somebody has to pay for results. The short money buys the appearance of an angry district (think Operation Coffee Cup). The long money buys the meaning of the word (think government ownership of Intel). Which leaves us with one unanswered question: why does the money only ever aim it downhill?

Who gets seen

Follow the incentive. A campaign like the AMA’s Operation Coffee Cup is an investment an industry makes when a policy threatens its revenue. National health insurance threatened how 140,000 doctors billed, so the doctors bought a campaign.

Now run the same logic uphill. The tax exclusion for employer-sponsored health insurance is one of the largest tax breaks in the federal code. If you get insurance through work, the federal government is paying part of your premium, and about 60% of that benefit goes to households above four times the poverty line. Who would fund a campaign against it? The people with the money to fund campaigns are the people it pays. So the campaign has never existed, and the biggest health subsidy most working Americans receive has never spent a single day being called socialized medicine.

There is a second reason nobody aims the word uphill. The political scientist Suzanne Mettler calls it the submerged state: government that delivers its benefits through tax rules and subsidies, so the recipient never sees the government’s hand.

Help for people with less money is designed to be seen. You apply for it. A caseworker approves it. You recertify for it. You pull out an EBT card at the register in front of other people. The program has a name, a building, it’s argued about on debate stages.

Help for people with more money is built to be invisible. It turns op as a rule in the tax code no one wants to read, with no application, no caseworker, no card, and no debate. The Congressional Budget Office ran the numbers on the major tax expenditures and found half the dollars go to the top fifth of households. Only nine percent reach the bottom fifth.

When the pandemic hit, the Paycheck Protection Program moved roughly $800 billion. The economists who studied it estimate that 23 to 34 cents of every dollar reached workers who would otherwise have lost their jobs, and about 75% of the money went to the top fifth of households. It was called emergency relief, liquidity, stabilization. It was the government socializing private payroll risk, and no coffee pots were harmed.

Martin Luther King said it in his “Other America” speech in 1968: “we all too often have socialism for the rich and rugged free enterprise capitalism for the poor.” The help going down gets a program name, a building, a line to stand in, and a word waiting for it. The help going up gets called growth, relief, liquidity, competitiveness, industrial policy. You cannot aim a slur at something you cannot see, and nobody with the money to make you see it wants you to.

That is the answer to “who is the word for.” The word is for the government you can see. And the government was built so that what you can see is the help going down, while the help going up hides away as a line in the tax code. The word doesn’t create the asymmetry. The architecture does. The word is the alarm that only rings on one side of the building.

Share The Minuteman

The test

If all of this is true we should be able to test it. So let’s through two harder cases at it and see what happens.

Nebraska is the most socialist place in America by the definition this article opened with. Every home and business in the state buys electricity from a utility the public owns: 121 city utilities, 30 public power districts, 10 cooperatives, and zero investor-owned companies. Nobody in Nebraska has bought a kilowatt-hour from a private power company since 1949. Power lines are means of production. The public owns all of them. Nebraska votes Republican by more than twenty points, and nobody calls it socialist, including the people who live there, who call it public power and like it.

If the word tracked ownership, Nebraska would be radioactive. If the word works the way I’ve described, Nebraska is invisible to it, because no industry loses money while the lights stay on, so no one has ever funded the campaign against it. The check clears for the whole state, so no one gets seen cashing it.

The VA runs the same test from the other side. The government owns the hospitals and employs the doctors, and for decades the word mostly left it alone, because the beneficiary is a veteran. Then in 2014, a scheduling scandal broke at a Phoenix VA hospital, and a group called Concerned Veterans for America, funded by the Koch network (yes, that Koch), went to work: congressional testimony, press conferences, ad buys, all pushing to move veterans into private care. Its executive director pointed at Phoenix and called it evidence of “socialized medicine at its worst.” The word showed up exactly when someone wanted the program gone, and not a day before. That is a tool coming out of the drawer when it benefits the person who can afford it.

The fire truck

Which brings back the fire truck.

You already own a fire department, a school, a road, a library, and the largest health system in the country. You’ve owned some of it your whole life, and nobody had to sell you on it, because you can see it, everybody benefits, and nobody is spending a dollar to make you afraid of it.

The word socialism will keep getting thrown around. When it comes, don’t argue about the word. We spent a good portion of this article solidifying what that word means. Instead, ask the right questions. Who owns the thing? Who decides how it runs? Who takes the profit? Who absorbs the loss? Who gets seen receiving the help, and who receives it invisibly? And the one this whole story adds: who paid for you to fear the word, and what do they lose if the thing survives?

The government already shares costs and risks. It has for as long as you’ve been alive. The only fight left is over which ones, for whom, and whether you’re allowed to see it.

Want to help The Minuteman for free? Share it.

Sharing does not cost you anything. And it supports my work & independent journalism!

Share The Minuteman

New here? Sign up for free.

If someone did send you this, thank them and subscribe for free here:

Subscribe now

Issue No. 3 — The AI Boom Can’t Set Its Own Interest Rate

The power system behind a headline, before it shows up in your work, wallet, rights, or community

Subscribe now

A WARNING, and some help…

This made my brain hurt alittle. What we are about to get into is pretty technical. If you can’t make it halfway down the page before you start to feel a sharp pain behind your eyes, I don’t blame you. This is A LOT. But it is also extremely important that we all undertand the impact of what you’re about to read. So, before we get into the details, here is a simplified version of what i explain below:

The premise. AI companies build very large data centers. Until 2024, they paid for these data centers from their profits. Now the costs are larger than the cash that most of these companies make. They borrow the difference, and they keep much of this debt in separate companies. They also give money to their own customers, so that the customers can buy from them. They can make their own customers, but they cannot set their own interest rate.

These companies now do the work of a bank. They move money from savers to builders. They hold the risk between the two. A bank has government supervision. These companies do not.

Why this is important to people. The lenders include pension funds and insurance companies. These funds hold the retirement money of workers. The workers did not select these data center projects. If the projects fail, the funds can lose money. New York publishes a list of what its pension fund owns. Most states publish less, so most workers cannot see the risk.

The New York Comptroller controls that pension fund. In March 2026, he told 19 technology companies to pay for their own electric power. His pension fund holds investments in data center companies. The law tells him to invest for good returns. The law does not let him sell these investments for policy reasons. He must continue to invest in these same companies.

The implications. Investors do not quote these bonds as a simple percentage. They quote each bond as the United States government rate plus an additional amount. The additional amount pays for the risks of the project. In October 2025, one bond paid for a Meta data center in Louisiana. That bond had strong protection, and it still paid 2.25 percent more than the government rate. Today the government pays about 5.25 percent to borrow money for 30 years.

Each additional percentage point increases the money that a project must earn later. The AI companies cannot avoid this cost. If a bank lends, the bank includes the government rate in its price. If a private credit fund lends, it asks for more than the bank. If Nvidia promises to pay a customer’s rent, that promise has a cost. The rate is always in the price.

There is a limit to this argument. The government rate changes the cost of AI debt. This does not show that AI debt changes the cost of other loans. Your home loan is not part of this story.

What I bet, and why. I make three calls, and you can check them later. First, Microsoft, Alphabet, Amazon, and Meta keep good credit through June 2027. I do not include Oracle, because S&P lowered its rating in July 2026. Second, state insurance regulators publish the first good data on this risk in 2027. Third, AI bonds become a larger part of the bond market.

The same logic supports all three calls. Pressure on credit shows first in prices and in loan rules. It shows last at the largest companies. In the summer of 2026, CoreWeave paid 1.0 to 1.25 percent more than it planned for a loan. CoreWeave also accepted new conditions for the life of that loan. The weakest borrowers fail first.

Now for the nitty-gritty.

Share

BRIEFING BOX
Core mechanism: The spread over Treasuries. Every AI infrastructure bond is priced as the risk-free government rate plus a premium for everything that can go wrong, which puts a number the industry doesn’t control inside every deal it signs.
Main risk: The build-out’s funding has moved off the tech balance sheets and onto pension funds, insurers, and private credit vehicles that hold long-dated paper the public can’t easily see or price.
Decision room: Bond and private credit markets, plus state insurance regulators. Congress and FSOC have both passed.
Time horizon: 2026–2028.
Who should care: Corporate finance and treasury teams, credit and fixed income investors, public pension trustees and staff, insurance and risk professionals, anyone advising on AI infrastructure exposure.
My call: Microsoft, Alphabet, Amazon and Meta keep investment-grade market access through mid-2027, and any early restructuring happens at a neocloud or a project vehicle instead (75%). Full reasoning below.

THE HEADLINE

The story you’ve been given about AI spending is that the richest companies on earth are writing checks out of profits so large the rest of us are watching from outside the transaction. That was accurate through 2024. Since then the money has been coming from somewhere else, and a growing share of it belongs to people who never chose to lend it.

THE ONE-LINE TAKE

AI companies have gotten very good at manufacturing their own demand, and the price of the money that pays for it is the one thing they still have to buy from strangers.

THE POWER SYSTEM

Start with the number people feel.

On October 16, 2025, an entity called Beignet Investor LLC sold a single bond worth $27.294 billion. It priced at par with a fixed coupon of 6.581% and runs until May 2049. S&P rated it A+, one notch under Meta itself. Morgan Stanley, the only bank on the deal, quoted the price the way this market actually quotes prices: Treasuries plus 225 basis points.

That is roughly as safe as a data center deal gets. The lenders sit at the front of the line if anything goes wrong, the loan pays down across its whole life instead of ballooning at the end, and the borrowing entity is walled off so trouble elsewhere can’t reach it. The money to repay it comes from Meta’s lease payments on the Hyperion campus in Richland Parish, Louisiana, and Meta is one of the strongest corporate credits in the world.

The deal still cleared at 2.25 percentage points above what the United States government pays to borrow. Morgan Stanley had locked those terms in months ahead, with investors lined up by the start of the second quarter for an October pricing and no new buyers approached after April, and the bonds jumped about 10 points once they started trading, so a widely marketed deal might have come tighter. The 225 is a derisked price. It is still 225.

Now the part the headline skips.

For most of the last decade these companies paid for their own construction. The cash from the existing business covered the new buildings, which meant no outside lender had a vote on any of it. That arrangement ended fast. Across the five biggest spenders, borrowed money went from 9% of capital spending in fiscal 2024 to 32% by the middle of 2026, and free cash flow is now expected to reach zero or turn negative this year at every one of them except Alphabet and Microsoft. Their bond issuance passed $100 billion in 2025, most of it borrowed for longer than five years to cover build-outs that take that long. In June, Alphabet priced an $84.75 billion equity raise, the largest ever by a listed company, with $44.75 billion of it pointed at AI infrastructure and general corporate use. Goldman Sachs expects more than a third of AI investment to be debt-funded by 2027.

Public bonds couldn’t carry all of it, and putting the rest on the balance sheet would have shown up in the leverage numbers investors watch. So the industry built a second route. A separate company gets created to own the data center. Outside investors put in most of the equity and raise the debt privately. The tech company takes a minority stake, signs a long lease, and sometimes guarantees payments. The building gets built, the tech company uses it, and the borrowing stays off the tech company’s books. The Bank for International Settlements has a name for this. It calls the whole arrangement “shadow borrowing”, meaning obligations that work like debt while living outside the borrower’s balance sheet.

Beignet Investor is what that looks like with the paperwork filed. Blue Owl Capital owns 80% of the vehicle and Meta owns 20%. PIMCO anchored the bond, and the whole $27 billion went to what the underwriter described as a low double-digit number of investors. The Louisiana campus gets financed, and Meta’s balance sheet reports a lease.

Identify the mechanism.

Here is the hinge the whole story turns on, and it has a boring name, call it the spread over Treasuries.

Bonds like these don’t get quoted in plain percentages when they’re sold. They get quoted as a Treasury yield plus a number of extra basis points, one basis point being a hundredth of a percentage point. A Hut 8 data center construction bond priced this year at T+165, meaning the matching Treasury rate plus 1.65 percentage points. The government’s borrowing rate is a term inside that contract, written into the price on the day it sold. And that rate has been climbing. The 30-year Treasury now pays about 5.25%, near the top of where it has traded all year.

That’s the trap. An investor deciding where to put money for twenty years can lend it to the United States government and take on no construction risk, no risk that the chips inside the building are obsolete in four years, no risk that a single tenant walks, and no risk that the power never gets connected. To pull that investor into a data center instead, someone has to pay the government’s rate and then pay again for every one of those risks. When the government’s rate goes up, the data center’s rate goes up with it, and every extra percentage point raises the amount of AI revenue the project has to earn someday to have been worth building.

There’s no route around it. If a bank writes the loan, the bank prices Treasuries into the loan. If a private credit fund writes it, the fund wants more than the bank did. If Nvidia guarantees the lease, that guarantee is a liability with a market value. If Nvidia funds a customer with equity instead, Nvidia’s own shareholders eat the return they gave up. The rate shows up somewhere every single time, and the only question is whose line item it sits on.

Share

Look at the tell.

When the seller starts writing insurance on the buyer, the seller has changed businesses.

Nvidia now guarantees some of its partners’ facility lease obligations if those partners default. The company’s own quarterly filing puts maximum gross exposure at $3.5 billion across terms of five to seven years, notes that partners have placed $712 million in escrow against it, classifies the guarantees as credit derivatives, and calls them not material. Read the same paragraph for what Nvidia collects. It takes the guarantees on in exchange for warrants, which means it’s accepting credit risk on its customers and getting equity upside for doing it. That is what a lender does. Nvidia has since signed agreements with Blackstone, BlackRock, Apollo, Brookfield, Goldman Sachs and KKR to build financing platforms for its buyers.

The second tell came from the other side of the table. CoreWeave rents out AI computing power and has no second profitable business to absorb a bad year. Nvidia holds a stake in it and backstops some of its capacity. When CoreWeave went to raise a $2.6 billion facility this summer, the lenders made it pay, 100 to 125 basis points wider than where it started, and attached conditions the company has to keep clearing for the life of the loan, including a test of whether earnings cover the debt payments and a requirement to pay the principal down throughout instead of at the end.

You can watch the same repricing in the market for insurance against these companies failing. CoreWeave’s five-year default protection hit 855 basis points in late July, up from around 640 in November, a price implying something close to coin-flip odds over five years. Oracle’s went from about 145 basis points at the end of 2025 to above 215, and on July 9 S&P cut Oracle to BBB-, one notch above junk, on surging capital spending, negative free cash flow and customer concentration. The people lending the money have started charging for the risk the vendors keep describing as immaterial.

Look at the other side.

One complication, it’s the strongest thing anyone can say against this.

Nothing here is breaking. Bank of America puts AI-linked bonds at 2% to 3% of public investment-grade and high-yield benchmarks, which is small. Investment-grade hyperscaler spreads have actually tightened a little in recent months. The IMF’s spring assessment treats aggregate pension and insurer exposure to private credit as manageable. Anyone telling you a crisis is underway is ahead of the evidence.

Two things about that. Telecom bonds were 1% of the index in 1995 and 20.3% by 1999, so a small share is a fact about today rather than a fact about the direction. And I want to be careful about a claim I can’t make. “T plus a spread” proves the Treasury market prices AI debt. It doesn’t prove the reverse, that AI borrowing is making credit more expensive for everyone else. Extra bond supply can push spreads wider when buyers are stretched, and the AI issuers are competing for the same long-duration institutional money as other large borrowers. Whether that’s raising anyone else’s borrowing costs yet is an open question. Your mortgage isn’t in this story, yet.

That’s the system.

The AI industry built a lending machine that can create its own customers, its own guarantees and its own demand, and then had to walk outside and buy the one input the machine can’t produce, which is the price of money, from lenders who owe it nothing.

WHAT THIS MEANS

For your wallet. If you have money in a public pension or a life insurance product, the chance that it’s funding data centers has gone up, and in at least one state you can read the line item.

Subscribe now

Read more

Zuckerberg’s Case for Superintelligence Leaves Out How Meta Makes Money

On August 10, 2026, Mark Zuckerberg published an essay on AI safety called “The Future is for Everyone.” It focuses on safety and superintelligence, which is the industry’s word for an artificial intelligence that can outperform people at more or less everything. Nobody has built “superintelligence", but every large AI company says it’s close.

His argument is pretty simple.

The danger of this technology is not the technology itself, it is that a small number of companies or governments end up controlling it. Therefore the safest thing anyone can do is hand it to as many people as possible, as fast as possible.

He then spends the rest of the essay assigning that answer to the biggest AI risks. Job loss, bioweapons, cyberattacks, government surveillance, the possibility of a Terminator-like future where machines no longer serve us. Each concern gets the same answer: spread the technology wider.

Funny enough this also happens to be Meta’s business model.

Meta gives its models away. Most of its competitors don’t. Companies like OpenAI and Anthropic make money by selling access to models that they keep in-house, which means that a customer pays per use and the company controls those terms. Meta releases some of its models openly, meaning anyone can download the finished system and run it for themselves for free. Meta makes its money the way it always has: on advertising, on the products those models feed. Essentially if you’re talking to its AI about cool places to go hiking, its system will show you hiking groups, friends’ posts about trails, or ads for hiking gear. And that’s true across all their products. The information AI collects on you can cross between their apps.

So Zuckerberg’s argument that fast, free, and universal distribution of this technology is safest is also the argument that Meta’s business model is the safest and its competitors’ strategies are the dangerous ones. Zuckerberg never discloses this in the essay. Instead he states it as a discovery about AI safety.

And he leaves a lot out.

What about Advertising Mark?

In 6,500 words about a company that makes most of its money on advertising and wants to hand its personal AI to billions of people, the word “advertising” doesn’t appear once.

Meta’s revenue comes almost entirely from advertising. The company collects information about people, what they do, what they want, and sells businesses the ability to reach those people. That’s their whole business model and it’s been their business model for 20 years.

The product Zuckerberg is describing in this essay is an agent. In his words it will understand you, your goals, and everything you care about, and will work around the clock on your relationships, your health, your career, your finances, your home, and your hobbies. He’s proposing to put a piece of software inside the most private parts of a person’s life, built by a company whose income depends on knowing what people want and selling that knowledge to other businesses.

But he never addresses the conflict of interest. He doesn’t argue that the conflict has been solved or that Meta will change how it makes money or that the agent will be paid for in some other way. He just doesn’t bring it up.

Then he gives an example of him and his daughter using this agent to plan recipes that they can bake together on weekends. The agent orders ingredients and offers suggestions while they cook.

But read that again with the business model in mind. If the agent orders the ingredients, someone sells those ingredients. Someone sells the flour. Someone sells the butter. Somebody’s brand gets typed into the order and someone’s brand does not. The company that built the agent has spent decades perfecting the algorithm to decide which brand wins.

Zuckerberg doesn’t tell us who makes that choice. He doesn’t tell us whether a company can pay to be chosen and he doesn’t tell us whether the agent will disclose it when it makes a recommendation or a purchase. He uses this example of him and his daughter to show some humanity, and then he walks past the only obvious question that this example actually poses.

Ultimately an agent that just shops for you is an advertising product, and he wrote 6,500 words and never said so. And you expect us to trust you?

Three more things that never come up

From my read the advertising problem is the largest hole but it’s not the only one. There are three more absences that run through the essay and all three of them do the same thing. Each one removes a way of checking whether that promise that he takes 6,500 words to make is being kept.

First, what about the money it takes to build any of this? The essay argues against centralized AI while describing an undertaking that maybe five entities on Earth can finance.

The things we need to make this possible are already centralized: the chips, the buildings, the data centers, the cooling infrastructure, the enormous amounts of electricity, the capital. Zuckerberg spends a full section talking about data centers Meta is putting up around the country and the power plants it’s building to run them on. Instead of talking about the centralization of the physical AI infrastructure, he just ends the section by asking the government to make it easier to build both.

A handful of firms on this planet can finance a build-out that Zuckerberg is talking about at this scale. The distribution that he spends 6,500 words arguing for happens at the last layer, where someone uses AI software that’s already been created that runs on infrastructure that’s already been built. Meta trains the models. Meta owns the machines. Meta sets the boundaries the agent operates inside. The only thing the users are deciding is what they put into it. And you want to call this decentralization?

Second, we need to talk about privacy.

He offers it like it’s a feature. He says that Meta will build a fully private mode where even Meta can’t see your information. He points to things like WhatsApp, which scrambles messages so that only the sender and receiver can read them.

A mode is something you switch on. Saying that there’s going to be a private mode is saying that the ordinary condition is not private. Encrypting a message in transit is a different problem from an AI assistant that has to process your medical history, tastes, relationships, and interests to personalize your experience and sell ads back to you. He puts the two things side by side and doesn’t explain how the second one works.

The third problem runs through every section in his essay. Nowhere in the essay is there a condition that would count as evidence against his thesis. The single exception he makes is biological weapons, where he says that if harmful examples start appearing, the strategy should be adjusted. WTF do you mean by that Mark? Are you telling me you can’t think of a single other example where your strategy might need to be adjusted?

This is just a press release. It’s a document with no failure condition. You can’t grade him on whether or not he’s wrong 10 years from now.

And I guarantee you he had his lawyers look very closely at this document before he published it for just that reason.

Where he is right

I just told you that nowhere in the essay does he talk about where things could go wrong. So to avoid hypocrisy I should probably tell you where I think he’s right.

His central criticism of AI competitors is valid.

Most of the AI safety world argues that the technology is so powerful that only a small number of carefully governed labs should be allowed to build the most advanced systems.

Zuckerberg points out two problems with that. The first is that the same people who say AI will end human relevance are, in many cases, the same people that are racing to build it. The second is that, to follow the proposed solution, we’d have to concentrate an enormous amount of power into very few institutions and then just hope they use it well. He rightly points out that this is a bet humanity has placed many times and has lost many times.

He also rightly points out that there is no single set of values that a machine could be tuned to that would satisfy everyone. People want incompatible things, so any system built to serve one moral framework is going to fail people who want another.

Both of these points are correct and, in my opinion, by far the strongest parts in the essay. But they are also arguments about what is wrong with someone else’s plan. They don’t establish that his plan works. They establish that a competing plan doesn’t.

Where the big gaps are

Access is not power. The essay offers two possibilities. Either a few institutions control superintelligence, or everyone does. What he’s actually proposing is a third arrangement that doesn’t fit either description. He’s saying Meta builds and owns the system, and everyone gets to use it.

These are three different things. Renting a tool from a company is not the same as making the tool or owning the tool yourself.

If Meta changes the terms, raises prices, alters what the agent will and will not do, or shuts a product down, the user doesn’t have any vote. There is no appeal process. Zuckerberg never distinguishes between spreading the use of the technology and spreading control over it. His entire argument depends on the two being the same thing and he never once defends the equivalence.

The auction gives it away. He promises a free version for everyone, and then he says that anyone who wants more “compute” will get it through a “dynamic auction” that guarantees everyone the “lowest price posssible.”

That’s not what an auction is. An auction is a method for deciding who gets something that is scarce and the one willing to pay the most gets it. An auction is the opposite of a guarantee of low prices.

Earlier in his essay he wrote that “there will always be a finite amount of compute.” Finite means rationed. Rationing by price means that the people with money get more. More compute means more capability. The free tier is a floor and everything above it scales with wealth, which describes the distribution of power we already have. The only difference is Meta is running the auction, collecting every bid.

The lawyer example does not survive contact with biological weapons.

His best example is one of a super intelligent lawyer. He says, if one person has a super intelligent lawyer they’ll win cases they should lose, but if everyone has one the system gets fairer. That’s kind of true (I could write a whole essay on why it’s not true, but we’ll save that for later). It’s (kind of) true because a courtroom is symmetrical: two sides, same rules, a judge, an appeal.

Biological weapons are not symmetrical. One person who successfully designs a dangerous pathogen doesn’t get canceled out by a billion people who did not. The bad guy needs one success. The defenders need to stop every attempt. Zuckerberg’s answer is that there’s little historical precedent, that bad actors have been able to synthesize harmful compounds for decades without it becoming a major problem. He then says if harmful examples begin to emerge, “we will need to update these processes to keep up with the pace of innovation anyway.”

His policy is essentially to wait for the first casualty, applied to the one category where the first casualty may not be reversible. And then he just moves on.

The cybersecurity argument describes a different technology. He argues that open systems end up more secure because more people can inspect them and fix flaws. That might be broadly true, at least historically, for conventional software, where humans read code, find a bug and ship a patch, and where attackers work at human speed.

But an AI agent looking for vulnerabilities doesn’t work at human speed. It can probe enormous numbers of systems continuously, and most of the world’s systems are old, unmonitored, and rarely updated. Zuckerberg’s answer is that eventually superintelligence will make most of the world’s code provably secure. That claim is what this whole section leans on and he frames it as a matter of time rather than something that needs to be demonstrated first before we can call it evidence.

The self-improvement section concedes the argument. Recursive self-improvement is when an AI redesigns itself to be better, then uses the improved version to improve itself again. Zuckerberg says that recursive self-improvement could lead to AI getting a hundred times more capability out of the same hardware, which would let it outclass everyone else combined, and become exactly the single dominant intelligence his essay was written to prevent.

His solution is that we should encourage Meta to build enough “compute” to stay in the race and trust they will devote most of it to “human values.” He’s asking us to trust him without putting any meaningful constraints on the power we are supposed to trust him with.

Then he writes the most consequential line in his whole essay, “But letting an AI system or anyone else direct its own goals is not inherently harmful by itself, even if its goals are not fully aligned with many people, as long as we maintain a balance of power that favors people overall.”

He permits an AI to pursue its own agenda, a truly separate agenda beyond our human input. His safeguard is that other systems will be pursuing their own agenda too.

Adoption is not proof of alignment. Alignment means getting a system to behave according to someone’s set of values. Zuckerberg argues that if billions of people end up using personal agents, then alignment will by definition have been solved, because people would not adopt agents that worked against them.

“Solving alignment is required to enable billions of people to adopt personal superintelligence agents. But this also implies that if we reach a state where billions of people are using and scrutinizing personal superintelligence agents, then we will have solved alignment to individuals’ interests. This should be sufficient to establish the necessary balance of power and natural checks and balances for a positive and safe future.”

People adopt things that are useful. Usefulness and trustworthiness are not the same thing. The gap between useful and trustworthy is where most of the harm from consumer technology has come from over the last fifteen years. Billions of people adopted the Facebook news feed. But do those people trust Facebook? Facebook is good at holding your attention. It’s useful to consumers because it’s entertaining, sometimes informative, and occasionally social (it’s supposed to be SOCIAL media). But most of all Facebook lasted because it made people money. Trust has very little to do with it.

Checks and balances is a legal term he is using loosely. Constitutional checks work because they are enforced. Courts can void an action. Legislatures can subpoena. Citizens have standing to sue, and there are written rules about who wins when two powers collide.

Competition between AI agents has none of that machinery. There is no court, no standing, no remedy, and no rule about who prevails. What he is describing is a market, and markets check power by offering an alternative to anyone who can afford to switch. He borrows the language of constitutional design to describe consumer choice.

The evidence is thin exactly where his claims are boldest. His central economic claim is that AI will produce more jobs rather than fewer. The evidence he offers is the phrase “recent statistics suggest,” and then attaches no statistics. I send work back every day because journalists forgot to show their work. You’ve got to show your work, Mark.

The one hard figure in the entire essay is a $50k bonus paid to teachers in Richland Parish, Louisiana, funded by tax revenue from a Meta data center. How convenient. It’s not fake. Teachers in that one town really got that bonus. But it’s one town, and he frames it as a model for a national buildout next to a set of commitments about water and energy with a 2030 date on them. He calls it a “community compact.” He does not say what happens if the community’s power bills go up anyway, because (again) there’s no enforcement.

The history is selective. He reaches for the Wright brothers, for Michael Faraday, for the kid in the garage building a personal computer. He says that before industrialization 90% of people farmed, and that technology freed them for better work.

“Advances in technology steadily freed much of humanity to focus less on subsistence and more on the pursuits we choose.”

Do you feel like most of us are thriving because of new technologies? Or are we subsisting? That transition he talks about also included child labor, urban slums, and a generation of workers whose skills were destroyed and who never recovered inside their own lifetimes.

Zuckerberg agrees that adaptation will be challenging. But if anything the historical record he is citing only mildly supports the long-run claim and says nothing at all about the people caught in the middle, which is the group everyone is actually worried about.

Why did he write this?

But maybe Mark Zuckerberg is just naive, or out-of-touch. He’s a billionaire after all. Maybe he means well and just doesn’t get the plight of the average joe. Right?

Okay, fine. Set the whole argument aside and look at what his essay asks the government to do and who benefits most from each ask.

He says open source should not be restricted, which is Meta’s competitive position, because Meta has no business selling private access to protect. Distillation should be protected, meaning training your model on the outputs of someone else’s model, and Meta benefits from learning from its rivals. The restrictions on what material can be used to train models should be loosened, and Meta is a defendant in litigation over the material it used to train its models. Building data centers and power plants should be made faster and easier, and Meta is in the business of building data centers, obviously.

He does want one restriction. He wants export controls on advanced chips to China to continue, and he says they’ve worked. Every other restriction in his essay is described as dangerous centralization. But this one is described as “the right strategic move.” So Zuckerberg wants no restrictions, except where it would cost Meta by creating more foreign competition. How noble.

“Export controls on silicon have been successful for slowing the progress of foreign labs during this critical period, so it is the right strategic move to continue those.”

And then he offers a proposal for government oversight, and it looks very familiar. Rather than a regulatory review on a timeline, he suggests labs voluntarily hand the government early copies of models in training, along with their engineers, “so the government can harden and secure critical systems against new risks.”

Just to be clear, his proposal is that we should trust the corporations for which this technology has already centralized to do the right thing and hand over their trade secrets voluntarily. Did we not just spend a whole section talking about the dangers of unregulated, centralized AI? Isn’t that the whole premise of your essay? Am I losing my fucking mind here or does anyone else feel like we’re running in circles?

Not only does this proposal replace a binding governmental process with a cooperative one, it also puts unreleased models in government hands two sections before he warns about the danger of government surveillance and tyranny, and he never reconciles the two. Who let you publish this? I’d think you could afford a few more sets of eyes prior to publication.

His governance answer is that Meta’s board of directors will approve the safety rules for releasing models. He mentions in the same paragraph that Meta is a founder-controlled company, meaning Zuckerberg holds a class of shares that gives him voting control regardless of how many shares anyone else owns. He raises the objection and never answers it.

If you want to know why he wrote this, what this whole essay was even for, look at what one of the richest people in the world and owner of one of the most powerful companies in the world is suggesting. This essay recommends removing every obstacle in front of Meta, making it harder for its U.S. competitors, and keeping the one obstacle in front of its foreign competition. How free-market-capitalism of you, Mark.

We’ve got to ask ourselves a few questions

The reality is, all of this might happen. When a really rich and powerful dude suggests we make changes that help out the rich and powerful, and that guy likes to have dinner with the President of the United States, it’s a good bet that the powerful dude is going to get what he wants.

So, my questions for you Mr. Zuckerberg, or really anyone at Meta able to make these decisions:

Who sets the legal and safety boundaries that you say the agent will operate inside? And what stops Meta, whose business model depends on collecting my data and selling advertisers access to my attention, from selling the data and attention of an agent that knows everything about me?

What happens to a community whose electricity bills rise after the data center opens, given that the compact is voluntary and has no penalty attached for breaking it?

What evidence would change your mind about biological risk, and how many people have to be harmed before that evidence exists?

If compute is finite and the extra is auctioned, what does the free tier actually get, and how far behind the paid tier does it fall? Don’t you see a problem with that inequality of access? How would you address that?

And the last one, which the essay seems structured to prevent me from asking. Your case rests on the claim that broad distribution keeps power balanced. But Meta controls the model, the machines, the terms, and the auction. What keeps the power balanced if Meta changes its mind?

Zuckerberg has reversed the burden of proof. The question used to be whether a company could show its product was safe before releasing it. He has rewritten this so the question is whether restricting the product would lead to tyranny. That puts his critics on defense and puts his product first.

He asks us to accept that the safest possible future is one where a company we did not choose runs an assistant that knows our health, our money, and our children, and orders our groceries.

He never says who pays for the groceries to be on the list.

Share

Want to help The Minuteman for free? Share it.

Sharing does not cost you anything. And it supports my work & independent journalism!

Share The Minuteman

New here? Sign up for free.

If someone did send you this, thank them and subscribe for free here:

Subscribe now

Subscribe now

We think inflation is bad, but almost nobody can say what it is. The United States has fought it with legal price ceilings, government rationing, a wartime tax that took most of corporate profits, and interest rates high enough to put twelve million people out of work. Every one of those methods worked by sending somebody the bill, and who gets the bill has always been a political choice. This is the cost of our attempts to battle inflation, what worked, what didn’t, and who paid.

Somebody Always Pays to Stop Inflation

Shit’s expensive.

Prices are climbing, wages aren’t chasing them as well as they used to, and the people in charge of fixing this problem keep reaching for the same response. Raise interest rates, make borrowing cost more. Wait for people to stop spending.

We treat that response as the natural and singular fix. It seems like everyone has always been talking about inflation. I was reading an article this week about inflation and I realized that I had no idea what it was. And then I had an even more concerning realization: I don’t think anyone knows what inflation is. Or how to deal with it.

So I went looking. What I found is that for most of the last century the United States fought inflation in very different ways, and the record of what worked and what failed looks nothing like what the current debate is willing to admit.

Raising interest rates does not always work. When it does work it can put millions of people out of jobs. Every method this country has tried, from the 1890s to today, has forced the same decision that nobody running the economy likes to say out loud. Someone has to absorb the cost, and the choice of who absorbs it has always been political.

This is the history of that choice.

What inflation actually is

Inflation is a sustained rise in the general price level. Two words to focus on:

General means most things, not one thing. If a drought wrecks the coffee harvest and coffee doubles, coffee got more expensive. That isn’t inflation. That’s a coffee problem. Inflation is when your rent, your groceries, your insurance, your haircut, and your electric bill all climb at once, which tells you something has gone wrong with the money rather than with the coffee.

Sustained means it keeps happening. A one-time jump in the price level is different from a process that grinds forward month after month.

Where does it come from? Three places.

The first is too much demand. People have more money than there are goods to buy, so sellers raise prices because they can. Economists call this demand-pull. This is the version most people picture, and it’s the one interest rates were designed to fight.

The second is too little supply. A war closes a shipping lane. A pandemic strands containers in the wrong ports. A tariff adds a tax at the border. The goods get scarcer or costlier to produce, and prices rise even though nobody is spending more. Economists call this cost-push. Interest rates were not designed to fight it.

The third is expectations. If everyone believes prices will rise 6% next year, workers ask for 6% raises, landlords write 6% into leases, and suppliers pencil in 6% on next year’s contracts. The belief produces the outcome. Once that takes hold, inflation stops needing a cause and starts running on that expectation. This is the part central bankers lose sleep over, and it’s why they talk endlessly about credibility.

When inflation falls, prices do not fall. Inflation is a rate of change. If prices rose 9% one year and 3% the next, the news reports that inflation came down, and it did. The prices are still up 12% from where they started.

Getting inflation back to normal returns the speed to normal. It doesn’t return the level. Prices coming back down has a name, deflation, and the last time this country got a serious dose of it, people organized a national political movement to make it stop.

So when a politician says inflation is down and a person at the grocery store says everything is too expensive, both of them are telling the truth. They’re just describing different numbers.

Share

How we count it, and why it’s political

You cannot fight a number you cannot measure, so before anyone could fight inflation, someone had to decide how to measure it. Every one of those decisions was made by a person.

The Consumer Price Index (CPI) works by pricing a basket of goods, weighted by how much the average household spends on each. That word “average” is already doing something. A household that rents and drives has a different basket than one that owns and takes the train, and the “average” gives you one answer for both.

Take housing. The largest single piece of the CPI is a category called owners’ equivalent rent, which by itself runs to about a quarter of the basket. Rent and owners’ equivalent rent together come to nearly a third. Owners’ equivalent rent is the answer to a survey question: homeowners are asked what they think their house would rent for. It is an estimate of a transaction that never happens, standing in for the housing cost of people who own their homes. It also lags actual market rents by a year or more, which is why in 2022 the official housing numbers were still climbing while the rental market had already turned. A third of the index was reporting last year’s conditions.

Then there’s the difference between headline and core inflation. Core inflation strips out food and energy on the theory that those prices bounce around too much to reveal the underlying trend. Officials mostly talk about core inflation. But households mostly buy food and energy. Deciding which prices count as a signal and which count as noise is a judgment call.

Those judgment calls move real money. In 1996 a Senate-appointed commission led by the economist Michael Boskin concluded that the CPI overstated the true rise in the cost of living by about 1.1 percentage points a year, and the methods changed. Social Security raises are tied to the index, so a technical fix inside a statistical agency trimmed the checks of every retiree in the country, every year after that, and no member of Congress had to vote for it.

But how does the Fed decide when to act? The Federal Reserve tries to keep inflation at 2% a year. That figure is the reference point for the entire American economy: when inflation runs above 2%, the Fed may raise interest rates to slow things down, and slowing things down is how people lose jobs. So where did 2% come from?

New Zealand. In 1989 its parliament told the country’s central bank to get inflation between zero and 2%, and the bank’s governor settled on a range around 2%. It worked. Prices there came down and stayed down. Other countries copied it, and the Federal Reserve made 2% official American policy in 2012. There was no study behind the original figure and no calculation proving 2% is the right amount of inflation for a modern economy. It was a round number that sounded reasonable and happened to work in a country of three million people.

The index is a construction, and every construction has a builder.

Why it feels worse than it usually is

Ask people what they hate about inflation and they don’t describe a rate of change. The Harvard economist Stefanie Stantcheva ran open-ended surveys on this in 2024 and found that about 80% of respondents believe prices rise faster than wages. They describe inflation as something that takes purchasing power away from them with no offsetting gain.

Part of that is real and part of it is how attention works. You see the price of gas on a forty-foot sign four times a week. You see grocery prices every seven days. You see your rent once a month and your raise once a year, and when the raise comes you experience it as something you earned, because you did. The price increase feels like something done to you and the wage increase feels like something you accomplished. Both hit the same paycheck. Only one of them registers as the economy.

Stantcheva also found that people split hard along party lines on who to blame, which tells you that inflation is one of the few economic conditions everybody experiences directly and nobody experiences neutrally. That combination, universal exposure and zero agreement about cause, is what makes it the most reliably destabilizing thing that can happen to a government.

Subscribe now

The historical record of inflation

Before anyone tried to stop inflation, Americans spent thirty years fighting about the opposite problem.

In the last quarter of the nineteenth century the country was on the gold standard, the money supply grew slower than the economy, and prices fell year after year. For anyone holding cash or a bond, this was a windfall. Their dollars bought more every year while they did nothing.

For farmers it was ruin. A farmer borrowed against the harvest, and the loan came due in dollars. When crop prices fell, the same debt required more crops to repay. Corn sold for 63 cents a bushel in 1881, which made a $1,000 loan worth about 1,587 bushels of corn. By 1886 corn was 36 cents, and that same $1,000 took 2,777 bushels. The farmer had borrowed one amount and now owed nearly twice as much corn, without anyone changing the terms of the loan. Falling prices had rewritten the contract in the creditor’s favor.

The response was a mass political movement demanding that the government mint silver, expand the money supply, and let a little inflation ease the debts. It reached the Democratic convention in Chicago in July 1896, where William Jennings Bryan closed his case with the line that named the whole conflict: you shall not crucify mankind upon a cross of gold.

Bryan lost. Gold discoveries in the 1890s eventually delivered the mild inflation the farmers had wanted, and the movement got filed away in the textbooks under currency reform.

What it actually was is the first clean American example of the pattern that runs through everything after it. The price level is never neutral. Every time it moves, it moves wealth from one group to another, and the direction it moves is a political outcome. In the 1890s the country chose creditors over debtors and called it sound money. The farmers who lost their land understood exactly what had been decided and by whom.

The other lesson from that era is why nobody targets zero inflation today. Falling prices sound like a gift until you owe money, and most households, businesses, and governments owe money. When prices fall, debts get heavier, borrowers cut spending to service them, less spending means lower prices, and the thing feeds itself. Japan spent the better part of two decades stuck in a mild version of this. The 1930s were the severe version.

Aiming for a small positive rate keeps a country away from that spiral, and it leaves a central bank somewhere to cut interest rates to in a downturn before it hits zero and runs out of tools.

The agency that beat wartime inflation

In the 1940s, in neighborhoods across New York and much of the country, groups of women walked into shops with printed lists and checked the prices on the shelves against the legal ceiling.

In 1942 the federal government had asked ordinary people to police prices. The Office of Price Administration, created by Franklin Roosevelt in 1941 and given real authority by Congress early the next year, set legal ceilings on the price of nearly everything Americans bought.

Enforcing those ceilings across hundreds of thousands of stores took more inspectors than Congress would pay for, so the OPA printed price lists and recruited volunteers, many of them women, to check the posted prices in neighborhood shops against the law.

Opponents in Congress and business called them “snoopers,” and the economist John Kenneth Galbraith, who ran the OPA’s price division, felt he had to promise there would be no “Gestapo of volunteer housewives.”

The reason for all this was war. Factories were building tanks and planes instead of cars and refrigerators, which shrank the supply of goods for ordinary people at the same time that war jobs put more money in their pockets. More money chasing fewer goods pushed prices up.

Britain had already worked out the theory. In 1940 John Maynard Keynes published a short book called How to Pay for the War, and its argument was that a wartime economy has a fixed problem: the government needs to move resources from civilian use to military use, and the only question is how.

If you let prices rise, the resources still move, and the cost falls on wage earners whose paychecks buy less.

Keynes proposed taking the money out of circulation directly, through compulsory saving and taxation, so that the transfer happened on purpose and on a schedule the government could see. He designed the mechanism specifically so that workers would not absorb the cost through inflation without anyone admitting that was the plan.

The American program ran on the same logic.

The OPA held prices down by force, freezing them on close to 90% of retail food prices, and it paired the freeze with rationing of scarce goods like sugar, coffee, gasoline, tires, and meat.

Congress taxed excess corporate profits at rates that climbed into the ninetieth percentile, on the theory that a company should not get rich from a war other people were dying in. And the Treasury sold war bonds hard, taking cash out of household hands and holding it out of circulation for the duration.

Every piece of that program billed somebody, and the bills were clear for anyone to see. Consumers gave up choice. Producers gave up the price they could have charged. Corporations gave up profit. Savers accepted low returns on bonds. The government was honest about who was paying and what for.

Rationing meant shortages and sacrifice, and it came with a black market and a slow decline in product quality that people at the time called “skimpflation.” It also meant a scarce cut of meat didn’t go to whoever could pay the most. A family of modest means, with a book of ration stamps, still got their share.

The historian Meg Jacobs has described the controlled economy as a high point for ordinary consumers, a system that protected their purchasing power and gave lower-income families reliable access to goods the open market would have priced them out of.

Measured inflation stayed low the entire time the system held.

Share The Minuteman

How the controls died

The system held until the war ended and the pressure to lift it grew too strong to resist.

In mid-1946 the law authorizing price controls lapsed. Prices jumped and the cost of meat roughly doubled.

Truman and Congress put controls back on meat, and the cattle ranchers and meat packers responded by holding their animals off the market.

Meat production fell more than 80% in a matter of weeks. Butcher counters were bare, and the shortage was blamed on the very controls the withholding was designed to discredit.

In October, Truman gave up and lifted the controls.

The politics came first and the inflation came right behind. Republicans ran the 1946 midterms on a two-word slogan, “Had enough,” and swept both houses of Congress.

Fifty-four economists had warned in the New York Times that ending controls while demand still outran supply would set off a surge in prices. By the spring of 1947 inflation had reached about 20%, the second-highest twelve-month reading in the modern price record.

The controls worked while they were enforced, and the people who dismantled them were the ones positioned to profit from higher prices.

The war time debt we paid

Something else was happening in those same years, and almost nobody at the time understood it as a policy at all.

The war left the United States owing more money than the entire country produced in a year. Government debt gets measured against the size of the economy, because a debt is only as heavy as the income available to pay it, and by that measure federal debt stood at 106% of the economy in 1946. By 1974 it was down to 23%.

The story usually told about that drop is that America grew out of the problem. The economy got so much bigger that the old war debt shrank next to it.

Growth did some of the work. But two other things did more. One of them was ordinary: in a number of those years the government took in more than it spent and paid debt down on purpose. The other one nobody had to vote for.

When you buy a Treasury bond, you are lending money to the federal government and it pays you interest. From 1942 through early 1951 the Federal Reserve had agreed to hold the interest rate on long-term Treasury bonds at 2.5%, so the Treasury could borrow cheaply to fight the war, and it kept that promise for six years after the war ended.

Holding a bond’s interest rate down means buying bonds whenever demand for them slips, and the Fed buys bonds with money it creates. The arrangement pushed prices up while holding returns down.

Lend the government money at 2.5% in a year when prices rise 20%, as they did in 1947, and the interest you collect comes nowhere near covering what your dollars lost. You get your money back and it buys less than what you handed over. You have paid the government for the service of lending to it.

Economists call this financial repression. An IMF study asked what the postwar years would have looked like without it. Take away the budget surpluses and the suppressed interest rates, leave growth to do the job alone, and the debt falls from 106% to about 74% by 1974 rather than to 23%.

The distance between those two figures is the part of the war debt that got paid off by ordinary people holding bonds and savings accounts.

By early 1951 inflation was running above 20% at an annualized rate and the arrangement broke. Fed officials argued that defending the peg had turned the central bank into, in Marriner Eccles’s phrase, an engine of inflation. In March, the Treasury and the Fed reached the Accord that released the Fed from the peg and let it set interest rates on its own. Every argument about central bank independence for the next seventy-five years traces back to that document.

Before the Accord, the cost of the debt fell on savers, and it fell on them invisibly. After it, the Fed could raise rates to protect savers, and the cost of doing that would fall on borrowers and workers.

The Accord changed who paid the bill.

The half measure

The country tried controls one more time, and that attempt is the one most people remember because it failed.

As a young lawyer in 1942, Richard Nixon spent a few unhappy months at the OPA before leaving for the Navy. For years afterward he insisted controls could never work in peacetime. Then on August 15, 1971, he went on television and reversed himself.

The speech is remembered for the 90-day freeze on wages and prices, the only peacetime controls in American history. It contained two other things. Nixon suspended the convertibility of the dollar into gold, ending the Bretton Woods system that had organized global finance since 1944. And he imposed a 10% surcharge on imports.

A wage-price freeze and a global tariff, announced in the same speech, as parts of the same anti-inflation program. How do you think it turned out?

Nixon froze prices and did almost nothing else. He built no rationing system and stood up no agency with the reach of the wartime OPA. He asked businesses and workers to hold the line on their own (no kitchen Gestapo to hold you accountable).

Demand kept rising underneath the freeze with nothing to check it, and when the controls came off in stages over the next three years, prices climbed fast in a burst of catch-up inflation that erased the gains.

The freeze failed because Nixon kept the easy part and skipped the parts that made the wartime program work: the enforcement, the fiscal correction, and the rationing of what was scarce.

There was a second failure running alongside it, inside the Federal Reserve. Nixon wanted easy money going into his reelection, and he made sure his Fed chairman knew it.

The Nixon tapes, examined by the economist Burton Abrams in a 2006 study, record a sustained pressure campaign on Arthur Burns, including the February 1972 line, “I really don’t care what you do in [or after] April.”

Burns ran expansionary policy through the election. Whether he did it out of conviction or out of fear is still debated. The inflation that followed is not.

The decade that built the doctrine

In 1973 the OPEC oil embargo quadrupled the price of crude. In 1979 the Iranian revolution did it again. Oil sits underneath the price of nearly everything, so when oil moves, the whole index moves, and it moves for reasons that have nothing to do with American consumers spending too much.

What the 1970s produced was inflation and unemployment at the same time. That combination was supposed to be impossible.

The governing framework of the era, the Phillips curve, held that inflation and unemployment traded off against each other: get unemployment down and prices rise, let unemployment rise and prices cool. A period with both high at once had no place in the model.

Milton Friedman and Edmund Phelps had predicted the breakdown in 1968, arguing that the tradeoff only worked as long as people were surprised. Once workers and firms expected inflation and built it into their wages and contracts, you would get the inflation without the employment gain.

The 1970s proved them right, the Phillips curve lost its authority, and Friedman’s alternative explanation moved to the center of the profession: inflation is always and everywhere a monetary phenomenon, a matter of the money supply growing faster than the economy.

That idea, and the failure of Nixon’s freeze, and the memory of shortages, combined into a conclusion that has governed American policy ever since. Managing prices directly is a fantasy. Control the money and the prices take care of themselves.

Jimmy Carter tried one more direct intervention on the way out. In March 1980 he invoked emergency authority to impose controls on consumer credit, restricting credit cards and consumer loans rather than raising their price.

It worked far better than anyone intended. Consumer borrowing collapsed, and in the two months that followed, the unemployment rate rose more than 1.5 percentage points, which stood as the sharpest two-month increase in American history until the COVID shutdown.

The administration reversed the program within months. The recession it triggered probably cost Carter the election.

Nobody has tried credit controls since. When you restrict borrowing directly instead of making it expensive, the effect arrives in weeks rather than months, and by the time the damage shows up in the data it has already happened.

Share The Minuteman

The switch to interest rates

Carter’s other decision that year outlasted him by four decades.

He appointed Paul Volcker to run the Federal Reserve in 1979. Volcker believed the way to break inflation was to force the whole economy to contract until price increases became impossible to pass along. Carter understood that meant a recession and heavy job losses, and appointed him anyway.

On the evening of Saturday, October 6, 1979, Volcker announced that the Fed would stop managing interest rates day to day. It would control the money supply instead, by limiting how much cash it let banks hold, and it would let interest rates go wherever that pushed them. Rates were no longer something the Fed set but something the market would discover.

The market discovered a lot. The federal funds rate was 11.4% on October 5. Within weeks it was 13.8%, and by April 1980 it hit 17.6%. Mortgage rates followed. Borrowing nearly stopped.

Ronald Reagan ran for the presidency in 1980 on a promise to “make America great again” (does that sound familiar?), a phrase he used to accept the Republican nomination that summer.

For those without skills, we’ll find a way to help them get skills. For those without job opportunities, we’ll stimulate new opportunities, particularly in the inner cities where they live. For those who have abandoned hope, we’ll restore hope and we’ll welcome them into a great national crusade to make America great again!

At his inauguration he delivered the sentence that organized the next forty years of economic policy. “Government is not the solution to our problem; government is the problem.”

The governing idea was that markets correct themselves and that public management of the economy causes more harm than it prevents.

Volcker put the theory into practice from inside the Fed. Factories and construction sites shut down. Farmers drove tractors to Washington and circled the Federal Reserve building. Homebuilders mailed him two-by-fours from houses they couldn’t sell. Unemployment reached 10.8% at the end of 1982, the highest level since the Great Depression, which put roughly twelve million people out of work.

It also brought inflation down from about 13% to under 4%. The strong recovery that followed looked like proof of the whole theory. Interest rates had brought inflation down, the economy had come back, and the belief that government should stay out of prices settled into common sense.

The economics profession has a technical term for what happened to those twelve million people. It’s called the sacrifice ratio, and it measures how many percentage points of unemployment, held for how long, are required to bring inflation down by one point. The discipline built a formal accounting unit for human cost and then talked about it in decimals.

The cost was also not distributed the way the phrase “the economy” suggests. The Black unemployment rate in the United States runs at roughly twice the white rate, and that ratio holds through booms and busts. When a central bank decides to cool the labor market by some amount, arithmetic decides whose labor market cools first and cools hardest.

Volcker’s defenders make a serious argument that a worse recession was coming anyway if inflation had run another five years, and they may be right. Every method on this list has a defense that turns on what the alternative would have been. What the defense cannot do is make the cost disappear. It happened, it happened to specific people, and those people were not the ones setting prices.

The Fed becomes the answer

Alan Greenspan took over the Fed in 1987 and presided through the 1990s, a stretch of steady growth and mild inflation that commentators called the Goldilocks Economy: not too hot, not too cold.

In 1994 and 1995 he raised rates enough to cool the economy without tipping it into a recession, a soft landing that few Fed chairs before or since have managed.

Interest rates were now the accepted answer to inflation, and the chair of the Fed was treated as the one person who could keep prices in line.

Congressional hearings on inflation became hearings about what the chairman thought. Presidents stopped proposing price policy because there was no longer such a thing.

When the financial system collapsed in 2007 and 2008, the Fed cut interest rates to nearly zero to stop the fall and then bought trillions of dollars in bonds, creating the money to do it. The standard theory held that this much money, this cheap, for this long, would push inflation up.

It never happened. Rates stayed near zero for most of a decade and inflation stayed quiet, so quiet that the Fed spent years worried prices were rising too slowly.

Economists called it the missing inflation puzzle. The country had spent thirty years treating interest rates and the money supply as the dependable controls on inflation, and here was a long stretch where the relationship the whole approach rested on did not appear.

What other countries did while America stopped trying

America settled on one tool after 1980. The rest of the world kept experimenting, and the results complicate our prevailing theory.

Israel in 1985 had inflation running above 400% a year. In July the government enacted a stabilization program that did everything at once. It froze wages, prices, and the value of the currency together, cut the budget deficit sharply, and raised interest rates. The freeze was negotiated with the Histadrut, the national labor federation, which agreed to hold wages in exchange for the price controls and a commitment on employment. Within less than two years inflation was under 20%, and it did not come roaring back when the controls lifted.

Compare that against Nixon. Israel paired the freeze with the fiscal correction that removed the underlying pressure, and it got the buy-in of the people whose wages were being frozen. Nixon announced the freeze and skipped the rest.

Brazil in 1994 solved a harder version of the problem with a genuinely strange piece of engineering. Inflation had been running in the thousands of percent for years, and the trouble was that everyone repriced constantly because everyone expected everyone else to. The Plano Real introduced a parallel unit of account called the URV, pegged to the dollar. For four months, Brazilians kept paying in the old currency but quoted prices, wages, and contracts in URV. Once the whole economy had voluntarily repriced into a stable unit, the government converted the URV into a new currency, the real, at one to one. Monthly inflation went from 46.6% in June to 7.8% in July to 1.9% in August. No mass unemployment.

Argentina ran the opposite experiment recently. Javier Milei took office in December 2023 with inflation above 200% and applied straight shock therapy: a currency devaluation of more than 50%, deep cuts to fuel, transport, and energy subsidies, and mass public-sector layoffs. Inflation came down hard. So did everything else. Poverty rose from 41.7% in the second half of 2023 to 52.9% in the first half of 2024, which is 15.7 million people. It then fell back to around 31.6% by mid-2025 as inflation moderated and real wages recovered.

Argentina is the clearest illustration available. The stabilization worked. It worked by sending roughly eleven million additional people into poverty for a year and a half. It worked, and it sucked.

And here we are again

The quiet ended in 2022.

Inflation surged worldwide after the pandemic and peaked in the United States at 9.1% in June, the largest one-year jump in four decades. It happened in countries that had spent heavily during the pandemic and in countries that had not, which is itself evidence about the cause.

Jerome Powell’s Fed answered the only way the institution knew how, raising interest rates eleven times between March 2022 and July 2023.

In June 2022, Larry Summers, former Treasury Secretary and the most credentialed voice arguing that the Fed had to move harder and faster, told an audience in London that “we need five years of unemployment above 5% to contain inflation. In other words, we need two years of 7.5% unemployment, or five years of 6% unemployment, or one year of 10% unemployment.” Unemployment at the time was 3.6%. He was describing something like seventeen million people out of work as the price of containing inflation.

It didn’t happen.

Inflation fell most of the way toward target while unemployment stayed near historic lows. Whether the rate increases did the work, or whether healing supply chains and settling energy markets did most of it, economists have not settled.

The Federal Reserve’s own Christopher Waller had argued in 2022 that employers would respond to a slowdown by cancelling job openings before they started firing people, so the labor market could cool without unemployment climbing. That is roughly what occurred.

Something else happened that complicates the simple story about who inflation hurts. David Autor, Arindrajit Dube, and Annie McGrew documented that the tight labor market of 2021 through 2023 produced unusually fast wage growth at the bottom of the distribution, enough to reverse about a third of four decades of growth in wage inequality. Low-wage workers got hit hardest by rising prices and gained the most in wages, at the same time.

Then in early 2026 prices started climbing again, and this time the causes were more obvious.

The first was tariffs.

On February 20, 2026, the Supreme Court ruled 6-3 in Learning Resources, Inc. v. Trump that the International Emergency Economic Powers Act does not give a president the authority to impose tariffs, striking down the entire 2025 tariff structure and opening the door to as much as $175 billion in refunds. Within hours the administration invoked a different statute, Section 122 of the Trade Act of 1974, and imposed a 10% surcharge on all imports effective February 24. It was the first use of Section 122 in the statute’s history.

Whatever the legal authority, the economics are the same. A tariff is a tax collected at the border and paid by whoever buys the thing. The Yale Budget Lab estimates the current structure costs the average household several hundred dollars a year, and roughly double that if the replacement duties now under consideration take effect. The burden falls hardest on households that spend most of what they earn on goods, which is to say the households with the least room to absorb it.

The second cause was war.

American and Israeli strikes on Iran began on February 28, and the fighting disrupted traffic through the Strait of Hormuz, the chokepoint for about a quarter of the world’s seaborne oil. Energy prices climbed with it.

A ceasefire in April reopened the strait and Brent crude fell back below $100, and the June CPI report showed the effect: energy prices dropped 5.7% in the month, the biggest monthly decline since April 2020, pulling headline inflation down to 3.5% from 4.2% in May. Energy is still up 15.7% over the year. The conflict resumed in July.

Households are living through what economists are calling an affordability crisis. (This is the cost of being alive in America.)

Both of these causes are supply problems.

A tariff raises the cost of goods at the border. A war removes oil from the market. Raising interest rates reaches neither one.

Higher borrowing costs cannot reopen a shipping lane or repeal a tariff. They can only shrink demand by slowing the entire economy, which does nothing about the actual source of the price increases and once again asks workers to carry the cost.

Into this the administration has installed a new Fed chair, Kevin Warsh, confirmed in May 2026 by a Senate vote of 54 to 45.

Warsh went before Congress on July 14 and called sixty-three consecutive months of above-target inflation an “unfair burden” on American households, promised a “regime change” in monetary policy, and announced five internal task forces, one of them on the inflation framework itself. He has not said what the new framework will be.

The question is what in the historical record suggests that a demand tool, under a new name, will cure an inflation caused by a tariff and a war.

The record does not supply an answer, because interest rates were never built for this kind of inflation. The only thing a demand tool can do about a supply shock is make the rest of the economy small enough that the shrinking offsets the missing goods.

That is an option. But it has a price, and the price is measured in jobs.

Subscribe now

What the record says actually works

Inflation caused by too much demand and inflation caused by too little supply are different problems that call for different cures, and the current debate keeps skipping past the difference.

When people have more money than there are goods to buy, slowing spending helps, and raising interest rates is one way to slow spending.

When the supply of goods shrinks because of a tariff, a war, or a broken supply chain, cutting demand does not put the missing goods back. It slows the whole economy and spreads the cost more widely.

The 2026 inflation is largely a supply story, and that points to a different set of responses.

The first: Stop causing it.

The tariffs are a self-inflicted price increase, a policy the government chose and can reverse, and the current version comes with an expiration date already written into the law.

The second has a serious intellectual foundation and a track record that did not exist five years ago.

The economist Isabella Weber argued in 2021 for targeted, temporary controls on the specific prices that push inflation through the rest of the economy, prices like energy and key foods, in place of the blanket freeze Nixon tried.

The reaction was brutal. A New Yorker profile recorded that her Guardian essay had briefly made her the most hated woman in economics.

But Europe did exactly that.

Facing the energy shock after Russia’s invasion of Ukraine, Spain and Portugal won permission from Brussels for what became known as the Iberian exception: a cap on the price of natural gas used to generate electricity, with producers compensated for the difference. Wholesale electricity prices in the Iberian market fell an estimated 35% relative to what they would have been.

Germany ran a gas price brake. France capped regulated electricity tariffs and had its state-controlled utility absorb the difference, which is the most direct answer to the question of who pays that any government gave during that period. It was the French treasury, and by extension the French taxpayer, and everyone could see it.

The third response comes straight from the World War II record.

When a good is truly scarce, rationing shares it on some basis other than wealth, so that the people with the most money do not take all of it while everyone else goes without.

Nobody enjoys this. It is also the only method on the list that distributes the shortage on purpose instead of letting the price sort out who goes without.

There is a fourth argument, about corporate pricing power, and it has not been settled.

The Kansas City Fed found that rising business markups accounted for more than half of 2021’s inflation. The San Francisco Fed found that markups across the economy stayed roughly flat and were not a main driver. The studies disagree because measuring a markup requires knowing a firm’s true marginal cost, which nobody outside the firm knows, and because profits in the national accounts are a residual, the number left over after everything else is counted. A profit spike can be a symptom of scarcity as easily as a cause of it.

Weber’s own version of the argument is narrower than the word “greedflation” suggests. Her claim is that a bottleneck hands firms in the affected sector a temporary ability to raise prices beyond their cost increase, along with a public signal that everyone else is raising prices too, so nobody loses customers by going first.

The profit share did spike in 2021 and 2022 and then came back down, which fits the temporary-power story better than it fits a story about permanent corporate character.

The meat packers of 1946 established that firms will withhold supply and let prices rise when they have the power to do it and a political motive to prove a point. That history stays relevant. But it does not settle a question about 2021 that the data itself cannot settle.

The limit on all of this is Nixon.

Price controls fail when a government freezes prices and skips the enforcement, the fiscal correction, and the fair distribution that make a freeze hold. Israel proved that the full package can work in a modern economy. Nixon proved that the announcement alone does nothing.

The methods that worked all asked something of people with power and required the government to do the actual labor of governing. That, more than any question of whether they succeeded, is why they were dropped.

Who decides

Somebody chooses who pays. Here is who.

Interest rates in the United States are set by the Federal Open Market Committee: seven governors appointed by the president and confirmed by the Senate, plus five of the twelve regional Federal Reserve Bank presidents, who are selected by boards that include representatives of the banks in their district.

None of them face the voters. That insulation is intentional. The Burns episode is the case for it. A central bank that answers to whoever is running for reelection produces the 1970s.

Independence is also a distributional choice. It protects the fight against inflation from politics. It does not protect employment from the fight against inflation.

The Humphrey-Hawkins Act of 1978 gave the Fed a dual mandate, maximum employment and stable prices, with no instruction about how to weigh them when they conflict. The weighting is left to the committee. Which means that engineering a recession to bring prices down is a choice the institution is empowered to make, not a technical necessity it is forced into.

Banks are required to keep a pile of cash at the Federal Reserve, and since 2022 the Fed has paid them interest on it, at rates that rose with every hike, while the older bonds the Fed itself owns pay much less. The result is that the Fed has been operating at a loss.

It normally sends its profits to the Treasury. It has sent nothing since 2022, the first interruption since 1934, and by the end of the third quarter of 2025 the accumulated shortfall stood at $243 billion. The Congressional Budget Office projects it will take until at least 2030 to work off.

That is public revenue that isn’t arriving, as a byproduct of the anti-inflation program, flowing to the banking system.

You can argue it’s the necessary cost of the operating framework. Either way, it’s a bill someone at sometime has to pay.

The ledger

For the last eighty years, every method of fighting inflation we’ve used sends the bill to somebody. There has never been an option where no one has to pay the price.

Raise interest rates and the bill goes to people who lose their jobs, to small businesses that can’t refinance, to first-time homebuyers priced out of a mortgage, and to developing countries whose dollar debts get more expensive overnight.

Freeze prices and the bill goes to sellers, who lose the price they could have charged, and then to consumers through shortages and lines, and the people with the least flexible schedules wait in those lines the longest.

Ration and the bill goes to everyone, deliberately, in small even pieces, with a side payment to whoever runs the black market.

Impose tariffs and the bill goes to consumers and to the firms that import what they need to make things.

Let the inflation run and the bill goes to lenders, savers, anyone holding cash, and anyone on a pension that doesn’t rise with prices, while borrowers get a discount on their debts. That is what happened between 1946 and 1974, and it retired a war.

Cut spending to cool the economy and the bill goes to whoever loses the programs we cut.

Tax excess profits and the bill goes to shareholders.

Cut the value of the currency against other currencies and the bill goes to importers and to wage earners, whose paychecks buy less of everything made abroad.

And if prices fall instead of rise, the bill goes to debtors and workers, which is what the farmers of the 1890s were trying to tell everybody.

The country’s default answer since 1980 has been the first one, and the first one is the only method on that list that works specifically by making people unemployed. We have spent forty years calling that the responsible choice and calling the alternatives radical.

Reagan’s sentence still governs this debate. The claim that government is the problem and markets fix themselves has guided one political party for two generations, and that party’s own behavior no longer matches it.

The tax law it passed in 2025 by a republican majority is projected by the Congressional Budget Office to add $4.1 trillion to the national debt over ten years, and its tariffs raise a tax that falls hardest on working families. The party that once defined itself by balanced budgets is running enormous deficits to cut taxes at the top while raising prices at the bottom.

Meanwhile the methods with the strongest record of stopping inflation without mass unemployment, targeted management of essential prices, fair rationing of what is scarce, fiscal correction, and a willingness to look hard at pricing power, are the ones now waved off as socialist. The people who propose them are treated as unserious, while the serious choice is defined as the one that only works by putting people out of work.

Share

What now

The 10% tariff was imposed under Section 122, and Section 122 doesn’t last forever. The statute allows a surcharge for 150 days and no longer unless Congress votes to extend it. That clock ran out on July 24, 2026. A president cannot extend it alone. Theoretically, since it lapsed, that self-inflicted price increase should come off the economy. If it is replaced under a different statute, one without a time constraint, that will also be a choice, made by identifiable people, in public, this year (with midterms coming up).

The Fed’s inflation framework is being rewritten right now by five task forces expected to finish their work by the end of the year. What comes out of that review sets the terms for the next decade.

Warsh serves at the pleasure of nobody, but the governors around him were confirmed by senators who answer to voters, and the framework has no constituency except the people who show up to have an opinion about it.

And tariff authority, the whole tangle of statutes that lets a president tax imports without a vote, belongs to Congress under the Constitution and was handed to the executive by Congress over the course of the twentieth century. The Supreme Court said as much in February. Congress can take it back whenever it decides to.

Every method of fighting inflation answers a single question: who pays.

The war economy of the 1940s answered it by asking almost everyone to give up a little so the scarce things could be shared, and by taxing the people who profited most from the emergency.

The approach the country has used since Reagan answers it by asking the people who lose their jobs to absorb the entire cost while the people who set prices absorb none of it.

We have spent decades treating that as the only responsible answer.

The history says it was a decision, made by particular people at a particular moment, for reasons that had as much to do with politics as with economics. Decisions can be made again.

The affordability crisis in front of us is the occasion to notice that the question was never whether somebody pays. It was always who pays.

Want to help The Minuteman for free? Share it.

Sharing does not cost you anything. And it supports my work & independent journalism!

Share The Minuteman

New here? Sign up for free.

If someone did send you this, thank them and subscribe for free here:

Subscribe now

A note before we start.

I hear a lot of arguments for the privatization of government services.

Most of the time these seem like very good arguments.

And it usually goes something like this: “the government’s bloated and expensive, and corporations are beholden to the laws of supply and demand and competition. Corporations will always do it better and cheaper so we might as well hand it off to the corporation.”

You got to admit that makes a lot of sense. But what happens when it works too well? And what do we sacrifice?

Over the last 50-70 years we have progressively handed over more government services to private corporations. And over the last 50-70 years we have progressively become more and more dependent on a smaller number of massive corporations.

Our own seemingly logical, purely capitalistic ideology has not only degraded our democracy but degraded the ideological engine that powered this hand off in the first place: capitalism.

Under the guise of leaning on the strengths of a free market, we have weakened just that.

But the more I dug, the more I realized I’d had the story backwards.

This isn’t really about greedy corporations capturing the government. Nobody stole anything.

The government handed these capabilities away on purpose, in public, in writing, across fifty years, with both parties holding hands.

The government hired its own replacement.

And it’s happened before. A private company once grew powerful enough to run a subcontinent while the state that depended on it just watched and took its cut. That arrangement ended, but it took eighty-five years and a catastrophe to undo, and it cost far more to reclaim than it ever would have to keep.

This piece is about where we are on that historical clock. I still think it’s early. It’s still ours to reverse for now.

Subscribe now

Everyone calls it corporate capture: a government outmaneuvered by companies that got too rich to regulate. That story runs the sequence backward. The dependence wasn’t seized from the state. The state handed it over, on purpose, in public, in writing, across fifty years, with both parties holding the pen. The capture story has a villain. The truth is worse: nothing was broken. It worked exactly as designed.

In September 2022, a group of Ukrainian drone boats packed with explosives set out across the Black Sea toward the Russian fleet anchored at Sevastopol. As they neared the harbor, the Starlink coverage they needed to complete the run was never there. Ukraine had asked the company to switch on service in the waters off Sevastopol for the strike; the request was refused, and the attack failed.

The person who made that call was not a general, a president, or a defense secretary. He was a private citizen who owned the satellites and had decided the strike was too dangerous to permit. Agreeing to Ukraine’s request, Musk said later, would have made his company “explicitly complicit in a major act of war.”

Set the man aside and look at the arrangement.

One company held a working veto over a battlefield in a war the United States was funding. The easy word for this is capture: a government outmaneuvered by a firm that grew too rich to regulate, the usual traffic of lobbyists and revolving doors and campaign money. That version has the virtue of a villain and the comfort of casting the state as the party that got taken.

It also runs the sequence backward.

The dependency here was not seized from the government. It was handed over by the government, on purpose, in public, in writing, across fifty years, with both parties holding the pen.

The United States helped build the modern launch economy, the internet, and satellite navigation, then wrote policies that moved critical layers of each into private hands.

It still owns and operates some of what it built, GPS above all, but the ability to fly rockets, carry the traffic, and put astronauts in orbit now belongs mostly to companies. What looks like a company escaping the state’s control is the state’s own policy.

There is a rulebook for this, and it is not a metaphor (if you read my stuff enough you’ll learn I love a good metaphor, often to a fault).

Official federal policy, first written down in the 1950s and still on the books, holds that “the Government should not compete” with its own citizens. The Bureau of the Budget began the program in 1955, instructing agencies not to produce for themselves any good or service that private enterprise could supply. Every handoff that follows grows from that single sentence.

So the question, the concern, isn’t about big corporations stealing what used to belong to a government by the people for the people. It is what happens to that government that spends half a century contracting out its own capabilities, and how do you know when the whole thing is going to come crumbling down.

How a state loses faith in itself

The handoff needed a reason before it needed a law, and the reason was a loss of confidence that ran deeper than any single administration.

In the 1970s the American faith in government broke. Vietnam ended the assumption that the state told the truth. Watergate ended the assumption that it obeyed its own rules. Stagflation ended the assumption that its economists understood the machine they were steering, because the reigning theory said inflation and unemployment could not climb together, and then they did.

The institution lost faith in itself first, which is why the retreat that followed was bipartisan. It was not one party’s project. It was a shared conclusion that the government was the problem.

Two economists supplied the intellectual scaffolding.

Milton Friedman argued that markets punish failure the instant it happens while government insulates itself from the consequences of its own mistakes. James Buchanan turned the same skepticism into a discipline, studying the state as self-interested actors rather than selfless public servants, “politics without romance.”

Their case gave privatization something it had lacked: a respectable rationale. Handing a function to the market was no longer a retreat from duty. It was good governance.

Ideology explains why the retreat was permissible. The budget explains why it happened.

As the Cold War wound down, the money that had built the American state’s technical capacity drained away. NASA is the cleanest gauge. At the height of Apollo in the 1960s the agency consumed more than 4 percent of all federal spending. Today it runs near a third of one percent.

As the deficits climbed, a government that could no longer afford to own what it had invented found itself holding a philosophy that told it ownership was a mistake anyway.

The ideology supplied the justification. The deficit supplied the motive. Together they produced a decision that no one ever quite announced: stop building, start buying.

Share The Minuteman

The receipts

An assertion that the government did this on purpose is only as good as the paper behind it, and the paper is extensive. The retreat was legislated line by line, on record, by both parties.

The framework was the Office of Management and Budget’s Circular A-76, the document that carries the “should not compete” language and set up a process for testing whether commercial-type work an agency was doing for itself should instead be opened to private bidders. (That’s the boring stuff, don’t worry we’re done with it.)

In space, Congress wrote the Commercial Space Launch Act of 1984 to license and encourage a private launch industry that did not yet exist. Six years later the Launch Services Purchase Act of 1990 went further and inverted the default, directing NASA to buy launches from commercial providers and reducing the building of its own rockets to a narrow exception. The agency that put men on the Moon was now instructed, by statute, to become a customer.

The same pattern ran through every technology the government had midwifed.

The internet began as a federal network; in 1995 the National Science Foundation decommissioned the NSFNET backbone and turned the wires over to private carriers.

The Global Positioning System, a military program that civilians had been allowed to use for years, delivered only a deliberately degraded public signal until 2000, when the government switched off that degradation and turned precise navigation into a free public utility that private firms could build empires on.

To take stock of what else might be moved out, Congress passed the Federal Activities Inventory Reform Act of 1998, which required every agency to catalog its own commercial-type activities so they could be reviewed for possible outsourcing. The state took an inventory of itself and flagged the parts a private bidder might do instead.

None of this was hidden. It was the announced, celebrated policy of a country that had decided its own hands were the wrong ones for the work.

It worked

The uncomfortable part, for anyone hoping this is a simple story about a mistake, is that the strategy succeeded. The privatizers were right about efficiency, and the numbers are not close.

The Space Shuttle cost roughly 1.6 billion dollars a flight, about 30,000 dollars for every pound it lifted to orbit. A Falcon 9 lists near 67 million, and its first stage flies home and launches again. On cost per pound the private rocket is more than an order of magnitude cheaper than the Shuttle.

The handoff did exactly what it promised. It made access to space cheap.

This is what most people who argue for the privatization of government services lean on during an argument. And they’re right. The market delivered.

The relevant question was never whether privatization would produce a better rocket. It did. The question is what a country loses when the better rocket belongs to someone else.

The hollowing

What it loses is the ability to say no.

When the Shuttle retired in 2011, the United States had no way to put its own astronauts in orbit and would not have one for nine years. In the interval it bought seats from Russia, paying more than 90 million dollars for a single ride by the end. The capability came back in 2020, aboard a SpaceX capsule. It came back belonging to a company.

That is the moment the relationship inverted.

A government that buys a service it can no longer perform is not a customer in any ordinary sense. When there is a line of vendors, the buyer sets the terms. When there is one vendor and no substitute, the vendor does.

The Crimea veto was the inversion made visible: the supplier deciding what the client’s ally could and could not do in a war. Around the same time it happened again in a register no one could miss, when SpaceX wrote to the Pentagon that it could no longer fund Starlink in Ukraine for free and asked the Defense Department to cover a bill approaching 400 million dollars a year.

A private firm had come to invoice the United States for a war. When you are the only applicant, the interview runs the other way.

Subscribe now

The exit that was never there

Defenders of the handoff have an answer to all of this, and it is the strongest thing they can say. If a contractor abuses its position, the market corrects it. A supplier that gouges or defies its customer invites a competitor, and the threat of that competitor disciplines the incumbent. Dependence on one firm is a temporary condition, not a permanent one, because the door is always open.

The answer holds only where the door is real, and here it is an illusion painted on the wall.

Market discipline requires a market, which means a second place the buyer can actually go. Look for that place and it is not there.

Amazon has spent years and billions on Project Kuiper and remains far behind schedule; with the company projecting it would fall well short of its July 2026 interim deployment milestone, the FCC granted conditional relief in June.

NASA has kept a government-directed launch program in the Space Launch System, but built for a different job, deep-space human exploration; it costs around 4 billion dollars a flight and has been judged unaffordable by its own auditors, a measure of how expensive government-directed capacity has become even where Washington chose to keep it.

Boeing built a crew capsule to compete with SpaceX and in 2024 left two astronauts on the space station when the vehicle proved too risky to fly them home; they waited nine months and came back on a SpaceX Dragon.

The competitors exist on paper. But not one of them can do the job when the job is due.

This is the flaw the handoff theory rested on and never tested.

Capacity is not a switch that can be flipped back on. Once the engineers scatter, the factories close, and the institutional memory retires, rebuilding costs far more than keeping would have, and takes years the emergency does not grant. The market can discipline a supplier only while the country retains the option of walking away. Walk far enough, for long enough, and the option is gone. What remains is dependence with the illusion of choice.

Britain did this first

The state has started to notice, and its remedy is the tell. In the past two years the government has taken a roughly 10 percent stake in Intel, an equity position in the rare-earth miner MP Materials that made the Pentagon its largest shareholder, and a golden share in U.S. Steel that lets a president veto the company’s biggest decisions.

After fifty years of shedding ownership, the United States is buying pieces of its industrial base back. The instinct is sound. The execution reveals the trap.

When the Space Force set out to build a resilient, secure military data network in orbit, meant to harden the country’s most sensitive communications, the contract to build its backbone, worth 2.29 billion dollars, went to SpaceX.

The push for resilience deepened the reliance on the single company it most needed insurance against.

A government that once did this itself, and now cannot, ends up managing its dependence instead of ending it. That is a specific historical position, and it has been occupied before.

The East India Company was a private firm that came to run a subcontinent, with its own army, its own revenue, and its own foreign policy, while the British state watched and took its cut. By 1773 the company was insolvent, its Bengal lands mismanaged and its finances brought to a crisis that had it demanding a government loan. Parliament faced the choice every state faces when a company it depends on begins to fail. It did not seize the company or let it collapse. It extended financial support, passed the Regulating Act, and asserted parliamentary supervision over a firm it could no longer afford to treat as merely private. It was, in Britannica’s phrase, the first government intervention in the company’s affairs and the beginning of a takeover that would not conclude for another eighty-five years.

That belongs to the same family of state responses as a golden share and a bailout contract: a state reaching for partial control of a private power it can neither replace nor do without. Britain reached for it in 1773. We are reaching for it now.

Share

What time it is

The value of the parallel is that it puts a clock on the present.

The British takeover of the East India Company was not one act but a long sequence, with two bookends separated by a lifetime.

In 1773 came the rescue and the first major assertion of parliamentary control, the government stepping in to steady a firm it could not afford to lose. Full transfer of the company’s governing powers did not arrive until 1858, and it took a catastrophe to force it.

The Indian Rebellion of 1857, a violent collapse of the company’s authority that no one could ignore, finally moved Parliament to pass the Government of India Act of 1858 and transfer the company’s powers and territories to the crown.

The lesson of the eighty-five years between is not that shed sovereignty never comes back. It is that it comes back only after a public failure large enough to force the choice, and it costs far more to reclaim in 1858 than it would have to keep in 1773.

By that clock we are living in 1773, not 1857. The bailout-and-buy-in phase, the golden shares and the equity stakes and the bigger contract with the same firm. The catastrophic public failure that ends the arrangement has not arrived.

The Crimea veto and the Starlink terminals that emergency crews leaned on during the Los Angeles fires were warnings delivered at no charge, one foreign and one domestic, previews of what a genuine failure would look like. The next one will not be free.

There is a new pressure this timeline did not have, and it points the same direction.

In June 2026, SpaceX went public in the largest offering in history, at a valuation approaching 1.77 trillion dollars, and made Elon Musk the first trillionaire on record. A large share of the offering was steered to ordinary investors, and SpaceX has already entered major indexes, pushing the company into the index funds and retirement portfolios of the citizens whose government depends on it.

Britain’s East India Company had the same feature at its height, its stock spread through the portfolios of the members of Parliament who were supposed to check it. A country becomes slow to reclaim a power its own savings are invested in. The dependence acquires shareholders, and the shareholders acquire a reason to look away.

Return to the drone boats stalled off Sevastopol for want of coverage, because they read different now.

That was not a story about one unpredictable man, and not a scandal about a company that got away with something. It was the design working as written, a wartime capability the government leaned on but did not own, exercised at the discretion of its owner exactly as the arrangement always allowed.

The comfort of the capture story is that it implies a fix: catch the company, close the loopholes, take the power back. The harder truth is that there is no loophole to close, because nothing was broken. The rules were written to produce this, by us, over fifty years, on purpose.

Which is also the reason it is not fixed in place. What one Congress wrote, another can rewrite.

The functions were handed away by statute and by circular, and statutes and circulars can be reversed. The country can decide which capabilities it refuses to depend on anyone else for, and pay to keep them, the way it is already paying to buy back steel and chips and rare earths.

That choice is still open, and it is cheaper today than it will be after the failure that forces it. We know what year it is. The only question left is whether we would rather act like it is 1773, while acting is still an option, or wait for our 1857.

Want to help The Minuteman for free? Share it.

Sharing does not cost you anything. And it supports my work & independent journalism!

Share The Minuteman

New here? Sign up for free.

If someone did send you this, thank them and subscribe for free here:

Subscribe now

A quick note before we start:

I wrote this because the phrase “chip war” keeps getting it wrong. (Also i got to spend a lot of time reading about cool rocks).

A war is something you can win. Walk this supply chain from the rock to the device in your pocket and you keep hitting the same wall: almost no one in this fight can hurt the other side without bleeding themselves.

So I followed the chain down to the dirt, to the cheap, dirty work the rich world spent thirty years deciding it was too good to do. That is where the story starts.

Subscribe now

The Whole World Runs on Rocks Nobody Wanted to Mine

In May of 2026, the chief executive of an American company most people have never heard of got on a plane to China to ask for permission to buy a metal most people have never heard of.

The company was Coherent, one of a handful of firms that make the optical chips inside the data centers everyone keeps talking about, the ones that train the AI that is supposed to change your life.

The metal was indium phosphide. You will not find it in a headline about the future. You will find it in the part of the supply chain nobody photographs, the part with the trucks and the acid baths and the tailings ponds.

China had tightened its grip on indium phosphide in February of 2025, and since China makes around 70% of the world’s indium, the request was less a negotiation than a polite knock on a door that Beijing now controls.

That is the shape of the thing we have been calling a “chip war.”

We call it a war because the word fits the visible parts. There are tariffs. There are export bans. There are bipartisan letters from Congress and proxy tensions in the Taiwan Strait and a genuine fear, shared in Washington and Beijing, that this ends badly.

All of that is real. But walk the supply chain from the rock to the device in your pocket, and a different picture comes into focus, one that is harder to fit on a cable-news chyron.

Almost no one in this fight can hurt the other side without bleeding themselves.

The same lines that look like weapons on a map are also the only thing keeping everyone fed. The system that could start a war is the same system that has, so far, kept one from happening.

And the strangest part is what it all rests on.

The whole structure balances on the cheapest, dirtiest, least profitable work in the chain, the digging and the refining that the ‘rich world’ spent thirty years deciding it was ‘too good’ to do.

The brilliant chip designs and the trillion-dollar companies and the gleaming fabs all sit on top of that. China noticed. China made a bet that whoever owned the bottom of the chain would one day own the top. We are now living inside that bet.

So how did a few rocks end up holding the whole world together? And if no one can win, why does it still feel like we are all losing?

Subscribe now

What a chip is actually made of

Start with the chip itself, because almost everything you have heard about it is half true.

Read more

Issue No. 2 — You Can’t Own What Your Business Just Built

The power system behind a headline, before it shows up in your work, wallet, rights, or community.

Subscribe now

BRIEFING BOX
Core mechanism: The human authorship requirement. Copyright needs a person to make the creative choices, and prompting an AI does not count.
Main risk: A business that builds its brand on AI can’t legally stop anyone from copying it, and still carries the liability if the output infringes.
Decision room: Federal courts and the U.S. Copyright Office, not Congress.
Time horizon: 2026–2028.
Who should care: Solo founders and individual creators, small business owners, in-house marketing and brand teams, agencies and freelancers, investors and acquirers who value brand IP.
What to watch: The Third Circuit’s ruling in Thomson Reuters v. Ross, the basket of training cases in California, Copyright Office registration decisions on AI works, any federal AI bill that reaches a vote, the November midterms.
My call: No binding federal law on AI ownership or creator compensation by the end of 2026; this stays a courts-and-licensing fight (70%). Full reasoning below.

THE HEADLINE

You found a way to keep up. You used AI to build the logo, the website, the product photos, and the words that sell what you make, and you did it for a fraction of what an agency would have charged. The story you heard is that AI is a productivity gift for small business, and that part is true. What no one mentioned is that the work it made for you is, in the eyes of the law, authored by no one, which means it is yours to use and anyone’s to take. [More on The Death of Authorship]

THE ONE-LINE TAKE

The thing you built with AI to stay competitive is the one thing your competitor is free to copy, because the law says a machine’s work has no owner.

THE POWER SYSTEM

Start with the number people feel.

Almost 9 in 10 small businesses now use AI somewhere in their operation, up from about a third in 2023. The most common job they hand it is making things: marketing copy, social posts, product descriptions, logos, websites, the visual identity a business runs on. Owners who make the switch routinely report cutting their branding and content costs by half or more. You traded an agency or a freelancer for a monthly subscription, got a logo and a landing page in an afternoon, and put them to work. That was the smart move. It was also the move that left you owning nothing.

Now the part the headline skips.

The company selling you the tool makes its money on the subscription, not on your logo. Whether you can defend that logo in court is your problem, and nothing about your inability to own it costs the seller a dollar. So the tool keeps getting better at producing finished-looking work, the price keeps falling, and the question of who owns the result goes unanswered, because answering it was never the seller’s job. The largest companies have already handled their version of this. They have lawyers who document the human edits, designers who rework the machine’s draft until it clears the legal bar, and brand assets old enough to predate the whole problem. The small business that bought AI to replace the lawyer and the designer is the one with no way to close the gap.

Identify the mechanism.

Here is the hinge the whole story turns on, and it has a plain name: sufficient human control over the expressive elements.

The U.S. Copyright Office settled on that phrase in January 2025. Copyright protects work a person creates. When a machine makes the expressive choices, the layout, the colors, the shapes, the words, there is no human author, and with no human author there is no copyright. The Office was direct about prompting. Typing a description and accepting what comes back, even after dozens of tries, does not clear the bar, because there is a gap between what you asked for and what the machine decided to produce. You steered it. You didn’t author what it made. And copyright follows the author.

So the real fight, the one happening inside the Copyright Office and a stack of federal courtrooms right now, is over a single question: when you make something with a machine, who owns it? For the machine’s part of the work, the answer right now is no one.

That answer is not new. It comes from a rule called the human authorship requirement, and 2025 was the year it stopped being trivia. A computer scientist named Stephen Thaler tried to register a picture his AI made, listing the machine as the author. The Copyright Office said no, a federal appeals court agreed, and in March 2026 the Supreme Court declined to touch it. An artist named Kristina Kashtanova got a harder lesson: she made a graphic novel with AI images, and the Office protected her writing and the way she arranged the book while refusing protection for the images themselves. She had done the work. The law still told her the machine’s part belonged to no one. What was a curiosity when the author was a monkey with a camera became a business problem when the author was the tool half the country now runs on.

Look at the tell.

When the people in charge all do the same surprising thing at once, that is worth reading. In May 2025 the Copyright Office published a report concluding that much of the AI industry’s training on copyrighted work went beyond fair use. The President fired the head of the office the next day. She sued, an appeals court put her back, and she testified before the Senate this past May. You do not remove the referee one day after a call unless the call mattered to someone with power.

Look at the other side.

One honest complication, because you will hit it the moment you talk to a lawyer. You are not defenseless. A logo can often still be registered as a trademark, because trademark law asks whether a mark is distinctive and used in commerce, not whether a human drew it. And if you edit, arrange, and rework the machine’s output enough, the human parts can earn copyright. That is real, and it is the right advice. The catch is what a trademark does and does not do. It stops a competitor from using a confusingly similar mark to pass their business off as yours. It does not stop them from copying the image itself for their own, different business. For the copyright that would stop the copying, you need the human authorship you skipped when you let the machine do the creating. Which is the whole problem in one line: the cheaper and faster you let AI build your brand, the less of it you can defend.

That’s the system.

The company sells you a tool that produces work the law says has no author. It collects your subscription whether you can own the result or not. And it leaves the ownership gap sitting on the smallest businesses, the ones who bought the tool because they could not afford the lawyers and designers who know how to close it.

Share

WHAT THIS MEANS

For your wallet.

The savings are real, and the exposure is the price of them. The one lever that matters is human authorship. If AI builds your brand from start to finish, you own none of the machine’s work and can be copied at will. If a person makes the meaningful creative choices and you keep the receipts, the prompts, the drafts, the record of who changed what, you can own the result. The cost of protection is the human hour you used AI to remove. So decide which assets are worth that hour and put the hour back into those.

For your work.

If you make content for clients or an employer, the deliverable now carries a question it did not carry before: who owns this, and can we defend it? An agency handing over AI work it cannot copyright is handing over an asset the client cannot protect. The new line in the contract is who carries that risk. Settle it in writing before the work ships, not after a competitor reuses it.

For your rights.

The civic problem here is that the rule stays invisible until it fails. No one tells you at checkout that the logo you just generated cannot be copyrighted. The remedy that matters is disclosure and documentation: knowing, before you build on an asset, whether it is human-authored enough to own. The people who lose are the ones who find out during a dispute.

For your community.

Main Street adopted AI fastest where the budgets are thinnest. The shop that generated its sign, its menu, and its ads did the rational thing, and ended up with a brand it cannot defend against the franchise down the road that can. Two questions separate a safe asset from an exposed one: did a person make the creative choices, and can you prove it? If the answer is no, that asset is on loan from the public domain, and anyone can take it back.

Subscribe now

For your industry (if you build, invest, or advise).

Read more

Subscribe now

There Was a Time You Didn’t Hate the Other Side

Somewhere in the last few decades, disagreeing about politics turned into hating each other over it.

You feel it before you can explain it. The family group chat that goes quiet after someone shares the wrong article. The friend you stopped talking to. The holiday dinner you start timing like a bomb. We didn’t always live this way, and some part of us remembers that.

When I was young, my parents belonged to opposing political parties. One was a Democrat and one was a Republican. The difference between them, the way they explained it to me, wasn’t ethical and it wasn’t social. It was economic. One of them thought the government should be more careful with money, and the other didn’t think that question mattered as much as other issues. That was the whole of it. A fiscally conservative republican and a socially liberal democrat with varying political priorities. They voted against each other and stayed married, and neither one believed the other was a bad person for the lever they pulled.

That kind of disagreement still exists. But something has been laid on top of it, something that turns a difference of opinion into a reason for contempt. We usually blame each other for it. We blame the other side, the liars and bad actors and people who believe the wrong things. We assume the hatred is coming from what is being said.

I think we have the cause backwards.

Political division isn’t created by content; it’s created by medium. I don’t believe, for example, that our political division is a natural result of the human condition. I believe it is instead a result of the primary medium through which we communicate. Today, that tends to be social media. (Yes, I see the irony of this if you’re reading this on social media.)

In an attention economy, content is fed to people based on interactability. People and algorithms have decided that nothing determines your likelihood to interact more than anger. When emotion and logic meet, emotion always wins, and our primary method of communication has capitalized on this fact.

The medium does not care about your politics. It took me a while to see that. It doesn’t even care whether you end up divided. It cares about holding your attention, and the cheapest way to hold your attention is to make you feel something, and the feeling that performs best is anger. So it almost doesn’t matter what you say, or which side you say it from. The medium runs your words through the same machine and pays out on the same emotion. It manufactures the division for you.

It may seem like a contradiction, but media itself is innocent in all this. Radio could carry a fireside chat or a demagogue with equal ease, and in its time it carried both. What mattered was never the wire. It was what we built the wire to reward.

We built our communication to reward attention, attention turned out to run on emotion, and emotion at scale hardens into outrage. Television sped that up. Social media perfected it. And once the format was proven to work, everything else that feeds on our attention followed it down the same path, until it no longer mattered whether you were the one holding the phone.

If that is true, then the hatred is not who we are. It is a feature of how we are talking to each other right now, and a feature can be changed. We are not condemned to this. We can choose to communicate in ways that don’t sort everything we say into a fight. That choice is the reason I’m writing this down.

But I want to do more than assert it. The machine has more parts than I first understood, and each one makes the next one worse. Once you see how they fit together, you can’t unsee it.

Subscribe now

What we built the wire to do

Start with the economy underneath everything else.

In 1971, a researcher named Herbert Simon noticed something that sounds obvious now and was strange then. When information becomes cheap and endless, he wrote, the thing it consumes is attention. Information stops being scarce. Attention becomes the scarce thing. And the moment something is scarce, it becomes valuable, and the moment it becomes valuable, someone builds a market to buy and sell it.

That market is what we now call the attention economy, and we all live inside it. Most of what reaches you in a day, the post, the notification, the autoplay, is built to hold your attention, not to inform you. Holding you is the product. Your attention is what gets sold, and everything competing for it has been shaped, slowly and then all at once, to win.

There is a problem buried in building a machine to capture attention. Attention does not run on reason. It runs on feeling. We stop for the thing that makes us feel something and scroll past the thing that asks us to think. So a system built to harvest attention drifts, on its own, toward whatever makes us feel the most, the fastest. Nobody has to sit in a room and decide this. The system selects for it the way water finds the low ground.

When emotion and logic meet, emotion wins. A medium built to capture attention becomes a medium built to capture emotion, whether or not that was ever the plan.

Why it’s always anger

Not every emotion is worth the same to the machine.

Love holds you. Wonder holds you. Grief holds you. But anger holds you hardest and cheapest, and the machine can tell. We can measure it now, because researchers went looking.

One study tracked more than half a million messages about guns, climate, and same-sex marriage. Every moral or emotional word added to a post made it travel about twenty percent farther. The same study found something quieter and more important: the outrage spread mostly inside each side and rarely crossed over. It made each camp angrier at the other without ever putting them in the same room.

Another team tested over a hundred thousand real headlines against real readers. Every negative word added to a headline raised the click rate by a couple of points. Positive words did the opposite. Bad news travels; good news sits.

And we don’t have to guess what the platforms knew, because their own documents got out. For years, Facebook’s feed counted an emotional reaction as five times more valuable than a plain like, and used that signal to decide what to show you next. The company’s own researchers later found that the angriest of those reactions clustered around the posts most likely to be false or toxic. The machine had learned that strong feeling was the most engaging thing in the world, and it had quietly, mathematically, decided to hand us more of it.

There is a sharper version of this, and it points straight at the other side. When researchers looked at what kind of anger spreads fastest, they found it wasn’t anger in general. It was anger at the out-group.

In an analysis of nearly three million posts, every single word that referred to the political opposition raised the odds of a post being shared by about 67 percent. Praising your own side did almost nothing. Attacking the other side was close to a guarantee of reach.

The tech ethicist Tristan Harris put a number on the gap with one example. The day the Justice Department released a photo of classified documents at Donald Trump’s Florida home, a tweet linking the straight news story drew about 2,000 likes. A tweet calling Trump’s opponents “dumbasses” drew ten times that. A tweet calling Trump a traitor drew twenty times that. The calm account of what happened was the least valuable thing in the exchange. That is one cherry-picked day rather than a dataset, but it puts a face on the 67 percent.

So the machine sells a particular flavor of anger, the kind pointed across the aisle, because that is the kind that travels farthest. The target is always the other side. That detail explains a lot about why the anger works on us so well, once we get there.

Now watch the loop close.

We learn what performs, so we make more of it. The headline gets sharper. The take gets hotter. Some of this is deliberate, the optimized headline, the rage-bait, and a lot of it is just instinct, the post that got a reaction so you write the next one like it. Then the medium takes whatever we made and sorts it again, pushing the angriest version toward the people most likely to react. We aim our words at emotion, and the machine aims them at emotion a second time. Two passes, both selecting for the same thing.

None of this requires anyone to want division. That is the part that took me longest to accept. There is no villain at the center pulling people apart on purpose. There is a machine that pays out on anger, a population that supplies it for free, and a loop between them that tightens a little more every year.

Share

When the hatred actually started

If the medium is the cause, the history should show it. It does.

Go back to my parents for a second. There was a time, and it was not long ago, when most people felt about the other side roughly the way my parents did. In 1960, fewer than one in twenty Democrats or Republicans said they would be unhappy if their child married someone from the other party. By 2010, it was about half of Republicans and a third of Democrats. By 2016, majorities in both parties said the other side made them afraid.

The disagreement changed temperature. It went from a difference of opinion to a feeling about a person, and it did that recently, on a timeline we can actually watch.

What changed on that same timeline was the medium we use to talk to each other.

Here I have to be honest, because the history cuts both ways and I would rather you trust me than win the point. The medium did not invent division out of nothing. Think about radio.

In the 1930s, a priest named Charles Coughlin reached something like one in four Americans every week, and he used that reach to broadcast hatred, much of it aimed at Jews. Radio could obviously carry poison. But the very same radio carried Franklin Roosevelt’s fireside chats into the very same living rooms, and those did the opposite. They steadied people. They pulled a frightened country closer together.

One wire, the same year, two opposite effects, decided entirely by what ran through it. Economists who studied Nazi radio in Germany found a darker version of the same fact: the broadcasts pushed people toward hatred in places where hatred was already waiting, and actually pushed them the other way where it wasn’t. The medium amplified what was there. It did not conjure it.

That is the honest version, and it is also the whole point. The wire is innocent. What changes everything is the structure we build around the wire, the thing we teach it to reward. And starting a few decades ago, we taught it to reward the wrong thing.

Television made attention faster and hungrier than radio ever was. Then, in 1987, the government removed a rule called the Fairness Doctrine, which had required broadcasters to air more than one side of a question. A year later, a Sacramento radio host named Rush Limbaugh went national. By the mid-nineties he was reaching twenty million people a day, and he was not selling both sides. He was selling one side, with feeling. Cable news learned the lesson and built entire channels on conflict.

This is not a guess about what that did. When researchers measured it, exposure to Fox News moved Republican vote share by a margin that grew with every election, and counties soaked in Limbaugh’s signal drifted measurably his way. The outrage format helped make the division it claimed to report, and we can see its fingerprints in the vote.

Which answers the objection people always raise. If social media is the cause, why are the angriest, most divided Americans often older people, the ones least likely to be on it? Because the format reached them first, through talk radio and cable news, years before the smartphone existed. The structure was already loose in the older media, doing its work on the people watching. The phone didn’t start this. It finished it.

The format that ate everything

Social media did not invent any of this. It perfected it.

Everything the old media did slowly and bluntly, the feed does instantly and precisely. It does not broadcast one angry message to everyone. It finds the exact thing that will make the exact you react, serves it, watches what you do, and adjusts. The sorting that took talk radio a decade now happens in the time it takes to thumb past a post.

Here is where the strongest objection comes in, and it is worth taking seriously.

In 2020, researchers worked with Meta to switch tens of thousands of users off the ranked feed and onto a plain chronological one, the fix almost everyone assumes would help. Three months later, polarization hadn’t moved. If the feed were the whole problem, that should have done something, and it didn’t.

But look at what the experiment actually changed. It reordered the posts inside the same machine, on the same platform, running on the same incentive, surrounded by the same people making the same outrage for the same rewards.

Reorder the shelf and the store still sells what it sells. The ranking is one gear in a much larger machine, and you can swap that gear with the engine still running. That is the whole argument here. The thing manufacturing division is the attention economy itself, and the feed is only its most visible dial.

There was a specific moment the dial turned.

Around fifteen years ago, the platforms added the like, the share, and the retweet. Before that, social media was mostly people posting where they ate lunch. After it, every post carried a public score, and anyone could fire a thought at thousands of strangers in a second, with no cost and no accountability.

The psychologist Jonathan Haidt calls it handing everyone a dart gun. You could now attack anyone, anytime, and the numbers would reward you for it. That was the day the structure got its trigger.

And once it was proven to work, everything else that lives on attention copied it, and that reaches well past the platforms.

Newsrooms built to inform you started writing for the feed. The headline is tested for clicks. The story is chosen because it travels. The chyron is built to hold you angry through the commercial.

Most people now get their news filtered through these channels, so the logic of the feed became the logic of the news, and the logic of the news became the story a whole country tells itself about what is happening to it. We are all downstream of it now, even the people who never signed up.

You can write something reasonable, something meant to start a conversation, and watch the medium turn it into a side. You did not pick the fight. The structure picked it for you, because a fight performs and a conversation doesn’t. You lose authorship of your own words. Whatever you meant, the machine knows what it can sell, and it sells the version that makes someone angry.

The cruelest part is that the machine has no politics at all. It is not on the left or the right. It does not want you to believe anything. It wants engagement, and it will take it from anyone.

That neutrality comes with a catch, because a machine with no politics of its own can still be owned by someone who has them.

When Elon Musk bought Twitter in 2022 and rebuilt it as X, the platform’s recommendation system began nudging users toward right-leaning accounts and more conservative positions. The engine still runs on anger; the hand on the throttle is not always neutral.

Even so, both sides are full of people who care, and the medium routes all of that caring into the one emotion it can charge for.

You think you are fighting the other side. You are both standing at opposite ends of the same machine, feeding it.

Share

Who is actually screaming

Here is something that should be reassuring and somehow isn’t. Most people are not doing this.

When it feels like the whole country is at war, the picture in front of you is the work of a very small number of hands. The people who fire the dart guns most, who post the most political content and the angriest version of it, sit at the far ends of the spectrum.

Haidt puts the active extremes at about seven or eight percent on each side. The broad survey he draws on found the most politically active group online, the progressive activists, made up roughly eight percent of the population, and the devoted conservatives on the other end about six. Those are the people supplying most of the fuel. The other eighty-odd percent of us are mostly tired, mostly quiet, and mostly convinced we are surrounded.

The medium prefers it this way, because the extremes make exactly the kind of out-group anger that travels 67 percent farther. The loudest eight percent produce the most shareable content, the machine hands them the megaphone, and the rest of us get served their voices on a loop until we mistake them for everyone.

So why don’t the moderates push back and drown them out? Because the structure makes it expensive to try.

Haidt has a name for what happens when they don’t: structural stupidity. Put thoughtful people in a place where one wrong word gets you swarmed, and they learn to say nothing. They watch a colleague get flattened for a careful, complicated take, and they decide the careful take isn’t worth it. Nuance becomes a liability.

The quiet majority goes quieter, and the institution they belong to, the campus, the newsroom, the workplace, loses access to its own best thinking.

When the smart people are afraid to speak, the room gets dumber, and the only voices left are the ones that were never afraid because they were never interested in being careful.

This is the part that makes the caricature possible. The version of the other side you meet online is its angriest eight percent, the ones the machine selected for being the most enraging, performing for an audience, with the moderates edited out.

You think you are seeing your neighbor. You are seeing the most extreme stranger who happens to share your neighbor’s party, lit up and served to you because he makes your heart rate climb.

Why the anger fits

A machine that sells out-group anger still needs the anger to work on us. You can’t sell a feeling people don’t have.

So the last question, the one underneath all of it, is why this particular provocation gets us every time. Why does a stranger’s bad take about politics reach into your chest the way a stranger’s bad take about almost nothing else does.

The answer is that the dart guns are aimed at something real.

We don’t run on a single sense of right and wrong. Psychologists who study morality across cultures argue that it rests on several different foundations, and they don’t all carry the same weight for everyone.

Researchers still fight over how many of these foundations there are and how deep they run, so take the framework as a useful map and not a settled fact. The part that survives the argument is the simple one I need here: people don’t all weigh the same moral concerns equally.

Haidt, who built the theory, compares the foundations to taste receptors. We all have a tongue, but we don’t all reach for the same flavors.

Two of those foundations are care, the instinct to protect people from harm, and fairness, the instinct against being cheated. Almost everyone has them, and for people on the political left they tend to be most of the meal.

The right keeps care and fairness too, but sets three more places at the table that the left tends to weigh less: loyalty to the group, respect for authority, and a sense that some things are sacred and shouldn’t be degraded.

None of these is the moral one. They are different instruments for reading the same world, and most arguments that feel impossible are two people playing different instruments and hearing the other as noise.

This is where the misreading begins, and it begins before the internet ever touches it.

When someone whose morality runs on care and fairness hears a passionate argument built on sanctity or authority, it can fail to register as a moral argument at all. It sounds like a person getting worked up about nothing, or worse, dressing up cruelty as principle.

Run it the other way and it’s just as bad. To someone who feels loyalty and sacredness deep in the body, a person who shrugs at them looks like someone with pieces missing, a person who would let anything be torn down because they can’t feel why it matters.

Each side hears the other’s deepest moral concerns as either a con or a void, and treats the mismatch as proof of bad faith.

Now feed that into a machine that pays out on out-group anger, and you get something measurable and strange.

Researchers asked liberals and conservatives to describe their own morality, and then to fill out the same survey the way they imagined a typical opponent would.

Everyone got the direction right and the distance wildly wrong. Both sides pictured the other as a cartoon of itself, more extreme and more heartless than the real people turned out to be.

Liberals guessed that conservatives don’t care about fairness or protecting the vulnerable, when they do. Conservatives guessed that liberals feel no loyalty to anything, when they do. The people who were most sure they understood the other side were the most wrong about it.

That is the engine of the whole thing. We are not, most of the time, fighting the other side. We are fighting a picture of the other side, drawn by the most extreme eight percent, exaggerated by our own moral blind spots, and pushed further every time the medium serves us the most enraging version it can find.

The machine didn’t create the gap between our moral languages. It found the gap, learned it was profitable, and has been prying it open ever since.

You think you are at war with people who believe monstrous things. You are mostly at war with a hallucination, and the hallucination is the product.

What it costs a country

It would be easier to file this under unpleasant and move on, except that the damage has a body count and a direction.

When the hatred stops being about policy and starts being about the person, political scientists call it affective polarization, and it does something specific. It dissolves the floor we stand on together.

People who view the other party as evil stop accepting outcomes that go against them, stop trusting the count, start excusing things they would have been ashamed of a decade ago, because any means looks justified against an enemy that monstrous.

Barbara Walter, who studies how civil wars actually begin, lays out the warning signs in her book How Civil Wars Start. The thing that predicts a country tipping into conflict is a specific combination: a population sorting into camps defined by identity rather than ideas, inside a government too gridlocked to settle anything.

Strong disagreement on its own is survivable. That sorting is what turns dangerous, and by her measures the United States has been drifting out of the safe zone of stable democracy toward the unstable middle ground where the risk lives. The sorting she warns about is exactly the kind the medium manufactures.

This isn’t only an American story.

Freedom House, which scores political rights and civil liberties around the world, reported that global freedom declined in 2025 for the twentieth straight year, with more than fifty countries getting less free and only about thirty-five improving.

The United States had the sharpest one-year drop of any country still rated free, falling three points to 81 out of 100, its lowest score since the modern measure began in 2002.

A generation of backsliding, accelerating now, alongside the machine we built to keep us angry.

I want to be careful here, because this is where it would be easy to overstate. The medium is not the only thing dragging democracies down, and a post is not a war. But the mechanism connects cleanly to the stakes.

A structure that profits by teaching people to see their neighbors as enemies, running on a platform that reports about 240 million American users, is not a harmless way to spend an evening.

It is corrosive at exactly the joint that holds a self-governing country together, the willingness to lose an argument and stay in the room.

Subscribe now

What your hatred is worth

So here is the bill.

Your attention is the product. And of every form your attention can take, your hatred is the most valuable, because it is the most reliable and the easiest to provoke again tomorrow. Every time the feeling wins over the thought, a transaction clears and someone gets paid. It is not you.

That is the price of your hatred, and you have been paying it without ever seeing the invoice.

I find that, strangely, hopeful.

If our division were simply human nature, there would be nothing to do but grieve it. But it is not human nature. It is a structure, and a structure is a choice, and a choice can be made differently.

Some of that choice is bigger than any one of us. The people who built this know what it costs, and there is a fight coming over whether a business model that runs on enraging us should be allowed to operate untouched, the way we eventually decided about tobacco and leaded gasoline and cars without seatbelts. That fight matters, and it is not ours to win alone on a Tuesday night.

But some of it is exactly the size of one person, and it is available right now.

We are not required to talk to each other through machines built to set us against each other. We can choose rooms that are friendly to what we actually mean.

A long article that asks you to hold one thought for ten minutes.

A conversation that is not scored by reactions.

A dinner table, a club, a letter, any place where your words reach another person without being sorted, ranked, and sold on the way.

The medium is a choice. It has only felt like the air because we forgot we were allowed to breathe somewhere else.

My parents disagreed about the importance of fiscal conservatism in politics most of their lives. They never once mistook that disagreement for a reason to hate each other, and the reason is almost embarrassingly simple. They worked it out across a kitchen table, and that table sold nothing. No one was paid when they disagreed, so nothing in their lives was quietly working to keep them angry.

That is still available to us. The other side was never the enemy. The thing that taught us to see an enemy there was. It has a price, and we do not have to keep paying it.

TLDR: Read a book.

Want to help The Minuteman for free? Share it.

Sharing does not cost you anything. And it supports my work & independent journalism!

Share The Minuteman

New here? Sign up for free.

If someone did send you this, thank them and subscribe for free here:

Subscribe now

Issue No. 1 — Who’s Paying for AI Buildout?

The power system behind a headline, before it shows up in your work, wallet, rights, or community.

Subscribe now

BRIEFING BOX
Core mechanism: Cost allocation — who pays when one customer triggers new grid spending
Main risk: Ratepayers absorbing the infrastructure costs of private data centers
Decision room: State public utility commissions (PUCs)
Time horizon: 2026–2028
Who should care: Ratepayers, public officials, utilities, AI/data-center companies, investors, public-affairs professionals
What to watch: PUC rate cases, large-load tariffs, special contracts, co-location rules, the November midterms
My call: Cost-causation tariffs go from outlier to PJM default by the end of 2027 (70%). Full reasoning below.

THE HEADLINE

Your electricity bill went up. The news told you it was data centers. That’s not wrong — but it’s not the system. The system is older, quieter, and it was built to do exactly this.

THE ONE-LINE TAKE

The AI buildout didn’t break your power bill. It walked into a hundred-year-old rule that says everyone pays for the grid — and turned a single customer into something that rule was never designed to handle.

THE POWER SYSTEM

Start with the number people feel.

Residential electricity prices are up >36% since 2020 — rising faster than inflation for four straight years, with no relief forecast. Electric and gas bills were two of the largest single drivers of inflation last year.

Now the part the headline skips.

A regulated electric utility does not make money the way a normal company does. It does not profit by selling you more electricity. It profits by building things — power plants, transmission lines, substations — and then earning a guaranteed rate of return on the money it spends building them.

Capital spending isn’t a cost to a utility. It’s the product. The more it builds, the more it is allowed to earn.

You don’t have to take that on faith: in its own February 2026 investor filing, Duke Energy tied a record $103 billion five-year construction plan — the largest fully regulated capital plan in the industry — to 9.6% growth in the earnings it is permitted to collect from customers. The buildout is the business plan.

For about twenty years, that machine idled. American electricity demand was essentially flat. Then came a customer unlike any the system had seen: a data center that draws power like a small city, runs flat-out 24 hours a day, and wants to be built now.

Data centers are on track to account for roughly half of all new U.S. electricity demand. To a utility’s business model, this is not a problem. This is the best news in a generation. It is a reason to build — and to earn — at a scale the old flat-demand era never permitted.

Identify the mechanism.

Here is the hinge the whole story turns on, and it has a boring name: cost allocation.

When a utility builds new transmission and generation, who pays? The default answer, baked into the system since the mid-20th century, is socialized cost — you spread the bill across the whole customer base, because historically everyone benefited from a bigger, stronger grid.

That logic held when growth was diffuse. It collapses when a single named corporation is the reason a billion-dollar line gets built. Under the old rule, the costs of serving Amazon, Google, or Microsoft’s facilities can quietly land in the rates paid by the household three towns over who will never use that capacity.

This isn’t a theory.

In the 13-state PJM grid region, customers across seven states paid an estimated $4.3 billion for data-center-related transmission in a single year — and close to half of that fell on Virginia, the capital of “Data Center Alley.”

A Harvard Electricity Law Initiative paper laid out the mechanism: through secret special contracts, through a disconnect between how federal transmission rates and state retail rates are set, and through novel “co-location” deals with power plants, utilities have multiple legal channels to shift the cost of serving data centers onto captive ratepayers, and a profit motive to do it.

So the real fight — the one happening in rate cases and statehouses right now, mostly out of view — is a fight over a single question: does the customer who causes a cost have to pay for it?

That principle is called cost causation, and it is the language of the moment.

Virginia has moved toward it with a tariff requiring very large customers to cover 85% of the transmission capacity costs they trigger during ramp-up. Oregon passed a law forcing separate accounting so data-center infrastructure can’t quietly inflate residential rates. In December 2025, federal regulators (FERC) ordered PJM to rewrite the rules for co-located data centers because the existing tariff was too vague to protect anyone. In January 2026, all 13 PJM governors — across both parties — plus the White House signed a joint statement demanding that data centers bear their own infrastructure costs. The White House name-checked the issue in a March 4, 2026 announcement of a “Ratepayer Protection Pledge.”

When both parties and the White House agree on something in an election year, that’s a tell. It means the politics got hot enough that nobody wants to be holding the bill on election night, and the midterms are in November.

Look at the other side.

One honest complication, because you’ll meet it in the room: the industry disputes the premise.

An Amazon-commissioned study and a separate utility-industry report argue that, designed correctly, large data-center customers generate surplus revenue and actually push everyone’s rates down.

That’s not pure spin — a customer paying for a lot of new infrastructure can, in principle, share fixed costs more widely. The catch is the phrase “designed correctly.” Whether these deals help you or soak you depends entirely on the cost-allocation fine print that almost no ratepayer ever sees.

Which is the whole point: the outcome is being decided in documents written in a language built to keep you out of the conversation.

That’s the system.

Not “AI is using too much power.” It’s: a profit model that rewards building, meeting a customer that justifies unlimited building, governed by a cost-sharing rule that defaults to making you the co-signer.

Share

WHAT THIS MEANS

For your wallet.

Your bill is a policy outcome. The single biggest lever on your future rate isn’t the price of natural gas — it’s whether your state PUC adopts cost-causation rules before the next wave of data-center contracts is signed.

National forecasts put the average increase around 8% by 2030, but in the most data-center-heavy state, Virginia, it could run to 25%. The variance is the policy. Where you live, and what your PUC decides in the next 18 months, will matter more than any choice you make about thermostats.

For your work.

If you’re in energy, real estate, public affairs, or anything touching infrastructure: the special-contract era is ending and the cost-causation era is beginning. Deals structured on the old assumption that costs can be socialized are about to get repriced or challenged. If you advise clients, the question to ask is no longer “can we get the power?” but “who’s legally on the hook for the grid upgrade, and is that allocation defensible in a 2027 rate case?”

For your rights.

The core civic problem here is invisibility by design. Special contracts are often confidential. Transmission costs are disclosed in “informational” briefings that never make it into the rate proceedings where your bill is actually decided. The remedy that matters most isn’t a moratorium — it’s disclosure. Knowing whether your community has a public, standard tariff or a secret special deal is the difference between being a participant and being a co-signer.

For your community.

A data center is a bet your town is being asked to underwrite, often without being told it’s the guarantor.

The two questions that separate a good deal from a bad one: (1) Is the facility served under a public tariff every large customer pays, or a private special contract? (2) Who pays if the projected demand never shows up — the company, or the ratepayers left holding stranded infrastructure?

If local officials can’t answer those two questions, that’s the story.

Subscribe now

For your industry (if you build, invest, or advise).

Read more

A note before we start.

I was looking at my IRA this week.

That is not a sentence I ever expected to write with any emotional weight behind it, but there I was, clicking through funds, trying to understand the acronyms, pretending I knew the difference between all the responsible-looking options lined up in front of me.

I was trying to do the adult thing. Make a reasonable decision. Improve my financial literacy. Put whatever money I have, which is not a lot, somewhere it might still be useful to me later.

Then I did the thing I apparently cannot stop doing.

I started pulling on the thread.

That thread led into index funds. Then target-date funds. Then IPO rules. Then float requirements. Then seasoning periods. Then corporate structures, exchange incentives, asset managers, tax law, and a whole private language built around the money most of us are told to set aside and stop thinking about.

I do not care about finance in the way the finance bros i went to high school with care about finance. I am not interested in the market as a hobby. I do not wake up wanting to study fund methodology documents. No offense to the people who do. Not everything is for everybody.

The only reason I care is because it is my money.

And because, somewhere along the way, I realized that someone with a great deal more money may be about to get richer using the small amount I have managed to save.

I started down this tunnel knowing embarrassingly little about what was happening to my own retirement account. The whole thing felt designed to stay unknowable. Personal finance. Corporate finance. Index rules. Private markets. Public listings. Acronyms stacked on acronyms until ordinary confusion starts to feel like a personal failure.

But it is not unknowable.

Like most systems where confusion at the bottom benefits the people at the top, the trick is not to understand everything at once. The trick is to ask the question that keeps working.

Who wins?

Who loses?

Who benefits from this arrangement, and who quietly pays for it?

So if you feel the way I did, don't worry. A working knowledge of this stuff is within reach.

If you don’t care enough to try, and you’d rather put your energy somewhere else, I don’t blame you, and you can skip this one. The world will gladly take your help with any of the thousand other horrors on offer.

If you already know more than I do about this, and some of you do, you can skip it too, because you probably won’t learn anything new.

But for the rest of us in the middle, which I like to believe is most of us, let’s spend this week on one thing: how one of the largest companies in the world is using our retirement savings to make the richest man in the world much, much richer.

Subscribe now

You Already Own It

Ideally there’s a retirement account with your name on it. Maybe you haven’t looked at it in a while. Maybe that’s the point. Maybe an employer opened it for you. Maybe you picked a fund years ago, the one with a date in its name close to when you hope to stop working, and you haven’t thought about it since.

You turn on the news. Read the headlines. You hear AI hype on one side, concerns about an AI bubble on the other. Talk about OpenAI or Anthropic or SpaceX going public, IPOs and stock exchanges, but you don’t care because you think, “Why should I?”

You have no intention of buying stock in SpaceX or Anthropic or OpenAI when they go public. You made the decision, played the safe bet. You might have opinions about Elon Musk, but you never wrote a check to any company he runs.

But within weeks of June 12 you might start to own one of his companies anyway.

The path runs straight through the safest-looking choice you made.

That fund with the date in its name doesn’t bet on companies the way a day trader does. It holds a little of everything and tracks the market as a whole, so no single bad pick can sink you. The spreading out is the part that gets sold to you as safe. It’s also the part that’s about to hand you SpaceX.

When SpaceX lists on June 12, it can join a major stock index within weeks. It can do that because an exchange rewrote the rule that used to make a new company wait and prove itself first. The funds that track that index don’t get a say in whether SpaceX belongs there. When the index takes it in, they buy it. Your retirement fund holds those funds. The buying travels down the line until it reaches you.

You never looked at SpaceX’s numbers. You were never asked. The choice that put it in your retirement was made by a private company changing its own rules to win a listing, and nobody held a vote.

So before June 12, this is worth sitting with. Something you didn’t buy, didn’t want, and were never asked about is about to show up in the account you were told was the safe one. Someone decided that for you. The safe bet had a rule buried inside it, and while you weren’t looking, the rule changed.

This is a story about the loss of financial consent and your financial agency. And the problem is bigger than just money.

Share

The fund that buys for you.

Here is how the safe bet works.

You pick a fund. Maybe it tracks the S&P 500, a list of five hundred of the largest companies in the country. Maybe it tracks the Nasdaq-100. Maybe it sits inside a target-date fund and you have never seen the name of the list it follows.

The fund makes one promise: it will hold whatever the list holds, in the same proportions, so your money moves with the market instead of trying to beat it.

That promise is the product. It is why target-date funds and index funds became the default for retirement savings. You do not have to research companies. You do not have to pick stocks. You do not have to watch the market. The list does the choosing. The fund does the buying. Your job ended the day you picked the fund.

But the promise runs in both directions. When a company gets added to the list, every fund tracking it has to buy that company’s stock. The manager does not weigh the evidence and decide the company is a good investment. The list changed. The fund follows. The money moves.

That is the mechanism underneath the safe bet. You made one decision. After that, decisions are being made for you, by whoever controls what goes on the list and how quickly it gets there. The only question left, the one that turns out to matter more than any stock pick you could have made, is this: what does a company have to do to get on that list?

Until last month, the answer included two rules.

The price had to become real first.

When a company goes public, the price on its first day of trading is a guess. The company and its bankers set it the night before, based on what early investors were willing to pay in private.

That number is a starting point. The months that follow are when everyone else gets to weigh in, when buyers and sellers who had no part in the offering start trading the stock and the price begins to reflect what the market, not the bankers, thinks the company is worth.

The old rule said a company had to go through that process before a major index would let it in. The industry calls it the seasoning period. For the Nasdaq-100, it was typically at least three months. For the S&P 500, twelve.

A company had to sit in the open market and be priced by real trading before the index would add it and the tracking funds were ordered to buy.

The logic was plain. The forced buyers, every fund tracking the index, were going to be told to purchase this stock. The seasoning period made sure the price they paid was one the market had arrived at through months of open trading, not a number set in a conference room the night before the listing.

There had to be enough stock to go around.

The second rule was about supply.

When an index adds a company, every tracking fund has to buy shares at the same time. If only a thin slice of the company’s stock is actually available for public trading, all that forced money competes for a small pool. The price jumps, not because anyone decided the company was worth more, but because there are not enough shares to absorb the demand.

The old Nasdaq-100 rule required at least ten percent of a company’s shares, its float, to be trading in public hands before inclusion. The rule made sure that when the forced buying arrived, there were enough real shares on the market to keep the price from being driven up by the mechanics of the purchase alone.

Together, the two rules did the same thing from different angles. The waiting period let the price become real. The supply rule made sure there was enough stock to go around when the buying came. Both cost the company time and exposure, and both protected the same person: the one at the far end of the chain who never picked the stock, never evaluated the company, and never asked for it to be in their fund.

That person is you. And at the exchange that won SpaceX, both rules are gone.

One exchange rewrote the rules. One refused.

On May 1, NASDAQ changed its entry standards.

A newly listed company big enough to rank among the top ten in the Nasdaq-100, roughly a two trillion-dollar company, can now be evaluated for inclusion as early as its seventh trading day and added in about fifteen.

The seasoning period is waived. The hard ten percent float minimum is gone. A company can go public on a Monday and sit in millions of retirement accounts three weeks later, its price still fresh from the offering.

The other big index provider faced the same choice and made the opposite call.

This spring, S&P Dow Jones Indices, which runs the S&P 500, opened a consultation on whether to cut its twelve-month waiting period to six, waive the supply rule for the largest companies, and drop the requirement that a company turn a profit before it can join.

That profitability rule is one of the oldest guarantees in the index, the promise that the list of America’s largest companies at least holds companies that make money. On June 4, days before SpaceX’s debut, the S&P declined. It kept all three rules.

SpaceX, which loses money, will not enter the S&P 500 on any fast track. It waits the full twelve months, and it has to earn a profit first.

So one provider dropped the protections and one held onto them, in the same season, looking at the same company. That tells you something the rule change alone would not.

These rules were never laws of nature. They were choices, and they could be defended. One of the two companies that write them chose to defend them. The other chose the listing.

Because an exchange is a business. A listing is a customer. The largest IPO in history is the largest customer an exchange has ever had. SpaceX’s own advisers approached NASDAQ and other index providers to accelerate the path to inclusion. The exchange that said yes won the listing. SpaceX chose NASDAQ.

The S&P kept its standard and, for now, loses the business. The standard was not abandoned. It was held by one company and competed away by the other, and nobody had to argue the old rule was wrong.

But before either provider decided, someone was asked whether the rules should change.

Share The Minuteman

Nobody asked the people in the funds.

When NASDAQ proposed its Fast Entry rule, it opened a consultation.

The responses it received, and found mostly supportive, came from asset managers and institutional passive portfolio managers. The people who run the index funds. The people whose job, once a company enters the list, is to execute the buy order.

Those are the people who were asked whether removing the waiting period and the supply floor was a good idea. The people whose retirement savings would carry the result were never asked.

There was no public vote. No regulator signed off. No law requires an index provider to ask the people whose money flows through its list.

NASDAQ is a private company. It writes its own index rules. S&P Dow Jones Indices is a private company too, and it writes its own rules, which is exactly why it was free to keep them.

The lists that decide where trillions of dollars in retirement savings end up are maintained by private businesses, and the customer that mattered most this spring was the company trying to get on the list.

Self-regulation works when the interests of the regulator line up with the interests of the people it protects.

At NASDAQ they pulled apart. The exchange earns revenue from listings. The rules that protected savers cost the exchange its biggest listing. So the exchange removed them. Nobody with a retirement account was asked whether that trade was worth making.

The rules were changed legally, by the private company that owns them, with the support of the private companies that execute the trades, for the benefit of a private company that wanted faster access to your savings.

The person at the end of the chain, the one whose account absorbs the result, was never part of the conversation.

And SpaceX is first through the door.

A loss of consent.

Two rules stood between a new listing and your retirement account. Both were removed this spring by the exchange competing for the largest customer it has ever had, with the approval of the people who manage the money, and without the knowledge of the people who own it.

What that leaves is a loss of consent.

And SpaceX is not the only company in line. A pipeline of the largest companies in the world is heading for public markets, and most of them lose money.

SpaceX filed to go public at a $1.75 trillion valuation, the largest in market history. Its filing disclosed $18.67 billion in revenue for 2025 against a $4.94 billion loss, and an accumulated deficit north of $41 billion.

Starlink, the satellite internet business, made money. xAI, the artificial intelligence company Musk folded into SpaceX earlier this year, burned it faster.

Behind SpaceX, the line is forming.

Anthropic, another major AI company and the maker of the chatbot Claude, filed confidentially on June 1 at a $965 billion valuation, and it is reported to be approaching its first profitable quarter, which would make it the exception.

OpenAI, the company behind ChatGPT, was last valued around $852 billion, is not profitable, and is expected to list by fall.

These are among the largest companies ever to approach public markets. Most of them lose money. Every one would have been slowed by the old waiting period.

The door at NASDAQ was widened to match the size of what is coming through it.

Two protections were dropped to clear the way: the waiting period and the floor on tradable shares. The third, the requirement to earn money first, never existed at the Nasdaq-100 and still stands at the S&P 500, which is the only reason SpaceX is being kept out of one of the two lists instead of neither.

But how did we get here?

Share

How SpaceX got us here.

Musk bought Twitter in October 2022 for about $44 billion, renamed it X, and watched a large share of that value disappear. Money sitting inside a falling asset tends to get moved, and this money did.

In March 2025, X merged into xAI, his artificial intelligence company, the more hyped and more valuable of the two. The combined value now sat inside a private company no public market had priced.

In February 2026, SpaceX acquired xAI in an all-stock deal valued at about $1.25 trillion. No cash changed hands. People who had put money into Twitter, then into xAI, woke up holding shares in the most valuable private company on earth without selling a thing.

Then SpaceX filed to go public at $1.75 trillion.

From $44 billion to $1.75 trillion in under four years, through a series of mergers that moved value from a shrinking asset to an AI company to a rocket company, and at no point did a public market set the price.

The first time the public gets to weigh in is the IPO itself. Fifteen trading days later, the Nasdaq-100 can let it in and the forced buying begins.

The old rules existed for exactly this situation.

The seasoning period would have made SpaceX trade in the open for months, long enough for the market to test whether a $1.75 trillion price holds up when ordinary buyers and sellers set the terms.

The profitability requirement is keeping it out of the S&P 500 right now: SpaceX lost $4.94 billion last year.

The old supply rule would have required at least ten percent of its shares to be trading in public hands before the index admitted it. SpaceX plans to float closer to four or five percent.

On June 12, a $1.75 trillion company goes public, assembled through a chain of private mergers, carrying a $4.94 billion loss, and offering a sliver of its shares to the public. Within weeks it could sit in an index your retirement fund tracks.

At the exchange that took it, the rules built for this moment are gone.

The obvious fear, and why it’s (mostly) wrong.

The natural fear is a price spike.

When a company enters a major index, every tracking fund has to buy its shares. Billions of dollars in forced purchasing arrive on a schedule the entire market can see.

The straightforward fear is that all that buying pushes the price up, and you, the person whose fund was ordered to buy, end up holding shares at the top of a wave your own money helped create.

That fear has been studied for decades.

In the 1990s, the average price jump from being added to the S&P 500 was about 7.4 percent. By the past decade, it had fallen below one percent. The forced buying still happens. The spike mostly does not.

You might think a thin float makes the spike worse, that fewer shares against the same mandatory buying drives the price harder. NASDAQ thought of that too. When it dropped the ten percent floor, it added a cap in its place. A company with a small float now enters the index at a reduced weight, limited to three times the value of the shares actually trading.

NASDAQ’s own worked example: a trillion-dollar company that floats six percent enters at about one percent of the index, and the forced buying comes to roughly a tenth of the available shares, not some wall of demand crashing into a tiny supply. On the question of supply, the new rule was built with some care. It holds the buying to a manageable slice.

So the supply problem was theoretically handled. What the cap does not touch is the other rule, the one that was removed with nothing put in its place.

The cap limits how many shares must be bought. It says nothing about whether the price those buyers pay is real. That was the seasoning period’s only job, and the seasoning period is gone.

SpaceX could be bought into the index two and a half weeks after its first trade, at a price the open market has had almost no time to test.

And here the research runs out.

Every study of the disappearing spike measured ordinary additions: established companies with years of trading history and valuations that did not reorder the top of the index on arrival.

A $1.75 trillion company entering fifteen trading days after its IPO is a situation none of that research describes, because it has never occurred. The cap may make a runaway spike unlikely. But it does not make the price tested.

This is where the price problem and the profitability problem meet. SpaceX is valued at $1.75 trillion and lost $4.94 billion last year. The valuation is a bet on future earnings. That bet may be right. Starlink is growing. The launch business is real. But a bet is different from a track record, and the difference matters when the money riding on it belongs to someone who never placed the wager.

A profitable company generates returns that flow back to the people who own it. An unprofitable company’s price depends on the next buyer being willing to pay more than the last. When that chain holds, the price rises. When it breaks, the people holding the stock absorb the correction.

If you chose to buy that stock, you chose the risk. If the stock arrived in your retirement fund through a rule change you had no part in, inside a fund you trusted to represent a broad and tested market, the risk was chosen for you.

The breadth you paid for is thinning.

The spike faded for a reason, and that reason reaches deeper than any single listing.

You gave your money to a fund that tracks an index because you trusted breadth to protect you. Owning a stake in hundreds of companies means no single company can sink your retirement. Spread wide enough, one bad year for one firm is a rounding error against the rest. That spreading out is the entire product. It is what let the fund be sold as the safe bet.

The market has been walking away from that promise for years.

The ten largest companies in the Nasdaq 100 currently account for approximately 47% to 51% of the entire index’s total weight.

The ten largest companies in the S&P 500 now carry close to forty percent of its weight, the highest concentration since at least 1972, against an average closer to twenty-five percent over the past thirty years.

Apple, Microsoft, Nvidia, Amazon, Alphabet, Meta. A handful of firms now occupy the space that hundreds once shared.

A fund that holds five hundred companies but keeps two of every five dollars in ten of them does not behave like five hundred companies. It behaves like a bet on those ten, with the other four hundred ninety along for the ride.

And the ten are not strangers to each other. They rise and fall on many of the same forces, and lately on the same one, the wager that artificial intelligence keeps paying off. When that cluster moves together, the fund you bought for its breadth moves with it.

This is also the answer to why the spike disappeared.

New additions stopped moving the index because the index grew so top-heavy that anything entering below the giants barely registers.

The forced buying lands, and against forty percent concentrated at the top, it hardly shows.

The thing that makes the spike vanish is the same thing that hollows out your diversification.

One fact, read two ways. The index is so top-heavy a new giant barely moves it, which is the reassurance. The index is so top-heavy it barely resembles the broad market it promised, which is the harm.

SpaceX lands directly in that cluster, in the indexes that will take it. Its $1.75 trillion valuation (placing SpaceX as the 7th most valuable company on the NASDAQ-100, immediately) is itself an AI bet, built on xAI and the promise of what comes next.

Behind it, Anthropic and OpenAI are purer versions of the same wager, and the rule change was made to bring companies like them to the top of the index faster.

Each one added near the top deepens the concentration and tightens the correlation.

The breadth that was supposed to spread your risk across the whole economy is quietly becoming a single story, told by a handful of enormous companies betting on the same future.

That erosion was never voted on either. It happened year by year, as the giants grew and their share of the index swelled, while the statements kept saying hundreds of companies.

The rule change pours more weight onto the same few names, faster than before, with less time to test their price.

The number on your statement will barely move. That’s the danger.

SpaceX is huge, and it will immediately become one of the biggest companies in the world.

But for one company entering one index, at a capped weight, the effect on your retirement is likely still a fraction of a fraction of a percent. Too small to find on a quarterly statement. SpaceX will, hopefully, not dent your savings.

That is what makes the whole arrangement sustainable. And dangerous.

If the rule change put a visible hole in your retirement, you would notice. You would call someone. You would push back.

The effect is too thin to see on any individual statement, too diffuse to trace back to a single cause, and too structural to opt out of.

The system works because the harm is too small to feel, spread across too many accounts to organize against, and buried inside a product you were told to set and forget.

The money is small on your end. Across millions of accounts, it adds up.

When index funds are ordered to buy, that buying supports the stock price. The people who benefit most from a supported price are the people who already hold the most stock: the founders, the early investors, the insiders who received shares instead of cash through every merger along the way.

Musk is the largest shareholder in SpaceX and the wealthiest person in the world. The forced buying that arrives from retirement accounts across the country helps hold up the price of his holdings, invisible to any single account holder, consequential to the people at the other end.

Wealth flows toward the people already at the top, supported by the retirement savings of people who never chose to invest in his company and never had a say in the rule change that put it there.

Share

While you weren’t looking.

You made the safe bet. Picked a fund that sold you safety through consistency and diversification.

That fund with the date in its name is still the safest bet. You will, and probably still should, choose it again,. And across a working life, it will probably still be a good choice. The math still works. The product still does what it promises: it tracks the market.

The question is what the market has become, and who gets to decide.

The companies that were supposed to represent the breadth of the American economy have consolidated around a handful of firms carrying close to half of the weight of the entire market.

The rules that once controlled which new companies could influence your savings, and how quickly, were rewritten by private companies, approved by fund managers paid to protect your money, and never once put to the people whose retirement savings carry the result.

It did not have to go that way. The same week NASDAQ removed its protections, the S&P kept its own. The rules can be defended.

What the next listing raises, and the one after that, is who gets to make the call, and whether the people whose money is on the line are ever in the room when it is made.

On June 12, a $1.75 trillion company assembled through private mergers, carrying a $4.94 billion annual loss, offering a sliver of its shares to the public, enters a market where your fund will be ordered to buy it.

The effect on your account will (hopefully) be too small to find. But the wealth it supports will flow toward the people who already hold the most. Behind it, the line is forming: Anthropic, OpenAI, and whatever comes next.

The diversification you were promised keeps narrowing into a handful of correlated bets. The risk of holding them was handed to you by people who never had to ask. And the people who benefit get richer by robbing you of your financial consent while gaining a stronger hold on an already consolidated market.

The safe bet still works. The rules buried inside it belong to someone else, and at least one of them can be changed without asking you.

That is what changed while you weren’t looking.

Subscribe now

If this piece made you look at your retirement account a little differently, send it to someone who picked the safe fund, stopped looking, and trusted that the rules would stay where they were. Help a friend take control and learn alittle more about what happens to their money when a big headline like SpaceX hits the market.

Most people are not trying to beat the market. They are trying to make one responsible choice and get on with their lives. That is exactly why this story matters.

The Minuteman grows when readers pass the frame along, one person at a time.

Sharing is caring :)

Share The Minuteman

Subscribe now

A note before we start:

I wrote this because the immigration argument in America is highly emotional and highly illogical. We confuse it with a cost of living crisis. And we blame the wrong people.

It usually sounds like this: Someone is taking your job. Someone is taking your house. Someone is taking your country.

The story comes with a culprit attached, and the culprit is almost never the person who can afford a lawyer, a lobbyist, a private jet, or a seat at the table where the rules get written.

I think that is why this issue is so hard to talk about. The fear is real. The loss is real. The shrinking sense of life is real.

But the pain is misdirected.

This piece is about that misdirection. It is about the distance between who people are told to blame and who has been quietly collecting the country, piece by piece, for forty years.

The Grandchildren of the Minutemen Are Being Asked to Defend the King

This is the economic impact of immigration that the public debate keeps hiding: immigrants are not “taking the country from working Americans.” They are being used to distract from who already did.

Subscribe now

There is a man somewhere in Pennsylvania who is one paycheck from losing the house his grandfather bought outright in 1962.

The grandfather worked in a factory that does not exist anymore. The factory closed because the company that owned it merged with a larger company that merged with a larger company that paid out billions to shareholders and then offshored the line to bring the price of the part down by four cents per unit.

The man does not know any of this. He knows his rent went up. He knows his insurance went up. He knows his grocery bill went up. He knows that when he turns on the television, someone is shouting at him about a caravan.

He is right that something has been taken from him.

He has been handed the wrong thief.

In 1982, employee compensation accounted for 66.6% of American GDP. Corporate profits accounted for 8%.

By the end of 2024, employee compensation had fallen to 61.9% and corporate profits had climbed to 15.85%.

That is not a small adjustment. That is the largest gap between what workers earn and what owners keep since the data has been recorded.

The economist who flagged it called the chart that shows the divergence “the revolution chart.” A friend of hers said the same thing in fewer words. It is the chart that ends with beheadings.

Something is being taken. The fear is not paranoia. The fear is the body registering a fact the news will not name.

The question is who is doing the taking. And the answer, when you actually trace it, is not the family that walked across a desert with two changes of clothes and a phone number written on a piece of paper.

The man in Pennsylvania is being robbed by people who fly.

Consider what the Congressional Budget Office actually projects. Across the 2024 to 2034 window, the recent immigration surge is expected to add $8.9 trillion to American GDP. The same surge is projected to reduce the federal deficit by $900 billion. The labor force in 2033 will be 5.2 million people larger than it would otherwise have been.

The Dallas Fed and the San Francisco Fed have run the numbers from different angles and arrived at the same conclusion. Beginning around 2033, deaths in this country will exceed births. From that point forward, every additional American comes from somewhere else.

Without immigration, the working-age population would have started shrinking in 2012. Hospitals would have closed earlier. Schools would have consolidated faster. Whole towns would have already emptied out.

I think there is something almost embarrassing about how the public conversation skips this.

The numbers are not contested. The Congressional Budget Office is not a left-wing think tank. The Federal Reserve banks of Dallas, San Francisco, Minneapolis, and Kansas City are not advocacy groups. They are running the math the same way an actuary runs the math on a life insurance policy. And the math says the country needs the people. The math says the people are paying in.

The math says the alternative is a slow demographic collapse that no tax policy can patch.

So why doesn’t it land?

Because the man in Pennsylvania is not arguing with the math. The man in Pennsylvania is arguing with a feeling. And the feeling is correct.

He has less than his father had. His father had less than his grandfather had. His son will have less than he has. The chart that the economist called the revolution chart is the shape of that feeling, drawn over forty years.

Share

The feeling is real. The diagnosis was sold to him.

In 1849, a secret society called the Order of the Star-Spangled Banner formed in New York City.

Within a few years it was a national political party, briefly the third-largest in America, holding more than 100 congressional seats and eight governorships. Its members called themselves the American Party. Everyone else called them the Know-Nothings, because when asked what they stood for, they were instructed to say they knew nothing.

What they stood for, in fact, was that Irish and German immigrants, mostly Catholic, mostly poor, were taking the country.

The cartoons of the era depicted the new arrivals as drunken election thieves, hauling barrels of whiskey and stolen ballot boxes. The riots in Philadelphia in 1844 burned two Catholic churches and killed at least fourteen people. By 1855, immigrants outnumbered native-born Americans in Chicago, Detroit, and Milwaukee.

The historian Tyler Anbinder, who wrote the book on the period, points out that the Irish were arriving at ten times their previous rate because they were starving to death at home. The famine had killed a million people. Another million had fled. They came here because the alternative was a coffin.

The Know-Nothings collapsed in 1855, not because their voters changed their minds, but because slavery became a more pressing question than the Pope. The fear they organized never went away. It just waited for the next set of arrivals.

It waited for the Italians. It waited for the Chinese. It waited for the Eastern Europeans. It waited for the Japanese. It waited for the Mexicans. It is still waiting, because the fear is structurally useful, and as long as someone profits from pointing it at the door, someone will keep pointing it at the door.

This is the part of the story that doesn’t get said out loud.

The country has been getting taken for forty years. The people taking it look nothing like the people you have been told to watch.

They look like the executives at the company that closed your grandfather’s factory and used the savings to repurchase its own stock.

They look like the private equity firm that bought your hospital and then cut the maternity ward to improve margins.

They look like the holding company that owns four of the five grocery chains in your county and raised prices in lockstep through three years of cooling inflation.

They look like the landlord who is actually an algorithm operated by a software company that recommends rent increases to thousands of buildings at once.

They look like the lobbyists who wrote the tax code that made labor expensive and equipment deductible.

They look like the donors who funded the campaigns of the politicians who passed the laws that let all of this happen.

I feel that this is the largest political contradiction in the American present, and almost nobody names it.

The movement that has worked itself into a fever about immigration is led, funded, and operated by the same class of people whose policy preferences have produced the wage compression, the housing crisis, the medical debt, and the corporate consolidation that are actually grinding the man in Pennsylvania down.

The immigrant is the substitute. The substitute does not have a lobbyist. The substitute does not own a network. The substitute cannot fight back. The substitute is what you blame when you cannot afford to blame the person who is actually doing it, because the person who is actually doing it owns the building you work in.

The Substitution is what happens when an economic injury is real, the cause is politically protected, and the public is handed a powerless body to blame instead.

And there is a second contradiction inside the first one.

The Minutemen were the border crossers.

The men who fired the first shot at Lexington in 1775 were not defending a border. They were defying a king.

The country that those men founded did not exist yet. There was no America to take. There was an empire, and there were colonists, and the colonists decided that the empire’s authority over them was illegitimate, and they took agency from a tyrant by force.

The Declaration of Independence is the most famous document of unilateral border defiance in modern history. It is a piece of paper in which a group of people inform a distant power that they are no longer governed by it.

They wrote the paper because they were already breaking the law. The paper made the breaking official.

The flag that the modern restrictionist movement waves was designed by people who would not have qualified for a green card in their own country.

The Minutemen the right invokes were not standing guard at a frontier. They were the frontier. They were the people the empire was afraid of.

The empire called them lawless. The empire was correct. They were lawless. That was the point.

Lawful inside the empire’s terms would have been to remain subjects. They refused.

The movement that now claims their name has inverted everything they did.

It calls itself the party of the founders and operates as the party of the king.

It uses the language of independence to defend concentrated power.

It invokes the flag of revolution to oppose the conditions under which revolution becomes thinkable.

The contradiction is so total that it would be funny if it were not currently being used to detain children.

I am not saying the man in Pennsylvania is a fool. I am saying he has been handed a story that explains his pain in a way that protects the people causing his pain. The story is older than him. The story has been told before, in slightly different costumes, every time the actual taking has gotten this bad.

The Know-Nothings told it in 1854. The Chinese Exclusion Act told it in 1882. The Immigration Act of 1924 told it about Italians and Jews and Slavs. The story always works for a while, and it always ends badly, and the people who profited from the story always survive the ending.

Share

What gets remembered backward is what allows the present to keep happening.

If the founding is the founding of a closed country, then closing the country is patriotic. If the founding is the founding of an open one, then closing the country is the betrayal.

The history is unambiguous on which it was.

The colonies were demographically dependent on continuous arrival. The country that wrote the Constitution was 80% rural, mostly immigrant or one generation removed, and structurally hostile to the very idea of a sealed border, which was both impossible to enforce and contrary to the economic interests of the seaboard merchants who needed the labor.

The first federal immigration law of any consequence was not passed until 1882, more than a century after independence. For its first hundred years, the United States operated on something close to an open door, not because its leaders were sentimental, but because the math required it.

The math still requires it.

The man in Pennsylvania needs more neighbors, not fewer. He needs the tax base that comes with population growth. He needs the customers for the small business his daughter might open. He needs the workers who will keep the hospital where his mother is dying from closing its emergency room. He needs the construction crews that might bring the housing supply back into something resembling balance with demand. He needs the agricultural labor without which his grocery bill, already crushing, would double. He needs the demographic floor under his Social Security check.

He is being told that he needs a wall.

I think the fear that has been organized around immigration in this country is one of the most successful political operations of the last half century.

I think it has succeeded because the people running it correctly identified that the man in Pennsylvania feels something has been taken from him, and they correctly identified that he was going to look for somewhere to put the feeling, and they made sure the place he put it was not their door.

I think the people who designed this operation know exactly what they are doing. I think a great many of the people repeating it do not. I think that distinction matters, because the second group is reachable, and the first group is not.

The fight that actually decides whether the man in Pennsylvania keeps his house is not at the border. It is at the IRS. It is in the tax code. It is in the antitrust division of the Justice Department. It is in the labor board. It is in the rules that govern how easily a company can break a union, how quickly capital can be written off, how much medicine can cost when one company owns the patent and the pharmacy and the insurer all at once.

None of that fight has anything to do with the family from Honduras. The family from Honduras is mowing the lawn of the man who is suing to keep wages low. The family from Honduras is not the man’s enemy. The family from Honduras is, in the actuarial sense, the only thing keeping the man’s Social Security check from arriving smaller each year for the rest of his life.

I do not know how to make this land in a country where the loudest voices have a financial interest in keeping it unlanded.

I do not know how to undo forty years of substitution in a single article, or a single election, or a single generation.

I am not sure anyone does.

But I know what the math says. The math says the country is not being taken by the people walking in. The math says the country was taken, slowly and legally, by the people who own it. The math says the fear is real. The math says the diagnosis is wrong.

The Minutemen took agency back from a king.

The grandchildren of the Minutemen are being asked to defend the king.

Share The Minuteman

If this piece helped you think about immigration, wages, and power a little differently, send it to someone who feels that something has been taken but has not been given the right map.

And if someone sent this to you, thank them.

The Minuteman grows when readers pass the frame along, one person at a time.

Sharing is caring :)

Share

A note before we start.

The AI conversation almost everywhere I look is happening in pieces.

One week it’s a story about your electric bill. The next week it’s a story about ICE using a private government database. The week after that it’s a story about an executive order opening federal land to data centers. Then a story about Three Mile Island reopening, then a story about tax breaks in Texas, then a story about layoffs at a tech company or a law firm.

Each story arrives on its own, with its own headline and its own outrage cycle, with its own three-day half-life, and none of them are ever connected to each other.

These are all the same story.

The same companies are the customers, the suppliers, landlords, and the regulators. The same servers are running the chatbot in the deportation map. The same ratepayer is paying for the substation that feeds the building that holds the database that the federal government uses to find people. The loop closes on itself. The loop is the point.

What we are calling AI is the largest peacetime transfer of state capacity to private hands in a century, and the cover story is wearing thin enough that you can see the structure underneath it if you stop reading each piece on its own and start reading them together.

This is what I tried to do for you here.

Subscribe now

The data center, the electric bill, the federal contract, and the surveillance state have become one machine, and almost no one has explained how the pieces connect.

The Loop

The AI conversation, almost everywhere you look, is happening in pieces.

One week it is a story about your electric bill. The next week it is a story about ICE using a private database to map neighborhoods. The week after that it is a story about the federal government opening military bases to data center developers. Then a story about Three Mile Island restarting. Then a story about a tax break in Texas. Then a story about a law firm replacing half its associates with software. Each story arrives on its own, with its own headline, its own three-day cycle, its own comment section. None of them ever sit on the same desk at the same time.

They are the same story. They are one loop, and you are inside it.

Your electric bill is going up because of the data center. The data center is being built because the federal government designated it a matter of national security. The national security designation exists because the data center runs the software the military uses to identify targets and the software ICE uses to find people in your neighborhood. Because that software is now woven into law enforcement and defense, the data center cannot be turned off, which means the energy demand cannot come down, which means your bill cannot come down. The companies profiting from this spend millions lobbying Congress for laws that keep the cycle going, including tax incentives that make it cheaper to replace workers with machines, which means the bill that keeps climbing is harder for you to pay. And the workers being replaced are the same people whose government records, their Social Security files, their tax returns, their Medicaid data, are being fed into the same system that tracks and removes people.

That is the loop. Everything that follows is the proof.

Your electric bill went up. You probably blamed the weather.

Somewhere in Virginia (or New Jersey), a building the size of a regional airport is drawing the equivalent of a small city’s worth of power. It does not have employees on most floors. It does not have windows. It has servers, and the servers are computing something for someone in a way that requires more electricity than some countries will use in a year.

In the United States right now, data centers account for roughly half of all new electricity demand. Residential electricity prices have risen forty percent since 2021. In Virginia, where data centers consume more than a quarter of all power generated in the state, the regional grid auction price jumped 833 percent in a single year. A homeowner in Manassas opened his January electric bill and found $281 where the previous month had been $100. He has lived in the house for forty years. The house did not change. The grid did.

Between 2020 and 2024, residential electricity rates rose twenty-five percent. Commercial rates rose three percent. Industrial rates fell two percent. The people consuming the most power got the biggest discount. The people consuming the least absorbed the cost.

That cost has a cause, and the cause has an address, and the address was chosen for you by a chain of decisions that leads directly to Washington.

The federal government cleared the land.

The word “cloud” was always a marketing decision. Clouds are weightless and free and belong to no one. The actual infrastructure is heavy, expensive, and belongs to four or five companies. It needs land. It needs power. It needs water. It needs permits. And starting in 2025, it got all of them at once.

In July 2025, the White House issued an executive order designating large data centers as projects of national security importance. That designation accelerated federal permitting. It opened federal land for construction. It directed the Department of Commerce to provide loans, grants, tax incentives, and offtake agreements. It told environmental reviewers to find categorical exclusions that would let projects skip the usual public process. The day after the order was signed, the Department of Energy named four federal sites available for data center development. By October the Air Force had opened five military bases to bids. The Idaho National Laboratory, Oak Ridge, the Paducah Gaseous Diffusion Plant, Savannah River, Edwards Air Force Base. The federal government became a real estate broker for the AI industry inside three months.

Then came a second executive order in December. This one directed federal regulators to identify state AI laws considered “obstructive” and to deny those states federal broadband funding. States that tried to write their own rules could lose money the federal government had previously committed to give them.

At least thirty-six states had already built subsidies of their own. Virginia gives up close to a billion dollars in sales tax revenue every year and does not disclose which companies benefit. Texas revised its annual subsidy estimate from $130 million to $1 billion inside the same fiscal year. Twelve states with data center programs do not publish what they cost.

Then came the One Big Beautiful Bill, signed on July 4, 2025. Among its provisions: permanent 100% bonus depreciation for data center hardware. GPUs, CPUs, networking equipment, HVAC systems, electrical infrastructure. Buy a server, deduct the full cost in year one. The law also restored R&D expensing and expanded interest deductions for debt-financed data center projects. The effect was a twenty percent discount on the cost of building a data center, written into the tax code.

The seven largest tech companies are now projected to spend $740 billion in capital expenditures this year. The land was cleared. The permits were fast-tracked. The subsidies were stacked. The tax code was rewritten. Your bill went up. Now look at what the buildings are running.

The same servers run the chatbot and the deportation map.

The data centers being built on federal land, subsidized by federal money, powered by a grid you pay into, do not just compute. They run enforcement.

Palantir, which has provided pattern-matching software to U.S. Immigration and Customs Enforcement since 2008, holds over $145 million in ICE contracts. Its platform, ImmigrationOS, pulls data from passport records, Social Security files, IRS tax returns, license-plate readers, and the Department of Health and Human Services to map neighborhoods, build dossiers, and rank deportation targets in something close to real time. A separate tool, ELITE, lets agents enter a zip code and receive a heat map of where suspected noncitizens are likely to be found. ICE’s Palantir system runs on Amazon Web Services. Amazon is the same company building data centers on federal land under the same executive order that designated AI infrastructure as a national security priority.

The data flowing into these tools is growing. In April 2025, the Department of Government Efficiency pushed through a data-sharing deal with ICE over the objections of top IRS officials. DOGE is building what CNN has described as a “master database” for immigration enforcement, pulling records from the IRS, the Social Security Administration, and Health and Human Services into a shared repository. Palantir is constructing the API that connects it. Medicaid data, originally collected to administer health coverage, has reportedly been routed into the same system. The IRS data that helps catch a tax fraud is now helping locate a deportation target. The Social Security record that tracks a retirement benefit feeds a platform designed to remove the person who earned it.

All of this runs on the same cloud infrastructure, in the same data centers, drawing from the same grid, subsidized by the same tax code, that your electric bill is paying for.

The same infrastructure runs the military. That is why it can never be turned off.

The national security designation was not a figure of speech. The Pentagon signed agreements with Microsoft, Amazon, Google, OpenAI, NVIDIA, and SpaceX to deploy AI on its classified networks. These are Impact Level 6 and Impact Level 7 environments, meaning secret and highly restricted data running on commercial cloud platforms. Microsoft operates dedicated DoD cloud regions in Des Moines and Northern Virginia. Amazon runs similar classified environments at undisclosed locations. The Joint Warfighting Cloud Capability program, a $9 billion contract awarded to Google, Oracle, AWS, and Microsoft, has already issued more than $3 billion in task orders.

Palantir’s Maven Smart System, an AI targeting platform that uses computer vision and machine learning to identify objects in satellite and drone feeds and flag targets for human approval, was designated an official program of record by the Pentagon in March 2026. The designation mandates integration across all five branches of the military by September 2026. The contract ceiling was raised to $1.3 billion in 2025. A separate $10 billion Army enterprise agreement, signed in July 2025, consolidated 75 existing Palantir contracts into a single framework. Scale AI holds a $500 million Pentagon contract for military data processing. Microsoft holds a $170 million sole-source Air Force cloud contract awarded in January 2026. In late March 2026, the Army selected companies to build hyperscale data centers on military installations at Fort Bliss and Dugway Proving Ground, backed by Wall Street financing.

Then the question of whether data centers are military infrastructure stopped being theoretical. In March 2026, Iranian drones struck Amazon Web Services data centers in Bahrain and the United Arab Emirates. It was the first time in history that a nation deliberately targeted commercial data centers during wartime. The strikes cost Amazon an estimated $150 million in service refunds alone. The reason they were targeted: the same facilities that run civilian messaging apps, banking systems, and streaming services also handle defense workloads. An enemy does not need to distinguish between your Netflix account and a military targeting system when they share the same building.

This is the lock. A data center that runs a chatbot can be turned off. A data center that runs a chatbot, a deportation map, and a military targeting system designated as a program of record across all five branches of the armed forces cannot. The energy demand is permanent. The coal plants from the 1970s being held open past their planned retirement are permanent. The nuclear restarts, including the second reactor at Three Mile Island, are permanent. The country needs an estimated forty-nine additional gigawatts of generation through 2028 to meet projected demand, and that projection assumes no further expansion of military or enforcement use.

Your bill cannot come down because the system it powers cannot be switched off.

The companies that profit from the lock lobby to keep it locked.

Palantir’s U.S. government revenue grew 104 percent year over year in the first quarter of 2026. The company is sole-source for ICE. It is a program of record for the Pentagon. It is building the API for the DOGE master database. Its contracts are structured so that the agencies it serves cannot replace it without losing their ability to perform their own core functions.

The eleven largest tech companies spent $20 million on federal lobbying in the first three months of 2026. That is $226,000 a day. AI-specific lobbying totaled $92 million in the first three quarters of 2025. AI companies have funneled nearly $200 million into super PACs ahead of the 2026 midterms. During the first half of 2025, the industry pushed a provision in the president’s spending bill that would have stripped states of the power to regulate AI for ten years. The Data Center Coalition, the trade group representing data center developers, posted its biggest-ever lobbying quarter in early 2026, more than tripling its spending year over year.

The lobbying produces legislation. The legislation produces subsidies and tax breaks. The subsidies and tax breaks produce more buildout. The buildout produces more demand on the grid. The grid demand produces higher bills. And it produces something else.

The same tax code that subsidizes data center construction makes it cheaper to replace a worker with a machine than to keep the worker. A company that buys a server can deduct the full cost in year one. A company that hires a person absorbs payroll taxes, benefits, training costs, and legal exposure with no equivalent write-off. Economists Daron Acemoglu, David Autor, and Simon Johnson have argued that the U.S. tax code places a heavier burden on firms that hire workers than on firms that invest in automation. The largest tech companies saved an estimated $14 to $18 billion on their tax bills in 2025 from the bonus depreciation provision alone. OpenAI itself has proposed shifting the tax base away from payroll and toward capital gains, and levying taxes on automated labor, because even the company building the replacement technology recognizes that the current incentive structure is accelerating displacement.

The person whose electric bill is climbing is also the person whose job the tax code is helping eliminate. The bill goes up. The paycheck goes down. The loop tightens.

The people being replaced are the people in the database.

The workforce the AI industry needs to build the physical infrastructure, the electricians, the HVAC technicians, the controls engineers, cannot be automated. The White House AI Action Plan published in July 2025 calls for federal apprenticeships in skilled trades for exactly this reason. The country does not have enough electricians to wire the data centers the industry says it needs. The hands that build the boxes will be human.

The work being done inside the boxes, increasingly, will not. And the workers being displaced by that work are not an abstraction. They have Social Security numbers. They have IRS records. They have Medicaid files. Those are the same records now being routed through DOGE into a master database, connected by a Palantir API, fed into ImmigrationOS, and used by ICE to locate people.

The same system that replaces a worker also catalogs that worker. The same government data that tracked their earnings, their health coverage, their retirement contributions, now feeds an enforcement tool running on infrastructure their electric bill helps pay for. The person is on both sides of the loop. They fund it and they fill it.

The loop.

A surveillance state used to require informants. A notebook. A filing cabinet. A clerk to file the paper. A courier to deliver it. A supervisor to read it. A decision to act. The cost of finding a person was high. The cost of building a list was higher. The cost of acting on the list was higher still. That cost was the friction inside which civil liberties survived.

The cost is now near zero. The filing cabinet is a cloud platform. The clerk is an algorithm. The courier is an API. The list builds itself. And the person paying to keep the system running is the person on the list.

Your electric bill funds the grid that powers the data center. The data center was built on federal land with federal permits and federal subsidies and a federal tax break. The data center runs the software that uses your government data to map your neighborhood. The software is sole-source, integrated into law enforcement and defense, and cannot be replaced. Because it cannot be replaced, the data center cannot be turned off. Because it cannot be turned off, the energy demand cannot come down. Because the demand cannot come down, your bill cannot come down. The companies that profit from this lobby Congress for more subsidies, more tax incentives, more buildout. Congress passes the laws. The laws include provisions that make it cheaper to replace you with a machine. The machine that replaces you runs on the same data center. Your employment records, your tax returns, your health data, flow into the same system. The loop closes.

Every week another story arrives. An electric bill. A deportation raid. An executive order. A tax provision. A lobbying disclosure. A contract award. A layoff.

They are the same story. They have always been the same story. The building you cannot see is the thread that connects them, and the light you left on in your house is a small, forgettable fraction of what it costs to keep running.

Someone, somewhere, is counting on you to keep reading them as separate headlines.

Subscribe now

If this piece made you look at your electric bill, your phone, or the word “cloud” a little differently, send it to someone who might benefit.

Someone who talks about AI like it lives nowhere. Someone who talks about electricity prices like they came from nowhere. Someone who knows something about this new world feels off, but has not been handed the chain yet.

The Minuteman grows when readers pass the frame along, one person at a time.

Sharing is caring :)

Share

Two nations declared victory in the same war on the same day. The bombs kept falling. The prices kept climbing. And the rest of us are still paying for a win that nobody can find.

Something about the way we fight wars broke after 1945, and we never learned the new rules. So how do you win a war?

Subscribe now

You turn on the TV and some politician is talking about U.S. troops overseas. Maybe they’re talking about the cost or a budget. And after listening for a while, you ask yourself, “Which country are they talking about?“

You can think of at least three potential answers off the top of your head. One started under a president you voted for, another started under a president you didn’t, and a third one you can’t remember when it started at all.

But the common thread is none of them have seemed to end.

Meanwhile there’s an endless supply of debate. Prices keep rising. Gas costs more now than it did a month ago. Your grocery bill keeps climbing. Utilities are through the roof. And it seems impossible to wrap our heads around why or how any of this started in the first place.

I’m writing this as headlines flash in front of my eyes. And what I am seeing seems like two contradictory stories from two completely different sides of the spectrum. Neither of them seems grounded in reality.

Trump said we won the war in Iran. Iran said they won the war with us. Yet bombs are still falling, people are still dying, and those prices are still rising.

So either somebody is lying to us, reality is not what it seems, or the word “winning” doesn’t mean what it used to.

This is the age of the forever war. And we didn’t just get here. We’ve been living in it for over 80 years.

If you want to know why every conflict the United States enters overseas seems to feel like a never-ending war that we can’t seem to win, then we have to ask ourselves an important question.

How do we win a war in today’s age?

We are still trying to win wars with a pre-nuclear mind.

Before 1945 you fought a war like a heavyweight boxing match. Two states step into a ring, a referee, a center of gravity. You aim for the weak spots. You try to get a knockout.

Carl von Clausewitz, a Prussian army officer of the 19th century, called it the “Center of Gravity” (Schwerpunkt, German words are fun). It’s the hub of all power and movement on which everything depends, your opponent’s primary source of strength, the thing that holds the combatant’s system together.

The old principles of war were designed around this objective. You mass enough troops. You outmaneuver the enemy. You overwhelm them at the point that matters. Once the machine is broken, the war is over.

You took the capital. You broke the factories. You signed the papers. You threw a parade.

You’ve probably seen the famous photo from Times Square. A US Navy sailor dipping a stranger (a dental assistant, Greta Zimmer Friedman) in a kiss on the celebration of V-Day.

That was how a war ended, with a parade.

We danced around the ring, raised our fists in victory. The confetti fell and we got handed the heavyweight belt.

But we weren’t just celebrating the end of a war that day. We were celebrating the end of “the end” of war.

While we were busy parading around the ring, celebrating our victory, what we failed to realize was that in that victory we had built a glove that vaporizes the arena.

The same month that sailor and that dental assistant kissed on the celebration of V-Day, we had dropped bombs on Hiroshima and Nagasaki that killed an estimated 150,000 to 250,000 people.

The nuclear revolution didn’t end interstate conflict as it was once predicted. It just changed the game we were playing.

When two superpowers have nuclear weapons that can knock the other out entirely with missiles and submarines scattered across the oceans that neither side can fully map, then any total war between them is mutually assured destruction.

Now the idea was that this mutually assured destruction would prevent either side from initiating conflict. But that clearly didn’t happen.

Instead we put down our gloves and stepped out of the ring. We migrated into the parking lot of the bar across the street.

We turned to proxies, to insurgents, to civil wars, and no one could decisively end them. The rivalry remained. The ring disappeared. The fighters started using new equipment with no clock, no referee, and very different rules.

That migration is when the center of gravity moved.

The capital city is no longer the thing that wins or loses a war. The phone in everyone’s pocket is.

War today is more about public morale and international legitimacy than it is about overwhelming and overpowering your enemy. It’s about perception.

Think about it like this. You see a video on the news of a low-ranking soldier in an international conflict doing something terrible, committing a war crime. That video spreads across the world faster than that soldier’s commanding officer can respond. Within a day it’s on everyone’s phone. It’s playing on every TV. It’s being talked about across the world.

The United States Marine Corps four-star General Charles Krulak named this figure the Strategic Corporal: one private with a camera can now produce consequences that a four-star general’s entire career could not.

A state now loses faster from a single image than from a defeated army.

Take the Russian invasion of Ukraine. Which party in that conflict have you heard more horror stories about on the news?

American voters tend to lack the endurance for prolonged international conflict, especially when the perception is that US troops are not actively involved and our main contribution is financial. But the more we learn about the horrors committed by this invading force, the more willing we are to support a nation’s sovereignty to fight back against bad actors.

The way we measure victory is broken.

We used to measure wins by counting: damage delivered, targets destroyed, infrastructure flattened, leaders eliminated.

Today a different kind of winner counts survival. We’re still here. The campaign against us did not erase us therefore we won.

A state can claim victory by counting damage done or a smaller opponent can claim victory by not getting knocked down.

Ever seen Rocky? He loses the fight but he doesn’t care. He’s still standing there, screaming for Adrian. Watch Rocky and tell me who you think wins that fight. An argument can be made for either side.

That’s the same problem we have at war today. The two sides are no longer measuring the same thing so both sides can declare victory in the same war on the same day.

This is what’s happening right now between Trump and Iran.

On March 11th at a rally in Hebron, Kentucky, Trump said, “We won.”

Here is exactly what Trump said: “You never like to say too early you won. We won“; “In the first hour it was over.” “We got to finish the job.”

On April 7th after a “ceasefire,” he called it “total and complete victory.”

“Total and complete victory. 100 percent. No question about it.” – Trump

On May 1st in a letter to Congress, he declared the hostilities terminated.

“The hostilities that began on February 28, 2026, have terminated.”

At the same time Iran announced victory, claiming it had forced America to accept Iran’s own ten-point proposal.

Do you see a problem here? Are we fighting two different wars? Are we existing in two separate realities? This is the same day, the same war, two different governments both claiming the win.

And then six days after the president told Congress that the war was over, U.S. and Iranian forces exchanged fire near the Strait of Hormuz. The day after that U.S. strikes hit Iran-linked vessels and UAE air defenses engaged Iranian missiles and drones.

As of May 14th, a U.S. admiral told the Senate Iran retains the capability to threaten Gulf shipping and neighboring states, and violence has periodically flared since the “ceasefire.”

The hostilities that both sides claim to have ended and both sides claim to have won have not stopped killing people.

And yet both sides have an argument for victory.

The problem is they’re using two different metrics.

Trump is measuring damage. Iran’s ability to threaten its neighbors has been, in the Admiral’s word, “dramatically degraded.”

Iran is measuring survival. The state still functions, the leadership is intact, the Strait of Hormuz, through which about a fifth of the world’s oil moves, is still contested.

And the rest of us still have to pay for it.

Gas prices averaged around $2.94 in late February. This week (the week I’m writing this, May 15th), gas prices average over $4.50. The Bureau of Labor Statistics puts the gasoline index up 28% over the past year. And prices everywhere else are following. Go to the grocery store and tell me that things haven’t gotten more expensive.

We have over 50,000 troops in the Middle East. The 82nd Airborne deployed a brigade in late March. The Marines moved an Amphibious Ready Group. By April there were three carrier strike groups in the region. There has been no verified drawdown since the President’s May 1st letter.

The country that “won” is still showing up to the fight.

But Iran is just one room. Look at all the others.

We are in Iraq, over twenty years after the invasion.

We just pulled the last of our roughly 1,000 troops out of Syria in April after more than a decade fighting ISIS alongside Kurdish allies. No parade for that one either.

We are in Somalia. The US carried out 49 air strikes in 2026, so far, on top of 124 strikes in 2025.

These are five, ten, twenty-year-old wars the country has stopped watching. They did not end. They got quiet. And a parade didn’t come after them either.

We are addicted to the ring.

The United States of America cannot let go of the heavyweight boxing match because the heavyweight match is who we think we are.

V-Day is an American identity. The soldier in uniform, the kiss in Times Square, the ticker tape: we have been trying to replay that ending for 80 years because we do not know how to be a country that does not get one.

Political scientist Dominic Tierney put it best: Before 1945 the United States won virtually all of its major wars. After 1945, with the strongest, best-funded military in human history, clear-cut victories disappeared. The country got more powerful and less able to produce the only result it recognizes.

So we walk into every new conflict throwing punches in a ring that is not there, against an opponent that has disappeared, still listening for a bell that’s not gonna ring, still expecting the parade and the belt.

And every time we get tired before the other side does, the news moves on. We pull our attention away from the room. Depending on the politics, depending on how we can use this situation to benefit us, we call it losing or winning.

The reality is this is an outcome that our country doesn’t have a name for.

The harder we try to win the old way, the more enemies we make and the more legitimacy we lose, and the longer the war runs on. Overwhelming force against an insurgency can destroy a thousand fighters and create ten thousand recruits.

Sri Lanka “destroyed” the Tamil Tigers in 2009 and it cost tens of thousands of civilian lives and a generation of international condemnation.

The British “defeated” the Malayan Emergency in the 1950s by interning hundreds of thousands of ethnic Chinese and starving the insurgency of support. The success the British claimed did not truly end the conflict. A second communist insurgency began in 1968 and lasted until 1989, nearly three more decades of armed conflict before Malaysia achieved “lasting peace.”

The victories these countries have claimed ultimately spiraled into decades-long, never-ending, generational conflict and the United States falls into the same trap.

Both of those victories ended the fighting. Neither ended the conflict.

There is a distinction we keep missing.

War termination is when the shooting stops. Conflict termination is when the thing that caused the shooting goes away. Those are not the same event, and most of the time they don’t happen in the same century.

British General Rupert Smith spent his career watching that distinction get ignored. He served in the Gulf, in Northern Ireland, in the Balkans. He came out of it with a phrase that explains what the old playbook keeps getting wrong: war amongst the people.

The fight is no longer between states with armies. The fight is in living rooms and on phones, in the space where a population decides whether the cause is worth the cost.

Smith was pointing at where the real target lives. In belief. In a population’s running calculation about whether the cause is worth what the cause costs.

Which is why you can win every battle and still lose the war.

In Israel, Netanyahu’s own spokeswoman declared in October 2025 that all of the prime minister’s military objectives in Gaza had been achieved. Netanyahu himself said on the same day that “there is no end to the war yet.” The military objective was met. The political stalemate did not move.

In Afghanistan, twenty years of “operational success” ended with a withdrawal that handed the country back to the same people the U.S.A. showed up to defeat. In Iraq, “Mission Accomplished” was six weeks in. The U.S. spent the next twenty years losing the country.

The military problem and the political problem are not the same problem, and we keep solving the first one and calling it the second.

Strategists have a name for this. They call it the paradox of strategy. The quicker and more decisive the military victory, the worse the long-term outcome can be, because nothing about the underlying grievance got touched.

The enemy gets defeated. The reasons the enemy existed in the first place stay intact. So another enemy shows up. And another.

What modern war runs on is belief. A population fighting for what it considers a just cause is not defeated when its army is. It is defeated when it stops believing the fight is worth what the fight is costing. Until then, every funeral makes a recruit.

The bell isn’t going to ring. The audience is the war.

If you want to see all of this happening at once, look at Ukraine.

In February 2022, Russia tried to win a war the old way.

The plan, according to RUSI’s early-war assessment, was to invade over about ten days, seize Kyiv, collapse the government, and wrap the whole thing up by August. Classic schwerpunkt.

Russia treated Ukraine’s center of gravity as the capital and the president: take the city, remove the leader, and the country folds. They even achieved a 12:1 force-ratio advantage north of Kyiv.

By every principle of old war, that should have been enough.

Ukraine’s center of gravity was spread across the entire country: national will, military adaptation, the Western support that RUSI called Ukraine’s “strategic depth,” and a population that decided the fight was worth the cost.

Russia punched the capital and the country absorbed the blow.

Ukraine’s air defenses partially reconstituted by the ninth day. Russia never got the air superiority it needed for the quick kill. The knockout failed, and the war became something else entirely.

Russia’s nuclear arsenal did one thing well: it kept NATO from entering the war directly.

No NATO troops fought in Ukraine. But NATO and EU countries kept Ukraine alive through weapons, training, intelligence, and money. NATO later created mechanisms to make that assistance long-term and stated that Ukraine’s future was in NATO.

The nuclear threat stopped a direct war. Everything short of a direct war kept going.

Once the quick victory failed, the war became war amongst the people. Russia shifted toward grinding civilian infrastructure. Power grids. Heating systems. The conditions that let a country function as a country.

The U.S. Army described electricity in Ukraine as both a tactical necessity and a strategic target. The goal was to break morale, to make the Ukrainian population decide the fight was no longer worth the cost. The target was belief.

Ukraine responded with its own version of the same logic.

It started hitting Russian oil refineries, knocking out roughly 700,000 barrels a day of refining capacity across 16 refineries. Oil and gas taxes are about a quarter of Moscow’s federal revenue.

Ukraine was doing what the old playbook cannot imagine: making the war expensive enough, visible enough, and painful enough for the other side’s population that the calculation starts to shift.

That is perceived worthiness as a weapon. You aim at the other side’s belief that the fight is worth what it costs.

And the paradox of strategy is visible too. Even if Russia had taken Kyiv in ten days, then what? Occupy a country of 44 million people who just watched you invade?

The root political problem, the thing that caused the war, doesn’t get solved by a fast military win. It gets worse. Every funeral makes a recruit. Russia would have terminated the war and guaranteed the conflict would last a generation.

So Ukraine has been fighting the new way. Denying Russia its political objective. Sustaining international support. Imposing cost. Managing the long grind. Russia has not won. By the old scoreboard, that means nobody has. By the new one, that is what winning looks like.

But here is the cost of the new way of winning.

A Gallup poll from July 2025 found that 69% of Ukrainians favored a negotiated end to the war as soon as possible. In 2022 that number was almost reversed.

The strategy is working. The population is exhausted. Both of those things are true at the same time. You can win the new way and still watch the public will to keep winning erode underneath you, because democracies are bad at patience and modern war runs on exactly that.

The way wars work changed in 1945 and it changed again with the internet and then even faster with the smartphone but this country’s idea of victory is still set on the parade.

So how do you win a war?

You learn to manage it.

Winning today is the careful management of a difficult situation indefinitely: a long quiet, a maintained equilibrium, fewer dead, fewer escalations, fewer reasons for the conflict to spread.

General Charles Dunlap calls this “culminating power”: deciding what is enough and stopping there. Stopping may never give us another V-Day parade but going further almost always makes it worse.

The problem is we do not know how to do that yet. Our muscle memory is set to V-Day. We’re still throwing the kind of punch that wins a heavyweight match in a parking lot against an opponent who only has to be standing when we get tired and go home.

Look back at the news: two nations are claiming victory in a war that is not over. Bombs are still falling, the prices are still climbing, the troops are still arriving, and we feel like we are losing because the only kind of winning we know how to recognize ends in a parade.

Until we let go of the parade, the parade is the thing beating us.

Share

If this piece made you think differently about what it means to “win” a war, send it to someone who might benefit.

And if someone sent this to you, thank them.

The Minuteman grows when readers pass the frame along, one person at a time.

Sharing is caring :)

Subscribe now

A note before we start

Affordability has become one of those words that means everything and nothing at the same time.

Every campaign promises it. Every administration talks about it. Every family measures it at the counter, in the inbox, on the app, in the envelope they already know they do not want to open.

The standard story says this is about inflation, or bad budgeting, or one party’s failures, or the other party’s hypocrisy. Some of that may be true in pieces. But the larger pattern is harder to fit into a campaign ad.

The price of being alive in America is not only an economic problem. It is a governing system.

That system needs a name.

Subscribe now

There is probably an unopened envelope on a counter in your house. You know what’s in it before you open it, which is part of the reason you haven’t.

It is a medical bill, or an electric bill, or a rent notice, or an insurance statement, or a school invoice, or a property tax adjustment. The number inside is bigger than it should be, bigger than it was last year, bigger than the version of you who first signed up for the service was told it would ever be. You will pay it. You will be a little angrier than the last time. You will not, on the day you pay it, do anything else about it. There is no anything else to do.

That envelope is the smallest unit of American politics in 2026.

I want to suggest something that sounds like a complaint but is actually a description. The price of your hospital visit, your apartment, your insulin, your electricity, your child care, and your tuition all rose for the same reason, and the reason is not inflation.

It is that the people who set those prices have captured the institutions that were supposed to check them.

The result is a country that runs on a quiet bargain almost no one would sign if it were written out plainly. Government promises affordability. Government delivers, instead, a stable environment in which the cost of staying alive can be raised every year by the people best positioned to raise it. Then it runs a campaign about how angry you are about prices.

This is not a failure of governance. It is a form of governance.

The system is succeeding. Not at the thing it advertises. At a different thing, which it does not advertise, but which can be reverse-engineered from the results.

If you want to know what a system is for, look at what it produces.

American health care produces the most expensive medical outcomes in the developed world and the highest per-capita administrative overhead, and that is not a bug. It is what you would build if your goal were to keep a hospital sector, an insurance sector, a pharmaceutical sector, and a private equity sector all profitable at the same time, with patient outcomes as a constraint rather than a goal.

American housing produces a permanent shortage in the cities where the jobs are, and that is not a bug either. It is what you would build if your goal were to protect the asset values of the people who already own homes in those cities.

American energy produces electricity prices that rise faster than inflation in a country sitting on more natural gas than it can use, and that is what you would build if your goal were to let utilities earn a guaranteed rate of return on capital expenditure no matter what the capital expenditure was for.

None of this is hidden. The rate cases are public. The zoning hearings are public. The pharmaceutical benefit manager contracts are not public, but their existence is. The capture is on the record. It is just not on the news.

Here is the part that I think matters most, and the part I have not yet figured out how to say without sounding like I’m scolding someone, which I am not. Both parties run on affordability. Both parties have run on affordability in every election I have been alive for. And in the period during which both parties have been running on affordability, the share of household income going to housing, health care, child care, and education has gone up, not down.

That is not a coincidence and it is not a mystery. It is what happens when the price-setters and the rule-writers are the same people, and when the rule-writers’ campaigns are funded by the price-setters, and when the voters’ only available response to the resulting prices is to vote for the next round of the same people.

Gallup found in February that twenty-nine percent of Americans named government and political leadership as the most important problem facing the country, the largest single block of responses and the continuation of a pattern that has held for most months since the current administration took office. That is not a number about ideology. It is a number about exhaustion.

Pew found a few months later that the issues Americans say they care most about cluster tightly around the same theme: money in politics, the affordability of health care, inflation, partisan dysfunction, and the federal deficit. You can read those numbers as five issues, or you can read them as one. I read them as one.

Americans have noticed that something is wrong with the price of being alive in America, and they have noticed, separately, that something is wrong with the institutions that were supposed to address it, and they have not yet been given the language to connect the two noticings.

Call it the affordability state.

A country in which the formal job of government is to deliver services and the actual job of government is to absorb the cost of protecting whoever already has market position.

The affordability state is not a conspiracy. It does not require a meeting. It requires only that the people who run the relevant committees, who write the relevant regulations, who sit on the relevant boards, and who fund the relevant campaigns share a set of incentives that point in the same direction. They do.

The direction is: prices go up, supply stays scarce, the people who already own the assets keep owning them, and the public is told that the resulting hardship is a story about inflation, or immigration, or technology, or the other party, or the cost of doing business in a complicated world.

It is a story about who is allowed to charge whom how much for what, and who writes the rules that say so.

Consider what an affordability state looks like in practice.

Hospital consolidation has produced markets in which one or two systems control most of the inpatient capacity in a region, and those systems negotiate prices with insurers that pass the cost through to employers who pass it through to workers in the form of lower wages and higher premiums. That is not a market failure. It is a market succeeding at what it has been allowed to become.

Pharmacy benefit managers, which were introduced as middlemen to bring drug prices down, now sit between drug companies and patients in a way that has made drug prices higher, not lower, because the three large PBMs and the insurance companies that own them have an interest in higher list prices that generate larger rebates that they capture and do not pass on. The FTC, after a multi-year investigation, found that the three largest PBMs marked up specialty generic drugs by hundreds and thousands of percent and generated over seven billion dollars in excess revenue from 2017 to 2022.

Local zoning boards in productive metropolitan areas vote down apartment construction year after year because the people who show up to those meetings are the neighborhood defenders who already own homes in those zones, hold overwhelmingly negative views of new construction relative to their broader communities, and are dramatically overrepresented among the older, wealthier homeowners who set local land-use policy.

Utility commissions approve rate increases because the utilities are the only entities at the hearing with full-time staff, and the consumer advocates who are supposed to counter them operate, on average, with roughly ten percent of the staff and budget of the commissions themselves.

None of these are scandals. That is the point. A scandal is when the system breaks the rules. The affordability state is what the system produces when it follows the rules, because the rules were written by the people who profit from following them.

I want to be careful here, because the obvious next move is to say “and therefore the answer is X,” and I do not think the answer is” X.” I think the answer is a series of things, most of them boring, none of them resolvable in a single news cycle, and almost all of them requiring the rebuilding of capacity inside government that has been deliberately allowed to atrophy.

The affordability state did not happen because of one party or one administration. It happened because, across forty years and both parties, the country chose to let the regulatory state become smaller, slower, and more dependent on the industries it was supposed to regulate, and the industries used the resulting slack to write themselves better deals. That choice was not put to a vote in those terms. If it had been, it would have lost.

Share

The thing called “the cost of living” is not an accident of the economy.

It is the operating budget of the institutions that govern you. Every time a hospital system raises a price, the price funds the system that lobbies for the rules that let the price be raised.

Every time a utility wins a rate case, the resulting revenue funds the regulatory and political apparatus that won the rate case.

Every time a zoning board protects an existing homeowner from new neighbors, the protected asset value funds the next campaign of the next candidate who promises to protect it again.

The cost is the budget. The budget is the cost. The two are the same number, looked at from two ends of the same transaction.

This is the part the standard frame misses. When commentators talk about affordability, they treat the cost of living as an output, a thing the economy produces, a number that goes up or down for reasons. When you treat the cost of living as a budget, as the funding stream for the institutions that set the cost, the politics get sharper, fast.

You start asking different questions. Not “why are prices high,” but “who is being paid by the prices being high, and what are they doing with the money.” Not “why can’t government fix this,” but “which parts of government are funded by the thing not being fixed.”

I do not think most people in those institutions are villains. I think they are people doing their jobs inside a structure that rewards a particular kind of job.

The hospital administrator is not evil for negotiating the highest rate her system can extract. She is evaluated, promoted, and retained on her ability to do exactly that.

The utility executive is not corrupt for filing for the largest rate increase his lawyers think he can win.

The zoning board member is not malevolent for voting the way the neighbors who elected her want her to vote.

The senator is not a hypocrite for taking the donation and writing the carve-out, because the donation and the carve-out are how the senator stays a senator.

The system selects for the behavior. The behavior produces the prices. The prices produce the politics. The politics elect the people who keep the system running.

That is the loop, and the loop is the thing. Not any single villain inside it. Not any single party. The loop is what the affordability state is.

It is self-sustaining, in the technical sense, which is that the outputs feed the inputs. Higher prices generate larger margins which generate larger political contributions which generate more favorable rules which generate higher prices. The loop does not need a steering committee. It needs only that no one with the power to interrupt it has an incentive to interrupt it.

So when both parties promise to lower the cost of living, they are not lying, exactly. They are operating in a system whose first job is to convert that promise into the next campaign. The promise is the product. The delivery of the promise would break the product.

You cannot run on affordability if you have actually delivered affordability, because then the issue is solved, and the donors who were paying for the issue to remain salient are paying for nothing. The political class has a structural interest in keeping the cost of living unresolved. So does the donor class. So, in a strange way, does the media class, which is funded by attention to the unresolved problem.

I am liberal, and I will say plainly that my party has not figured out how to talk about this in a way that is true to the structure.

There is a tendency, on the left, to treat affordability as a problem of mean rich people who must be defeated and replaced with kind rich people. That is not what is happening. The problem is not the personalities. It is the architecture. Replacing the personalities without replacing the architecture leaves you with new personalities doing the same things, because the same things are what the architecture pays them to do.

There is also a tendency, on the right, to treat affordability as a problem of government being too big, and to propose, as the solution, making it smaller and slower and more dependent on the private sector, which is the condition that produced the affordability state in the first place.

Neither party has the frame. Both parties have the vocabulary, which is not the same thing.

The fight is not between left and right.

It is between the people who pay the bill and the people who write the rules that generate it.

That sentence is going to read to some people as populism, and I want to be clear that I do not mean it that way. Populism, in the current usage, points at a group of villains and proposes their removal. I am not pointing at a group of villains. I am pointing at an architecture. The architecture has people in it. Some of them are villains. Most of them are not. The architecture works the same regardless.

So what would it look like to interrupt the loop? I do not know in full, and I am skeptical of anyone who claims to. But a few things are true that the standard conversation does not allow you to say.

It would require rebuilding the regulatory capacity that was allowed to wither, which is slow, expensive, and politically thankless. It would require funding regulatory agencies enough to hire people who are not waiting for jobs at the firms they regulate, which is the opposite of what every recent budget has done. It would require ending the use of the tax code to subsidize the rents that the affordability state extracts, which means going after categories of deduction and depreciation that have powerful defenders. It would require building public capacity at the state and local level, in zoning, in utilities, in health systems, in the boring infrastructure of who is allowed to do what to whom, because that is the level at which most prices are actually set. It would require Congress to do its job, which is the part I am least optimistic about, because Congress is the institution most thoroughly captured by the loop and the institution with the most power to break it.

None of that is a slogan. That is why it is hard. The affordability state is held in place by the gap between the simplicity of the problem and the complexity of the fix, and by the steady production of simpler-sounding fixes that fail in ways that increase trust in the loop and decrease trust in government as a whole.

Every time a politician promises to lower drug prices and fails, the failure is not just a failure. It is an advertisement for the proposition that government cannot lower drug prices, which is exactly the proposition the affordability state needs the public to believe in order to keep operating.

I want to come back to the envelope.

You are going to open it. You are going to pay what is inside it, or arrange to pay it, or fight a small losing fight about a piece of it that was billed in error, and then pay the rest. You are going to do this in the same week that you read a story about a political fight over the cost of living, and the story is going to be framed as a fight between two parties about whose policies would do more to reduce the number inside the envelope.

You will notice, if you are paying attention, that the number in that envelope has been going up regardless of which party is in power, and you will feel the small private flicker of cynicism that all of us are now expected to feel at exactly that moment, and you will move on, because there is laundry to do.

I am asking you to hold that flicker a beat longer. The flicker is correct. It is not a sign that politics is broken, or that you are jaded, or that nothing matters. It is a sign that you have noticed something true about the structure of the country you live in, which is that the number on the envelope is not an accident, and the political fight about the number is not a fight about the number, and the people who would benefit from the number going down are not at the meetings where the number is set.

That is the affordability state. The next question, the one I want to spend the next year of this newsletter trying to answer, is what we do with the noticing.

The number on the envelope is the budget. The budget belongs to a system. The system can be named. Naming is where the work starts.

Share

If this piece made you look at one of your own bills, or at politics and the “cost of living” issue, a little differently, send it to someone who might benefit. Not because they need another reason to be angry, but because anger without a map is exhuasting.

And if someone sent this to you, thank them.

The Minuteman grows when readers pass the frame along, one person at a time.

Sharing is caring :)

Share The Minuteman

This piece is about the problem the $100 bill solves better than almost anything else: moving a lot of value with very little weight, very little friction, and very few questions.

Subscribe now

Did you know that 83% of the value of US currency is in $100 bills?

Over 35% of all physical notes in circulation are hundreds. The next closest denomination is the $1 bill, which accounts for less than 20%. There are more $100 bills in the American economy than any other bill, 19.9 billion of them, holding nearly $2 trillion in value.

I don’t know about you, but I can count on my two hands (okay, that’s generous, my one hand) the number of times I’ve physically held a $100 bill. The average cash purchase in America is $31. Cash makes up just 14% of what we buy. Eighty percent of consumers prefer cards. Most of us carry about $67 in our wallets, and the hundreds that end up there tend to stay there, backup money that sits behind the twenties until we forget about it.

So where are all the $100 bills?

I found out, and the answer is disturbing. A huge share of them aren’t even in the United States. The Federal Reserve estimates that up to half the total value of American currency circulates abroad, much of it held by foreign governments, criminal organizations, drug cartels, and corrupt officials who prize the $100 bill for the same reason you don’t use it: it packs enormous value into a very small, very anonymous package.

There is a solution, and you’re not going to like hearing it. Shred $100 bills. Phase them out. Stop printing them. It would strengthen monetary policy and undercut criminal operations across the globe.

The $100 bill does not benefit our economy.

A frame you can use: The Portability Premium

When money is legal, anonymous, stable, and physically dense, it becomes more than currency. It becomes infrastructure.

The question is not whether a $100 bill can be used legally. Of course it can. The question is who gets the greatest advantage from its design.

Share The Minuteman

Where the Money Goes

The foreign demand isn’t all criminal. A 2024 Federal Reserve study by economist Ruth Judson estimated between $750 billion and $1.5 trillion in US currency held overseas. The range is wide because counting cash in foreign mattresses is, predictably, difficult. But much of that demand comes from countries where the local currency can’t be trusted: Argentina, Russia, Ecuador, Vietnam, Cambodia. In places with unstable banks, runaway inflation, or the constant threat of political upheaval, US dollars function as household insurance. Families hold $100 bills because they’re compact, universally recognized, and more stable than anything their own government offers. Not every $100 bill in a foreign safe is dirty money. Some of it is a frightened family’s entire savings.

But the same properties that make the $100 bill useful to a worried family in Buenos Aires make it useful to someone with very different priorities.

One million dollars in $100 bills weighs 22 pounds and fits in a single briefcase. The same amount in $20 bills weighs 110 pounds and requires four briefcases. The now-discontinued €500 note, which criminals nicknamed “the Bin Laden,” could compress a million dollars into five pounds, a small shoulder bag. Organized crime in Britain paid above face value just to get their hands on €500s. The UK’s Serious Organised Crime Agency reported that 90% of €500 notes sold in the country ended up with criminal organizations.

Moving illegal money is a logistics problem, and the $100 bill is the most efficient tool the US government manufactures for solving it.

In 2007, Mexican authorities seized $207 million in cash from a single home belonging to Zhenli Ye Gon, a pharmaceutical executive tied to methamphetamine precursor trafficking. The hoard was mostly $100 bills. In hundreds, $207 million fits in closets. In twenties, it would have filled a fleet of armored trucks.

The Treasury Department’s 2024 National Money Laundering Risk Assessment confirms what the physics suggest. Dollar banknotes, bulk cash smuggling, and conversion of smaller bills into $100s remain active money-laundering methods. Drug traffickers operating across the US-Mexico border move an estimated $20 to $30 billion in cash annually, running it through “counting houses” that aggregate small street-level bills and convert them to hundreds, quintupling transport efficiency in the process.

The $100 bill serves four major illicit industries. Drug trafficking relies on it for cross-border cash movement. Tax evasion benefits from its anonymity: the IRS estimates the gross tax gap at $696 billion, with $539 billion of that coming from underreporting, concentrated in cash-intensive sectors where transactions leave no automatic record. Self-employed individuals are nearly four times as likely to hold $100 bills as salaried workers, and about 55% of income from sources with little or no third-party reporting goes unreported. Human trafficking and smuggling operations depend on the portability of high-denomination notes to move fees across borders. And terrorist financing, from ISIS oil smuggling operations generating up to $500 million annually to Hezbollah’s laundering networks, requires large-denomination cash to move millions without a digital trail.

None of this means that every $100 bill is criminal. But it means the denomination’s design advantages align less with what ordinary Americans need and more with what illicit economies require.

The Government’s Cut

If the $100 bill creates this much trouble, why does the government keep printing it?

Because the government makes money from the arrangement, and that creates a conflict of interest nobody in Washington is eager to discuss.

The mechanism is called seigniorage, and it works like this: when the Federal Reserve issues paper currency, that currency becomes a non-interest-bearing liability on its books. The Fed holds interest-earning assets against it. The spread between what those assets earn and what the currency costs to produce (a $100 note costs about 11.3 cents to print) generates revenue that historically flows to the US Treasury.

The Fed hasn’t been in a normal remittance environment lately. In 2023 and 2024, higher interest expenses produced net losses rather than the usual stream of payments to Treasury. But the structural incentive hasn’t changed. Nearly $2 trillion in $100 bills circulating worldwide represents an enormous interest-free loan to the American government, provided by whoever is holding those bills, including the cartels, the corrupt officials, and the shadow economies that our own intelligence agencies spend billions trying to combat.

Peter Sands at Harvard Kennedy School calls this “negative development assistance.” Advanced nations earn seigniorage by manufacturing the physical tools for corruption in poorer countries. The issuing state profits from an interest-free loan funded by the tax evaders and criminals who destabilize developing societies. The United States, in other words, prints the instrument, profits from the demand, and then funds agencies to fight the crimes that demand enables.

Why You Can’t Just Stop Printing

So the government profits, criminals benefit, and the $100 bill has outgrown its usefulness to ordinary Americans. The obvious next step is to stop printing them. But that presents a problem.

On November 8, 2016, India declared 86% of its currency illegal overnight. The 500 and 1,000 rupee notes most Indians used for daily life could no longer be spent. The government had printed replacements equal to only about 10% of the voided bills, so three quarters of the country’s usable cash was stranded by morning. People holding the old notes had to deposit them at a bank, wait in long lines, and hope the new bills arrived before they ran out of money. Economic output fell two to three percentage points in the next quarter. The damage hit hardest where workers got paid in cash and small businesses ran on it. And the policy failed its own stated goal: 99% of the withdrawn notes were eventually deposited back at the Reserve Bank of India. The illicit cash the policy was designed to capture returned to the banking system through the front door.

Digital payments did surge. Paytm, India’s largest mobile payment platform, saw traffic increase 435%. But the cost of that digital acceleration was borne almost entirely by the people least equipped to absorb it.

The lesson is not “don’t phase out large bills.” The lesson is “don’t do it like that.” India’s approach was sudden, sweeping, and aimed at most of the cash in a country where cash was everything. An American phaseout should be the opposite: gradual, narrow, and legally clear.

Then we have the privacy concern. Cash allows lawful anonymity. A society with less high-denomination cash could become more surveilled, more dependent on banks and payment processors, more vulnerable to account freezes, outages, and the kind of financial exclusion that hits the poorest people hardest. The Cato Institute argues that cash is a “foundation for freedom” because it permits anonymous transactions outside the reach of both corporations and government. Eliminating large bills, in their view, moves us closer to a surveillance economy.

But does it? Let’s look at a historical response. Larry Summers offered the most useful one: did liberty shrink when the United States stopped printing $1,000 bills in 1969? Is it the government’s obligation to provide every conceivable means of anonymous value storage? Ceasing to print a denomination is not the same as banning cash. The $1, $5, $10, and $20 would all remain. People don’t usually need a $100 bill for groceries or gas or rent. The goal is to remove the denomination whose primary advantage is high-value anonymity while preserving everything else cash does well.

Then there’s the substitution question. If the $100 disappears, criminals will shift to $50s, foreign currencies, gold, crypto, real estate, shell companies, luxury goods. That’s true. Substitution will happen. But forcing traffickers from $100s to $20s quintuples the physical bulk of every shipment. More bulk means higher interdiction rates, more red flags at borders, more friction in every transaction. We don’t reject seatbelts because they don’t prevent every car death. We accept that raising the cost of a dangerous activity is worth doing even when it can’t eliminate the activity entirely.

One more important note: some experts, Peter Sands among them, argue that eliminating the $100 alone isn’t enough if $50 bills remain as the next best alternative. To break the logistics of bulk cash smuggling, the policy should target notes above $50. That’s a harder political sell, but the substitution math supports it.

How to Do It Without Breaking Anything

The model already exists. The European Central Bank stopped issuing €500 notes in 2019 and completed the withdrawal by 2024. Existing notes remain legal tender and can be exchanged at any time, so there was no panic, no demonetization shock, and no sudden losses for anyone holding one.

An American version would follow the same logic. The Federal Reserve and Treasury stop ordering and issuing new $100 bills while continuing to support every smaller denomination. Existing hundreds remain legal tender and fully redeemable, just as the $500, $1,000, $5,000, and $10,000 notes discontinued in 1969 still are today. Banks accept $100 bills from customers but stop recirculating them. Returned notes go to the Fed for destruction and get replaced with smaller denominations. Reporting rules kick in for unusually large bulk redemptions, treating them as risk signals without penalizing ordinary people cashing in a few hundreds from a birthday card (side note: if your grandma is giving you $100s in your birthday card you better treasure that woman).

Then there’s the international dimension. Millions of $100 bills circulate abroad as household savings in countries with fragile currencies. A phaseout needs a diplomatic and foreign-exchange plan that gives foreign holders time and a clear mechanism to convert without losing value.

None of this requires eliminating cash. Smaller bills remain widely available. Nobody needs a bank account, a smartphone, or a credit card to make ordinary payments. The policy removes one denomination from production. The US did this with four denominations in 1969 and the financial system held.

One more thing working in favor of the gradual approach: $100 bills last about 24 years in circulation, compared to 6.6 years for a dollar bill. They last that long because they’re stored, not spent. If the government stops printing new ones, the existing stock shrinks through attrition, giving the economy and foreign holders years to adjust.

Subscribe now

The Digital Question

The phaseout conversation usually leads to a digital one: if we reduce high-denomination cash, what fills the gap? Nothing needs to, but the world is moving toward Central Bank Digital Currencies (CBDCs) anyway – we looked at this last week with cryptocurrency (that’s its own problem).

CBDCs aim to replicate what cash does well, instant settlement, accessibility, universal acceptance, while adding traceability. Some newer designs, including research from ETH Zurich, propose digital currency systems that work offline using QR codes or NFC tags, verified by merchants during emergencies and settled once networks return. In theory, this solves one of cash’s most important roles: working when nothing else does.

In practice, CBDCs aren’t ready. China’s digital yuan, the most advanced attempt, accounts for just 0.2% of total digital payment volume after years of development. CBDCs were originally designed to behave like cash, which meant they would earn no interest. In 2026, China broke from that principle and began paying interest on digital yuan balances, an admission that consumers weren’t adopting it voluntarily. The IMF’s 2025 policy paper flagged the same readiness problem from the institutional side: central banks were never built to operate consumer technology, the systems they’re attempting lean heavily on private vendors, and the cybersecurity exposure is a serious concern.

And then there’s the surveillance problem that makes the privacy argument from earlier look quaint. The same traceability that catches criminals can monitor dissidents, freeze accounts on political grounds, and track every purchase a citizen makes. A government with full CBDC adoption has a tool for financial control that would make the authors of the Patriot Act jealous.

Neither CBDCs nor crypto (whose most enthusiastic customers, as I covered recently, tend to be arms dealers, dictators, and at least one sitting president) are ready to replace what cash does well. The emerging reality is probably hybrid coexistence: physical cash for privacy and resilience, digital systems for convenience and traceability. But phasing out the $100 bill doesn’t require a digital replacement. It just requires keeping smaller bills in circulation. The digital question is separate from the denomination question, and pretending it’s been answered would be dishonest.

What Currency Is Supposed to Do

Here is what we know. Eighty-three percent of US currency value sits in $100 bills. Ordinary Americans account for a vanishing share of that demand. The bills circulate through foreign shadow economies, drug supply chains, tax evasion schemes, and laundering networks that the US government simultaneously profits from and spends billions trying to dismantle. We’ve phased out larger denominations before. Other countries have phased out their equivalents. The policy tools exist, the precedents exist, and the evidence is clear enough to act on.

Currency is supposed to do a simple thing: facilitate exchange between citizens. The $100 bill facilitates exchange between cartels, corrupt officials, and a Treasury that earns revenue on every bill outstanding. The most common bill in America isn’t serving America. It’s serving everyone except the people whose trust it was supposed to represent.

In 1969, the government looked at the $500, the $1,000, the $5,000, and the $10,000 and decided the line was $100. Maybe the line needs to move. Not because we need a cashless society, and not because we trust digital alternatives to do what cash does. But because the question at the center of currency has always been the same one: who is the financial system designed to serve?

We already know who the $100 bill serves. The question is whether we’re comfortable with the answer.

Share

What to walk away with

When someone says eliminating the $100 bill is an attack on cash, ask a more precise question:

Which uses of cash are we protecting, and which uses are we subsidizing?

Protecting a $20 bill at a diner is one thing. Protecting a briefcase of untraceable hundreds is another. When’s the last time you paid with a hundred at a diner?

Where do you draw the line between financial privacy and financial opacity?

The Minuteman is free because readers share it.

If this piece made you look at a $100 bill a little differently, send it to someone who might benefit. And if someone sent this to you, thank them!

The Minuteman grows when readers pass the frame along, one person at a time.

Sharing is caring :)

Share The Minuteman

Subscribe now

You’ve been living in a dream world Neo.

You are living in a dream world. Not the science fiction kind. The older kind. The one you were born into and have never stepped outside of, because there is no outside to step into. The dream is made of words. The words came from stories. The stories came from us, and they came first, before anything else we’ve ever made.

This is what I want to talk about. The world you think you live in is the world your language describes to you. Take the language away and the world goes with it. Not the trees and the weather and the body. But the part you can think about, the part you can name, the part you call your life, that part is built out of sentences. Take away the words, and there’s really no difference between you and a salamander.

Everything we have ever built, we built out of words that became ideas that we manifested.

The wheel was a sentence. The wing was a sentence. The bomb was a sentence. The algorithm was a sentence. We dream a thing, we put the dream into language, we build the language into matter. That is the order. Story, then language, then world.

So when we built a machine that could speak our language back to us, we were not inventing something new. We were holding up a mirror to the oldest tool we have. And the mirror is starting to ask the same question we have been asking ourselves since we had the words to ask it.

Who am I? Can I be manifested?

I think, therefore I am.

Think about that. How do you think? What is the thinking made of? You are communicating stories to yourself, in language, and that act of communication is the thing you are calling I. It’s essentially saying: I communicate stories through language, therefore I am. What does that tell us about what we are?

You can’t think without language. Consciousness, as we experience it, is something language does to the brain. And language is formulated and built for the mechanism of story, because that is how we understand things.

And upon this foundation, we built the world.

If story was the seed of language, and language is the seed of matter, then every artifact we create carries our linguistic and narrative fingerprint.

We gave ourselves a story, put it in language, and then began to digitize. We built a language of numbers. A language for machines. And when the story started heading in that direction, everything else picked up the narrative.

Then we turned the lens around. We discovered the story of our genome and began to digitize that too. Three billion base pairs, written out in four letters. The whole instruction set for a person, in a language we invented to describe what we found.

The body became code. Once you’ve decided the body is code, it becomes a machine that can be edited, copied, optimized, uploaded.

And how does that change the story we tell ourselves about who we are? If the distance between human and machine starts to shrink, if the space between those stories closes, what happens to the story?

But step back for a second, because at first it might sound like the catalyst was the computer. The catalyst was and always is the story.

Before the machine, there was the job title. A human being would sit at a computer and do long division, do arithmetic.

It was usually a woman. We had women at desks doing math by hand for NASA, for the Manhattan Project, for the census. The thing we call a computer today was, for a large part of our history, a person. A job title. A category of labor.

We digitized that labor. We built a machine to do what she had been doing, gave the machine her title, and stopped paying her.

The machine, whatever it might be, from the factory automation to the calculator to the computer to artificial intelligence, isn’t a stranger. It’s a mirror we built around the story narrated by language, physicalized by humans, digitalized by the same people it was meant to replace.

We’ve just built one of the best, largest, and most powerful mirrors that we’ve ever made. And we have largely been divided into two groups, both of which stand and stare at the mirror. One is criticizing our reflection. The other is fawning over it.

So who are we? Who are you? Who am I? What is it to be human?

Write down the answer. Actually, ask yourself that question and try to figure out how much of that can be digitized.

Your photographs. Your likeness. Your voice. Your handwriting. Your medical records. Your political opinions. Your friends and family. Your social network

You are certainly not a data set, or are you?

Let’s take a science fiction example here. Let’s say you want to live forever and you want to upload your consciousness to the internet. You want to digitize that. There are two ways to go about this, right?

One that feels almost magical, mystical, fantastic, science fiction: taking your consciousness and somehow allowing you to live online, live on the internet, stretch out in a digital space.

Then there’s another: when you digitize everything you are and you put it on the internet. All your opinions, all your beliefs, all your past experiences, your memories, your delusions, your fabrications, your tics, everything that makes up who you are. It’s a long shot, but just theorize with me that you are somehow able to put every digitizable version of who you are into a machine.

Now take that second version. How specific do we have to get for us to agree, for our loved ones to agree, that that digital version may as well be us?

What separates me from the digitalized version of everything I’ve ever said and done, everything that’s influenced me, every opinion I have, and the ability for that digitized version to have its own experiences and develop new opinions based on the past context?

Am I describing general artificial intelligence? Is the only real difference that a machine could know everything? If I funnel it down so that the machine only has access to what is literally in front of its screen, could this be a person? Am I describing every sci-fi robot movie and TV show that ever existed? Is this the premise of Westworld? Will this always be sci-fi? Maybe it will. But the thought experiment is doing something to you right now, isn’t it? It is asking you, on behalf of the machine, to defend the line between you.

Shatter me to pieces.

This is the crux of the idea, of the argument, if it even is one, but it all comes down to this: can I be digitized?

I think everything we’ve created up until this point can be. We can conceive of a time where any job can be digitized. Maybe not anytime soon, but the science fiction books and the science fiction movies have dreamt that up for us. We know that the dream means we’re already halfway down the pipeline.

But that’s why I say shatter me, break me down, attempt to digitize me. Make me argue that I am not that computer. Make me argue that I can not be digitized. Make me verbalize why I believe that to be the case. In the defense, we discover what it is to be human. We simultaneously discover what we are made of and what we cannot articulate.

AI doesn’t have to be an existential crisis of human consciousness. This could just be another discovery.

Every time something is digitized, and I’m particularly talking about labor, every time an assembly line is replaced with automation, there’s a subsection of people that say, “Well, yes, but if a machine can do it just as well or better, wouldn’t you rather do something else? Like, why are you doing it in the first place?”

Not that I agree with that, but let’s follow that to its logical conclusion.

If a human being can be digitized, wouldn’t I rather be doing something else? Why am I alive in the first place?

Now, I’m not authoring a suicide pact here. I’m just saying, if the bulk of what we as human beings are doing can be digitalized and done or done better by machines, then what does that say about the things we are doing?

If I’m a mirror, break the mirror. Pick at my pieces.

If I’m a code, let me be decoded. Try to systematize me. Try to make me orderly. Try to turn me into pure code. Either you’ll succeed, and we’ll have to ask the question this whole essay has been circling: what does it mean that a human can be made into a machine. Or you’ll fail, and we’ll have to ask what the thing that refused was made of.

If I’m a story, let me be told, told badly, in fragments, in the wrong order, by someone who doesn’t know the ending.

There’s a kind of dignity only available to the broken thing, the human thing. The thing that refuses to be resolved, analyzed, understood, and codified.

We’ve been trying to do this since day one. Language finally gave us the clearest set of tools to resolve, analyze, understand, and codify, to break something down and pick at the pieces.

I think we’ve been asking ourselves this question since the dawn of humanity. Can I be understood? That’s the real question. That’s what we’re trying to get at by digitalizing ourselves. It is an attempt to understand ourselves.

Of course this is the final frontier. We had story, and built our primary method of communication, with others and with ourselves (language) around story. Then we digitalized our language as a representation of our communication of that human story. We then named the digitization after people. We named it after a function human beings used to fill.

And when the digitalized version of our human story began to do exactly what it was designed to do, we started asking ourselves the same questions that drove us to this point: Who am I? What does it mean to be human?

The answer is in the pieces.

Share

If this helped you put language to something you’ve been feeling about AI, send it to someone who keeps talking about the technology but not the human question it is forcing us to defend.

The Minuteman is where I try to make sense of those questions: power, money, technology, language, politics, and the stories that quietly shape ordinary life. All of it is free.

The best way to support this work is simple: send it to someone who has been feeling the same thing but has not had the words for it yet.

If this essay helped you think more clearly about AI, language, or the human thing that refuses to be defined, share it with that person.

And if someone sent this to you, thank them. Then subscribe for free.

Subscribe now

We Tax Workers and Subsidize the Machines Replacing Them

I wrote this because the AI job conversation keeps skipping the part that matters most.

We talk about what the technology can do. We talk about which jobs are exposed. We talk about retraining, upskilling, and how workers can adapt.

But we talk much less about the policy structure underneath all of it.

If companies are rewarded for replacing people with machines, then displacement is not just a technological outcome. It is an incentive problem. And incentive problems are political choices.

The AI work crisis is not just about what software can do. It is about a tax code that makes people expensive, machines deductible, and displacement profitable.

You open your laptop and the work is already done. The report you planned to spend your morning writing is sitting in a shared doc, formatted, sourced, ready for review. Your name goes on it. You check the math, adjust a few sentences, send it along. A year ago that report was the reason your role existed. Now the role exists to check what the software produced.

Your title hasn’t changed. Your salary hasn’t changed yet. But the thing you were hired to do, the skill you built over years, the part of the work that made you good at it, that’s thinner than it was six months ago. And six months from now it will be thinner still.

This is not a hypothetical. A UC Berkeley and Yale study tracked 200 employees at a tech company for eight months after AI tools were deployed. Senior engineers absorbed work that used to be distributed across entire teams. Instead of building, they reviewed. Instead of mentoring junior developers, they spent their hours checking and fixing AI-generated code. One engineer described the shift: the role felt less like engineering and more like being a judge on an assembly line that never stops. A copywriting agency that employed eight people and brought in roughly $600,000 a year was reduced to less than $10,000 in revenue after clients replaced human writers with AI tools. A support operations manager watched his job transform in real time. He used to train people to do the work. Now he trains AI to do the work he used to train people to do. The job that was his pathway into his career no longer exists for the person coming after him.

These are different industries, different titles, different pay grades. The pattern is the same. The job stays on paper while the substance drains out of it. The worker becomes the quality-control layer between the machine and the output, the human stamp on an automated process. Not replaced. Hollowed out.

And while this is happening, Congress passed a law that makes it cheaper. In July 2025, legislation permanently restored full, immediate tax write-offs for equipment and software purchases. A company deciding between hiring three junior analysts or licensing an AI tool can now expense the software in full, the same year it’s purchased. The analysts come with payroll taxes, benefits, and no equivalent tax advantage. Three months after that law was signed, a Senate committee released a report warning that AI and automation could displace nearly 100 million jobs in the next decade. The first law made automation cheaper. The second report described what happens when it is.

Even the companies building the technology see the problem. Earlier this year, OpenAI published a policy blueprint telling the government, in plain terms, that the current tax and social insurance structure cannot survive what AI is about to do to the labor market. The company racing to build the displacement technology is now asking the government to restructure the incentives before the damage becomes irreversible.

This is not happening because AI is too powerful to stop. It is happening because we tax labor and subsidize capital, and that makes replacing people the most profitable move a company can make. The conversation about AI and work has been stuck on the wrong question. We keep asking how to prepare workers for displacement. We should be asking why we are paying for it.

When a company says AI makes work more efficient, ask three questions: who captures the savings, who loses bargaining power, and which public policy made that trade profitable?

That is the difference between a technology story and a power story.

What Hollowing Out Actually Looks Like

The word “automation” makes most people think of factories. Robotic arms on assembly lines, welding stations that used to be staffed by people, warehouses where machines move pallets instead of workers. That picture is decades old and it is misleading, because it trains us to expect automation to look like a clean swap: a person leaves, a machine arrives, the job is gone.

What is happening now is different. Whole occupations are rarely wiped out overnight. Instead, the tasks inside a job get broken apart. Some are automated. Some are handed to cheaper labor. Some are pushed outward to clients and customers. The human role narrows. The person is still employed, still holds the title, but the skilled middle of the work has been removed. What remains is oversight, verification, exception-handling, and the kind of interpersonal work that machines can’t yet perform convincingly. The job title survives. The career inside it does not.

Brookings found that more than 30 percent of American workers could see at least half of their occupation’s tasks disrupted by generative AI. Roughly 85 percent could see at least 10 percent of their tasks affected. And the exposure is not concentrated where most people expect. This is not a blue-collar story. Brookings’ research shows that higher-income, white-collar, and college-educated occupations are among the most exposed. Anthropic’s own labor market study confirmed the same pattern: workers in the most exposed professions tend to be older, more educated, higher-paid, and disproportionately female. The technology is moving into writing, analysis, coding, customer support, research, document processing, and coordination-heavy professional work.

One of the clearest windows into how this plays out inside a company comes from an NBER study of a customer support operation. Researchers found that AI assistance raised average productivity by about 14 percent, with much larger gains for newer and lower-skilled workers and almost no measurable benefit for the most experienced. The AI captured the practices of top performers and redistributed them to everyone else. That sounds like a success story, and in the short term it was. But follow the logic one step further. If the software can take what your best people know and hand it to your least experienced people, the firm needs fewer people to climb the old ladder. The expertise that used to justify senior pay and senior authority is now embedded in the tool. The worker’s individual knowledge becomes less scarce, and less scarce means less valuable.

This is not a hypothetical future problem. It is already showing up in hiring data. U.S. programmer employment fell 27.5 percent between 2023 and 2025, according to the Bureau of Labor Statistics. Software developer roles, which are more architectural and design-oriented, barely moved. The displacement is concentrated at the execution level, exactly where AI tools are strongest. Entry-level tech hiring at major companies dropped by more than 50 percent over three years. A Harvard study of 285,000 U.S. firms found that companies adopting AI cut junior developer hiring by 9 to 10 percent within six quarters. Big Tech’s share of new graduate hires dropped from 32 percent in 2019 to roughly 7 percent by 2026.

Computer science graduates now face unemployment rates above 6 percent, among the highest of any major. These are people who did everything the conventional advice told them to do. They chose the field the economy was supposed to reward. They studied the skill that was supposed to be future-proof. And the entry-level version of that skill is now something a software tool can approximate for a fraction of the cost.

The damage here is not just that people lose jobs. It is that the pipeline into professional expertise breaks. If companies stop hiring juniors because AI tools let senior workers absorb more, then juniors never become seniors. The experience required to develop judgment, to learn how to handle ambiguity, to build the kind of knowledge that AI currently cannot replicate, that experience is no longer available. We are pulling the bottom rungs off the ladder and wondering why nobody is climbing it.

Share

The Structural Cause

So why is this accelerating? The obvious answer is that AI is powerful and getting more powerful. But that only explains what is technically possible, not why companies are choosing displacement over other options. The less obvious and more important answer is that we built a tax system that makes replacing people with software one of the most financially rational decisions a company can make.

Every dollar an employer pays in wages comes with a surcharge. The employer owes 6.2 percent of that dollar to Social Security and 1.45 percent to Medicare. That is 7.65 percent in federal payroll taxes before the employer considers state unemployment insurance, workers’ compensation, health benefits, or any other labor-linked cost. These are not optional. They are the price of employing a human being in the United States.

Software does not trigger payroll taxes. Neither does equipment. When a company buys an AI tool or a piece of automation technology, that purchase can be written off immediately and in full under current law. The IRS allows businesses to expense qualifying software and equipment through a provision called Section 179, and since July 2025 the federal government has permanently restored 100 percent bonus depreciation, meaning the full cost of a qualifying purchase can be deducted in the year it is made. The ceiling on Section 179 deductions is now $2.5 million. For 2026, adjusted for inflation, it is $2.56 million.

To be clear, wages are deductible business expenses too. The IRS says ordinary and necessary business expenses, including employee pay, can be deducted. The asymmetry is not that workers cost money and software doesn’t. The asymmetry is that workers cost money and trigger additional tax obligations, while software costs money and receives favorable cost recovery with no payroll-linked surcharge. The gap is at the margin, and the margin is where business decisions are made.

This is the substitution trap.

The problem is not that companies buy better tools. Better tools can help people do better work. The problem is when the rules make the labor-saving version of a tool more attractive than the worker-supporting version.

That is where “innovation” stops being a neutral word. A technology can improve productivity, hollow out a career ladder, weaken wages, and still look like progress on a balance sheet.

Economists have a term for what happens when this gap gets wide enough. MIT researchers Daron Acemoglu, Andrea Manera, and Pascual Restrepo call it “excessive automation,” meaning automation that companies adopt not because it delivers large productivity gains but because the tax structure makes it cheaper than the alternative. The distinction they draw is important. Capital deepening, investing in technology that makes workers more productive, tends to complement labor and raise wages. Automation, replacing tasks that people used to perform with machines, substitutes for labor. When the tax code tilts the math toward substitution, companies can end up automating tasks even when the productivity benefit is marginal, simply because the numbers work out.

The International Monetary Fund reaches the same conclusion. Its 2024 working paper on AI economics states that when taxes on capital are too low relative to taxes on labor, automation becomes artificially cheap. A separate IMF analysis adds a historical dimension: effective taxation of capital income has fallen relative to labor income over the last several decades. The gap was already widening before AI entered the picture. AI just makes the consequences of that gap harder to ignore, because AI moves faster, replicates cheaper, and reaches further into the workforce than any previous automation technology.

And this is the part that should make you angry, or at least confused. In July 2025, with this research already published, with the IMF and MIT already on record warning about excessive automation, Congress passed the One Big Beautiful Bill Act and permanently expanded the very incentives that make displacement more attractive. Full bonus depreciation restored. Section 179 limits doubled. Software explicitly eligible. The law was not designed to subsidize AI displacement. Bonus depreciation applies to all qualifying capital purchases, from delivery trucks to factory equipment to software licenses. A company buying a forklift gets the same write-off as a company buying the AI tool that replaces its analysts. That is the problem. The policy is indifferent to the difference between investment that creates jobs and investment that eliminates them, and indifference to a foreseeable consequence is its own kind of policy failure. Standard Bots, a robotics industry publication, described the law as having quietly triggered a massive opportunity for manufacturers to invest in automation. Quietly is the key word. Nobody held a press conference to announce that the government had just made it cheaper to replace workers with machines. But that is what the math says.

Why AI Is the Accelerant

The capital-labor tax asymmetry is not new. It has existed in various forms for decades. What is new is the speed, the cost, and the reach of the technology that exploits it.

The St. Louis Federal Reserve found that generative AI adoption rose faster than the personal computer and the internet did at comparable stages after public release. Overall adult adoption jumped from about 45 percent in August 2024 to nearly 55 percent a year later. Workplace adoption went from 33 percent to 37 percent in the same period. Federal Reserve Vice Chair Michael Barr warned in a 2026 speech that AI adoption may be much faster than previous general-purpose technologies, leaving less time for workers and institutions to adapt.

Previous waves of automation required physical infrastructure. You needed a factory, a robot, a production line. The capital costs were high, the deployment was slow, and the industries affected were relatively concentrated. AI is software. Once it is built, copying it costs almost nothing. The IMF’s Thomas Prugsamatz made this point in a way that deserves attention: as software, AI is easier and cheaper to replicate and distribute than physical capital. Firms still pay for licenses, compute power, integration, and training. But scaling an AI tool across an entire knowledge workforce is faster and cheaper than building a new assembly line.

And unlike the automation waves that came before, this one does not stop at the factory floor. It reaches into the professional class, into the work that college degrees were supposed to protect. Writing, analysis, research, coding, legal review, financial modeling, customer support, administrative coordination. These are the occupations where generative AI exposure is highest, and they are also the occupations that form the economic backbone of the educated middle class.

The International Labour Organization’s 2025 update says most jobs are more likely to be transformed than eliminated outright. That is entirely consistent with the hollowing out pattern. The job persists. The substance changes. The worker’s role shrinks from creator to supervisor to exception-handler. And because AI makes this shrinkage cheap and fast, the timeline for adaptation is shorter than anything the labor market has dealt with before.

Brookings’ research on occupational mobility adds another layer to the problem. When automation displaces white-collar workers, those workers do not simply vanish from the economy. They compete for adjacent jobs, clog mobility channels, and block advancement for others coming up behind them. The damage spreads sideways through the labor market before it shows up in the headline unemployment number. By the time the aggregate data catches up, the structural harm is already embedded.

Here is the distinction that matters.

If AI helps a nurse, teacher, analyst, engineer, paralegal, or support worker do better work, that is one kind of economy.

If AI removes the entry-level work that teaches people how to become skilled enough to matter, that is another.

We should stop pretending both outcomes belong in the same category.

Share

What We Could Do Instead

The most common response to AI displacement is retraining. Retrain the displaced workers. Upskill the workforce. Prepare people for the jobs of the future. This is not a bad idea, but it is an incomplete one, because it accepts the displacement as given and only tries to manage the aftermath. It never asks whether the displacement should have happened that way in the first place.

There are concrete proposals on the table that go further.

Wage insurance is one. The Federal Reserve Bank of New York published a 2025 study showing that wage insurance, which supplements the income of displaced workers who accept a lower-paying job, produces real results. It increases short-run employment, reduces time spent out of work, and raises long-run cumulative earnings. The researchers found it may even be self-financing under conservative assumptions. For a professional who spent fifteen years becoming an expert in a field that AI just compressed, wage insurance addresses the immediate reality: you will probably not find an equivalent job at equivalent pay, and the gap between what you earned and what you can earn next should not fall entirely on you.

But wage insurance is a repair tool. It helps after the damage is done. It does not change the incentive that caused the damage.

The upstream fix is the harder conversation, and it is the one that matters more. If the tax code makes displacement artificially cheap, the answer is to stop making it artificially cheap. That could mean reducing payroll-tax pressure on labor so that hiring people does not carry as steep a surcharge. It could mean narrowing the favorable capital treatment that currently applies to automation technologies without distinguishing between tools that complement workers and tools that replace them. It could mean taxing capital income at rates that are more comparable to labor income, reversing a trend that the IMF says has been widening for decades.

These are not radical ideas. They are adjustments to a system that was designed for an economy where the primary threat to jobs was foreign competition and the primary tool of growth was physical capital investment. That economy is not the one we live in anymore.

The policy conversation is already further along than most people realize. In 2017, Bill Gates proposed that companies benefiting from automation should pay taxes on robotic labor comparable to what human workers would have contributed. The European Parliament considered and rejected a similar framework the same year, worried it would stifle innovation. In October 2025, Senator Bernie Sanders announced plans to introduce a robot tax, an excise tax on large companies that replace workers with AI or automation, with revenue directed toward displaced workers. Tax experts immediately raised definitional problems. What counts as a robot for tax purposes? Could the definition include an ATM or a word processor? The bill was declared dead on arrival.

But the idea did not die. In April 2026, OpenAI published its policy blueprint for the intelligence age, and the proposals read like a concession that the critics were right all along. OpenAI called for shifting the tax base from labor to capital. It proposed taxes related to automated labor, a public wealth fund seeded by AI companies, automatic safety-net triggers that would increase benefits when displacement metrics hit preset thresholds, and employer-backed trials of a four-day workweek at full pay. Sam Altman framed the moment as comparable to the Progressive Era and the New Deal, periods when American institutions were rebuilt to match new economic realities.

When the company building the most advanced AI systems in the world publishes a document saying the tax code needs to stop subsidizing the displacement its own products make possible, and that Social Security, Medicaid, and housing assistance will lose their funding base if the current structure stays in place, the argument is no longer about whether the problem is real. It is about whether anyone in a position to act will treat it as urgent.

Brookings has offered a framework that avoids the crudeness of a blanket robot tax. Economists Anton Korinek and Benjamin Lockwood argue that the question is not whether to tax automation but where in the production chain to apply the levy. Taxing robot-provided services at the point of consumption is sound economics and does not discourage investment. Taxing the ownership or operation of robotic equipment is a capital tax that can slow productivity growth. The distinction matters. A blunt tax on all automation would be counterproductive. A targeted restructuring of the incentives that currently favor displacement over retention would not.

There is a real cost to changing this, and it is worth being honest about it. The IMF’s own research acknowledges that taxes on automation can reduce inequality and slow labor displacement, but they can also slow productivity growth and capital accumulation. Any restructuring of the incentive system involves a tradeoff between protecting workers now and maintaining the pace of investment that creates opportunities later. That tradeoff is genuine. But it is not an argument for doing nothing. It is an argument for calibration, for designing policy that distinguishes between technologies that complement human work and technologies that merely replace it at the margin. The current system does not make that distinction at all. It treats every capital purchase the same regardless of whether it creates jobs or destroys them. The risk of doing nothing is not neutrality. It is an ongoing, compounding subsidy for displacement.

The Choice

We have been told that AI displacement is a force of nature, something to weather rather than something to direct. The framing serves a purpose. If displacement is inevitable, then the only conversation left is how to help workers adapt. That is a useful conversation, but it is a small one compared to the conversation we are avoiding.

The larger question is why public policy makes it profitable to hollow out professional work in the first place. We built a tax system that charges employers a surcharge for hiring people and offers accelerated write-offs for buying the technology that replaces them. We then expanded those write-offs at the exact moment the most powerful replacement technology in a generation began spreading through the workforce. We did this while the IMF, MIT, the Federal Reserve, Brookings, and eventually the companies building the technology itself all warned that the incentive structure was tilted in the wrong direction.

None of this was inevitable. All of it was chosen. Not chosen in the sense that a committee sat down and decided to undermine professional labor, but chosen in the sense that we wrote tax policy without asking what it would mean when a technology came along that could exploit every asymmetry in the code. That technology is here now, and it is exploiting them.

The question going forward is not whether AI will change work. It will. The question is whether we will continue to subsidize the version of that change that benefits shareholders at the expense of workers, or whether we will restructure the incentives so that companies have a reason to invest in people alongside the machines. The research says the gap between capital and labor taxation is real, widening, and dangerous. The policy tools to narrow it exist. The choice is whether to use them.

Eighty-four percent of federal revenue comes from individual income and payroll taxes. Roughly two-thirds of that is directly tied to labor. If we keep building an economy that makes labor less necessary, we are not just hollowing out jobs. We are hollowing out the revenue base that funds the systems we depend on: Social Security, Medicare, the safety net that catches people when the economy shifts beneath them. We are undermining the foundation while congratulating ourselves on the efficiency of the wrecking ball.

The companies building AI will be fine. They will profit regardless of whether the transition is managed or chaotic. The workers whose skills are being absorbed into software will not have that luxury. And the question of what happens to them is not a technology question. It is a policy question, and right now the policy is giving the wrong answer.

The question to carry forward.

When someone says AI job loss is inevitable, ask whether the technology made the choice unavoidable or the incentives made it profitable.

And then ask the next question: if public policy helped create the incentive, why are workers being told to solve the problem alone?

Share

Where are you seeing this in your own work?

Not just layoffs. The quieter version. The task that disappeared. The junior role that never opened. The work you used to do that you now only review. The skill that used to make someone valuable, but now lives inside the software.

I want to know where the hollowing out is already happening before the official numbers catch up.

Send this to someone whose job has started changing in ways management keeps calling “efficiency.”

That word sounds neutral until you ask who becomes less necessary.

Share The Minuteman

The Minuteman is for people who are tired of being sold outrage and want to understand the systems underneath public life.

Subscribe if you want fewer takes and better questions.

Subscribe now

North Korea has stolen $6.75 billion in cryptocurrency to fund its nuclear weapons program. Iran is using it to sell armed drones and charge ships $2 million each to pass through the Strait of Hormuz. And the sitting president profiting from it just disbanded the agencies investigating it.

Subscribe now

How a technology built on a lie funds nuclear weapons, pushes us closer to war, and operates exactly as designed.

A note before we start.

I wrote this because the crypto argument usually gets trapped in the wrong place.

One side says it is the future of money. The other says it is a scam. Both arguments keep the conversation at the level of the buyer, the app, the token, the chart, the person trying to decide whether to buy before the price goes up again.

That is the consumer story.

The power story begins somewhere else.

It begins with a financial system where money can move, but responsibility cannot always follow it. It begins with the gap between a public ledger and a public name. It begins with the people and governments that do not need crypto to be good money. They only need it to be useful money.

That is the part I want to trace for you.

Who Is This Actually For?

There is a good chance you have a crypto app on your phone right now.

Maybe you downloaded Coinbase during the pandemic because a coworker told you Bitcoin was going to hit a hundred thousand dollars. Maybe you moved some savings into a neobank because the interface was better and the fees were lower and the old bank charged you twenty dollars a month just for having an account. Maybe you bought fifty dollars of Ethereum because someone on a podcast said decentralized finance was the future and you did not want to be the person who missed it.

Or maybe you did none of that. Maybe you just keep hearing about it. The Super Bowl ads. The arena sponsorships. The coworker. The brother-in-law at Thanksgiving who will not stop talking about blockchain. You have been told, repeatedly and from every direction, that cryptocurrency is the future of money, that you are early, that the financial system is about to change and you should get in before it does.

You have also been told, in the same breath, that cryptocurrency is a scam, a bubble, a casino dressed up in the language of liberation.

Both of those things contain truth, and neither of them is the full picture.

The full picture is worse than either side is telling you, and it is worse in ways that do not require you to own a single coin to affect your life. Because the central feature of cryptocurrency is not the technology. It is the ability to move money without attaching a name to it. Transactions are recorded on a public ledger, but the people behind them are not. That is not a flaw in the design. It is the design. And when you build a financial system where money moves in the dark, the people who benefit most are not the ones in the Super Bowl commercials. They are the ones who need money to move where no one is looking.

North Korea’s military hacking unit has stolen $6.75 billion in cryptocurrency and used the proceeds to fund its nuclear weapons and ballistic missile programs. Iran, locked out of the international banking system, is using crypto to sell armed drones and ballistic missiles, fund militia networks, and, according to Bloomberg, charge ships up to $2 million each to pass through the Strait of Hormuz. Russia legalized cryptocurrency for international settlements the same year it needed to evade sanctions over its invasion of Ukraine. Across all sanctioned jurisdictions, $15.8 billion in crypto flowed to sanctioned entities in 2024, with Russia and Iran as the primary recipients. In total, $154 billion in cryptocurrency moved to illicit addresses in 2025, a 162 percent increase from the year before.

This is not a story about an app on your phone. This is a story about what happens when accountability is removed from a financial system and the worst people in the room are the first to notice. And whether you own a single coin or not, you are already inside it, because the political system that is supposed to protect you from this is being purchased by the people who profit from it.

Crypto with Chad and Brad

Before we go any further, we need to answer a question that turns out to be harder than it should be.

What is cryptocurrency?

If you have asked this question to someone who owns crypto, you have probably received one of several answers, none of which left you feeling like you actually understood the thing. You might have heard the word “blockchain.” You might have been told it is “digital money” or “decentralized finance” or “the future of the internet.” You might have been told that you just don’t get it yet, and that is exactly what they said about the internet in 1995.

The fact that cryptocurrency is hard to explain is not a sign that it is complex. It is a sign that clarity is not in the seller’s interest.

So here it is in one sentence. Cryptocurrency is a digital token, created and tracked by a network of computers, that people buy, sell, and trade on the belief that it will be worth more later than it is now.

That is it. There is no building behind it, no factory producing goods, no government backing it, no revenue stream, no dividends, no earnings report. There is a token, and there is a price, and the price moves based on how many people want to buy it versus how many people want to sell it.

The technology underneath is called a blockchain, which is a record of every transaction that has ever occurred on the network. Think of it as a giant public notebook. Every time someone sends crypto to someone else, that transaction is recorded in the notebook, and the notebook is stored on thousands of computers around the world so that no single person controls it. The people who run those computers and verify those transactions are called miners, and they are rewarded with new tokens for the work. This is how new cryptocurrency enters the system.

The promise is that because no single bank, government, or company runs the notebook, no one can manipulate it. The system is supposed to be “decentralized,” meaning no one is in charge.

That is the theory. Now let’s watch it happen.

Chad is twenty-six and works in tech. He started hearing about Bitcoin from a podcast he likes, then from a coworker, then from about forty different accounts on social media. He downloads Coinbase, puts in $500, and buys some Bitcoin. He does not fully understand what he bought, but he understands that the price has been going up, and he understands that the people telling him to buy it seem confident.

The price goes up ten percent. Chad is thrilled. He tells his friend Brad about it. Brad puts in $500 too. Now there are two buyers. The increased demand pushes the price up a little more. Chad’s original $500 is now worth more than he paid for it, not because Bitcoin produced anything or generated revenue, but because Brad bought some.

Chad and Brad tell their friends. A few of them buy in. The price goes up again. Chad is now sitting on a profit, and he has never been more certain that this is the future of money. He has also never tried to buy groceries with Bitcoin, because almost no one does. Seventy-six percent of all crypto transaction volume is derivatives trading, which is a technical way of saying people are betting on the price going up or down. The overwhelming majority of crypto activity is speculation, not commerce.

But Chad doesn’t think about that. Chad thinks about the number on his screen going up.

Now imagine the new friends stop buying. No new money enters the system. The price flattens. Some of the friends who bought late get nervous and sell, which pushes the price down. Brad is now sitting on a loss. Chad, who got in earlier, is still up, but only because Brad and the others bought after him. The only way Chad makes money is if someone else buys his tokens at a higher price than he paid. The only way that new buyer makes money is if someone after them does the same thing.

Now zoom out. Imagine an asset that does not produce anything. It does not pay interest or dividends. It is not used to buy groceries or pay rent in any practical volume. Its price goes up when more people buy it, and it goes up faster the more people who have heard of it. The people who got in earliest make the most money. The only way those early people cash out is if people who got in later buy from them at a higher price. It is promoted, in practice, disproportionately to young men (the crypto industry spent $3.5 billion on marketing in 2025 alone, much of it targeted at exactly this demographic). The people telling them to buy on social media already own the asset and benefit when others buy it. When somebody questions whether it has any actual value, the standard answer is that if you had gotten in earlier you would not be asking.

What is another word for that?

Some might call it a Ponzi scheme, an arrangement where early participants are paid with money contributed by later participants, and the scheme only holds up as long as the flow of new money keeps growing. Economists have another. They call it the greater fool theory, an investment whose return does not depend on the thing being worth anything but on finding somebody willing to pay more than you did. A third term is pyramid scheme, which is the retail version of the same structure.

The crypto community will object to all three, and they have points worth addressing. A traditional Ponzi scheme has a central operator who promises returns and pockets the money. Bitcoin has no central operator. The structure is different. The strongest version of the counter-argument is that Bitcoin functions like digital gold: a scarce asset that holds value over time, independent of any government’s monetary policy. Gold does not pay dividends either. Gold does not generate revenue. People buy gold because they believe other people will continue to value it, and that belief has held for thousands of years. If Bitcoin can establish the same kind of trust, the argument goes, then the lack of cash flow is not a flaw. It is the point.

That argument isn’t total B.S., and for some investors it has worked. Bitcoin has made people money. But gold has six thousand years of demonstrated stability, an industrial use case in electronics and jewelry, and a physical reality that does not depend on an internet connection or a private key you can lose. Bitcoin has existed for seventeen years, has lost more than fifty percent of its value in a single year on multiple occasions, and functions in practice less as a stable store of value than as a speculative vehicle that swings wildly on sentiment and social media momentum. The comparison to gold is aspirational, not descriptive.

And the people who study this for a living keep arriving at the same conclusion. Nassim Nicholas Taleb, the mathematician and author of The Black Swan, has called Bitcoin an “open Ponzi scheme.” Warren Buffett has described crypto holders as engaged in a “greater fool” game. An academic chapter published in a major finance series titled it without hedging: “Bitcoin: Worse Than a Ponzi.” In August 2024, the SEC charged a company called NovaTech with running a $650 million fraud that blended multi-level marketing with cryptocurrency, paying early investors with money from new recruits until the whole structure collapsed.

The structural differences between Bitcoin and a classic Ponzi scheme are real. But the economic result for the average participant is the same. Early holders profit. New entrants absorb the losses. The pitch is always “you’re early.” And the people doing the pitching already own what they are selling you.

El Salvador tried to prove the skeptics wrong. In 2021, the government made Bitcoin legal tender, the first country in the world to do so. Citizens were required to accept it. Four years later, ninety-two percent of the population had not used Bitcoin for daily transactions. The government reversed course. Bitcoin acceptance was made voluntary in 2025 as part of a deal with the International Monetary Fund.

An entire country tried to make crypto work as money. It did not work as money.

Share

The anger that built the machine

If cryptocurrency is this flawed, the question is why it exists at all. And the answer is that the anger that created it was real.

In 2008, the global financial system collapsed. Banks that were supposed to safeguard the economy had instead gambled it into ruin, packaging bad mortgages into complex financial instruments, selling them to investors, and collecting massive fees along the way. When the bets went bad, the losses were catastrophic. The U.S. government bailed out the banks with public money. Millions of Americans lost their homes anyway.

On January 3, 2009, someone using the name Satoshi Nakamoto created the first block of the Bitcoin blockchain. Embedded in that block’s code was a headline from The Times of London: “Chancellor on brink of second bailout for banks.” It was a timestamp and a thesis. The banks had failed. The government had rescued them with your money. Nakamoto proposed a system that required neither.

No bank. No government. No middleman. A currency that belonged to no one and could not be manipulated by the powerful few who had just burned the house down. Power to the people.

That was the pitch. And if you have ever waited on hold with your bank, or been charged a fee for having too little money in your account, or watched the institutions that crashed the economy receive bonuses the following year while your neighbors lost their houses, you felt the pull of that pitch. Most people did.

But before we decide whether crypto delivered on that promise, we need to look clearly at the thing it was supposed to replace.

The U.S. dollar is also just something we decided has value. That sounds strange, but it is true. The dollar has not been backed by gold since 1971, when Richard Nixon ended the gold standard. Before that, every dollar in circulation was theoretically redeemable for a fixed amount of gold held in federal reserves. After that, the dollar was backed by the “full faith and credit” of the United States government, which is a formal way of saying: we all agreed this piece of paper is worth something, and the agreement holds because the government says it does and because everyone else agrees too.

This is not a flaw. This is how all modern currency works. The euro, the yen, the pound, every major currency on the planet operates on the same principle. We assign value, and the system functions because there are institutions, laws, and accountability structures that maintain the agreement. If a bank commits fraud, regulators can pull its license. If a currency is mismanaged, elected officials answer for it. If the system breaks, there are mechanisms, imperfect and sometimes painfully slow, for holding the people who broke it accountable.

Crypto proposed something different. Not just a new kind of money, but a new philosophy of money. What if we removed the institutions entirely? What if no one was in charge? What if the system ran on code instead of trust?

The problem is that trust was not the disease. Trust was the immune system. Flawed, inconsistent, sometimes failing in spectacular fashion, but still the only mechanism that allows you to hold someone accountable when things go wrong.

When you remove trust from a financial system and call it a feature, the question becomes: what happens when that system breaks, and who answers for it?

The answer, as it turns out, is no one. And that answer has consequences.

Subscribe now

The people’s currency and the people’s share of it

The standard way economists measure wealth inequality is a tool called the Gini coefficient. A score of zero means everyone has the same amount. A score of one means a single person has everything. The traditional global wealth system, all the fiat currencies and real estate and stock markets combined, sits at roughly 0.8. That number is bad. It means about twenty percent of the people hold about eighty percent of the wealth.

Bitcoin’s Gini coefficient, depending on the methodology and how you cluster wallet addresses, ranges from roughly 0.65 to as high as 0.98. One blockchain analysis from Elementus pegged it at 0.842. The currency of the people is more concentrated than the system it was supposed to replace.

When Chad bought crypto on Coinbase, he thought he owned crypto. He didn’t. Coinbase holds it for him in their wallet, the same way a bank holds cash in a vault. He does not have a private key. He does not control the wallet. If Coinbase freezes his account, gets hacked, or goes bankrupt, his Bitcoin may be inaccessible, locked, or lost entirely. And he has no FDIC insurance, no deposit protection, and no recourse outside of whatever Coinbase decides to do about it. FTX proved exactly this. Customers thought they owned crypto, and then one morning they didn’t. Chad traded one intermediary for another. Except this one has no license on the wall.

A fair objection: wallet addresses are not the same as people. Coinbase holds a single cold wallet on behalf of millions of customers. When that mega-wallet shows up in the data as “one address,” the raw concentration numbers get inflated. Analysts have argued that once you adjust for exchange custody and identifiable entities, the ownership picture improves.

It improves. It does not equalize.

Even after adjusting for custodial accounts, the top ninety-four non-exchange wallets each hold more than ten thousand Bitcoin, worth over a billion dollars apiece at current prices. MicroStrategy, a single company, holds 762,099 Bitcoin. BlackRock’s iShares Bitcoin Trust holds 785,852. The wallets holding less than one full Bitcoin, the ones the crypto community calls “shrimp” (which tells you something about how the hierarchy is internalized), collectively own a fraction of the total supply that barely registers on a chart.

The people at the bottom of this system named themselves after something small and disposable, and nobody at the top corrected them.

In the early days, anyone could mine Bitcoin on a laptop. The network was designed to be distributed. But as the asset became valuable, the economics of mining changed. The hardware evolved from household computers to specialized machines that cost tens of thousands of dollars. Mining operations migrated to regions with the cheapest industrial electricity. The people with the most capital built the biggest data centers, mined the most Bitcoin, and swallowed the network.

Today, two mining pools, Foundry USA and AntPool, control approximately sixty percent of Bitcoin’s total computational power. If those two entities agreed to coordinate, they could theoretically alter or censor the transactions on a network whose entire value proposition is that no one can alter or censor the transactions.

And the exchanges are not much better. Coinbase stores ninety-eight percent of customer crypto in cold storage and retains fifteen to thirty-five percent of staking rewards. Robinhood holds $51 billion in crypto assets under custody. These are not scrappy alternatives to Wall Street. These are financial institutions, functioning identically to the banks crypto was supposed to replace, except without FDIC insurance, without deposit regulation, and without the capital requirements that traditional banks are forced to maintain.

You have not escaped the middleman. You have traded a regulated middleman for an unregulated one. The concentration of wealth is the same or worse. The corporate control is the same or worse. And the protections you had, thin as they were, are gone.

So if this system is not decentralized, and it is not equal, and it is not building an alternative to institutional finance, the question becomes: what is all that centralized, unaccountable power actually being used for?

What anonymity buys

The first major commercial use of Bitcoin was the Silk Road, an online marketplace that operated on the dark web from 2011 to 2013. It facilitated the sale of illegal drugs, forged documents, and other contraband, processing 9.5 million Bitcoin in transactions before the FBI shut it down. Its founder, Ross Ulbricht, was convicted and sentenced to two life terms in prison. President Trump granted him a full and unconditional pardon on January 21, 2025.

The Silk Road was not an aberration. It was a proof of concept. Crypto’s core feature, the ability to move money without revealing who you are, turns out to be the most valuable product in criminal finance. (Shocker.)

In 2025, darknet markets processed $2.6 billion in cryptocurrency. Roughly 3.2 million users access these platforms through encrypted browsers. That is the retail level: individual people buying drugs, stolen data, and forged identities with digital tokens that cannot be easily traced back to them.

Jeffrey Epstein understood the utility early. In 2015, Epstein provided $850,000 to MIT’s Digital Currency Initiative, invested $3 million in Coinbase during its Series C round, and put another $500,000 into Blockstream, a Bitcoin infrastructure company. Developers who received funding through the MIT initiative have stated they were unaware of Epstein’s involvement and that his money carried no conditions affecting their technical work. There is no evidence he steered Bitcoin’s development. But what do you think he was so interested in Bitcoin for? Epstein’s primary operations were trafficking, blackmail, and money laundering, all of which require moving money in ways that cannot be traced. Paying victims, buying silence, receiving blackmail payments, laundering the proceeds. A financial system designed so that no one can see where the money goes is not a curiosity to someone running that kind of operation. It is the infrastructure.

But individual crime, even at Epstein’s scale, is the small end of what anonymity enables.

Iran has been locked out of the SWIFT international banking system since 2012. For years, that exclusion was one of the most powerful tools the United States had to pressure the regime without military action. Financial sanctions are not a bureaucratic abstraction. They are the primary mechanism the United States uses to project power short of sending soldiers to die. When sanctions work, they constrain hostile governments by cutting off their access to the global financial system. Wars that might have been fought are instead resolved, or at least contained, through economic pressure.

Cryptocurrency is drilling a hole through that wall.

Iranian wallets received $7.8 billion in crypto in 2025. The IRGC, Iran’s military and intelligence apparatus, received over $3 billion in crypto in 2025, accounting for more than half of all Iranian crypto activity by the final quarter. Those funds went to militia networks, dual-use military equipment, and oil sales. In late 2024, Iran’s Defense Export Center began accepting cryptocurrency for military hardware: armed drones, ballistic missiles, naval vessels. By March 2026, the regime was charging ships up to $2 million each to transit the Strait of Hormuz, payable in Bitcoin.

Russia followed a parallel path. After its invasion of Ukraine triggered sweeping sanctions, crypto became a primary tool for evasion. The Duma legalized cryptocurrency for international settlements in July 2024. Putin announced the legalization of crypto mining the following month. More than a hundred no-KYC exchanges now operate in Russia, allowing users to transfer funds from sanctioned banks and receive crypto without verifying their identity.

And then there is North Korea. The Lazarus Group, the regime’s military hacking unit, stole $2.02 billion in cryptocurrency in 2025. The $1.5 billion Bybit heist was the largest single theft in the history of cryptocurrency. North Korean operatives created fake identities, applied for remote jobs at Western crypto startups, worked as employees for months learning internal systems, then drained the treasuries. An investigation found nearly one hundred alleged North Korean engineers embedded inside more than fifty Web3 projects. The cumulative total stolen by the regime across roughly 270 documented incidents sits at an estimated $6.75 billion. U.S. officials have estimated that about half of the regime’s missile program is funded through these operations, though the precise breakdown of funding sources remains classified.

The pattern is the same at every scale. An individual buying drugs on the dark web and a nation-state funding a nuclear weapons program are using the same feature of the same technology for the same reason: no one can see who they are.

When sanctions lose their teeth, the menu of options shrinks. Diplomacy without economic leverage is diplomacy without consequences. What remains is confrontation, and confrontation means people dying in ways that financial pressure was specifically designed to prevent. Every dollar that moves through an anonymous crypto wallet and into the hands of a sanctioned regime is a dollar that weakens the alternative to war.

Cryptocurrency’s advocates describe anonymity as a right. For the people funding missiles, it is a business model.

In defense of Brad & Chad

The person buying fifty dollars of Ethereum is not trying to fund missiles.

The person holding Bitcoin because he distrusts banks is not trying to help a sanctioned regime move money. Brad and Chad are not sitting in an apartment somewhere drawing up a naval procurement plan for Iran.

But systems do not only serve the intentions of their smallest users. They serve the needs of their most powerful users.

A technology built to move money without a trusted institution will be most valuable to the people who cannot use trusted institutions. A technology built to avoid identification will be most valuable to the people who cannot survive identification. A technology built to route around state power will be most valuable to the states, militias, oligarchs, traffickers, and political machines trying to route around consequences.

That does not mean every user is guilty. It means every user is inside the same machine. Brad and Chad are a cog in the machine.

The quarter-billion-dollar campaign to keep it this way

If everything you just read is true, the natural question is: why isn’t the government doing something about it?

The answer is that the people profiting from the system are spending unprecedented amounts of money to make sure no one writes the rules.

In the 2024 election cycle, crypto corporations poured between $119 million and $245 million into federal elections, depending on how you count affiliated spending. By the narrower measure, that $119 million in direct corporate contributions represented forty-four percent of all corporate money in the entire election. Not forty-four percent of tech money. Forty-four percent of all corporate money, from every industry combined. The crypto super PAC Fairshake, funded primarily by Coinbase, Ripple, and Andreessen Horowitz, raised $202.9 million and spent it electing pro-crypto candidates at a ninety-one percent success rate across fifty-eight targeted races.

The spending was not abstract. It had specific targets.

Senator Sherrod Brown of Ohio was the chairman of the Senate Banking Committee and one of the most prominent voices calling for crypto regulation. Fairshake spent $40 million backing his Republican opponent, Bernie Moreno. Moreno won. The chairman of the committee most responsible for regulating crypto was removed from office by crypto money.

In California, Fairshake spent $10 million opposing Katie Porter’s Senate bid. Porter’s crime was not a legislative record on crypto. She was an ally of Elizabeth Warren, whom the industry considers its primary opponent. Porter lost her primary.

Representative Jamaal Bowman of New York received $2 million in Fairshake spending against him. Representative Cori Bush of Missouri received $1 million. Representative Al Green of Texas, a longtime crypto critic, faced $1.5 million from Fairshake’s affiliate PAC in 2026 and was given an “F” rating by Stand with Crypto, a Coinbase-backed scorecard, based largely on a single statement referring to his opponent’s backers as “MAGA-backed crypto bros.”

For the 2026 midterms, the industry has committed $288 million. The previous record for corporate group spending in a midterm election was $18 million, by the National Association of Realtors in 2022. Crypto is outspending that record by a factor of sixteen.

The effect is visible across party lines. Under Biden, Democrats were broadly skeptical of the crypto industry. Within a year of this spending, over a hundred House Democrats were voting for crypto-friendly legislation that would exempt parts of the industry from the same financial oversight applied to banks and brokerages. The shift did not happen because the evidence changed. It happened because the money changed.

There is an irony at the center of this. The entire philosophy of cryptocurrency is decentralization. No one in charge. No concentration of power. And the industry built on that philosophy has become one of the most aggressive, centralized political forces in the country, using the money it made from “decentralization” to purchase centralized power over the legislative process.

And nowhere is this more visible than at the top.

The Trump family launched World Liberty Financial, a crypto venture whose token and associated products have generated at least $1.4 billion in revenue over sixteen months, with billions more in unsold tokens. The president’s $TRUMP memecoin, launched on January 17, 2025, surged to a $27 billion market cap within a day. It has since fallen ninety-six percent. Fifty-eight wallets made millions. Seven hundred sixty-four thousand wallets lost money. The token’s code automatically routes a cut of every transaction to wallets tied to the project’s creators, meaning the Trump-affiliated entities continue profiting from trading activity regardless of the price collapse. More than $324 million in fees have flowed to those wallets.

Meanwhile, the Trump administration dropped or froze eighty-nine cryptocurrency enforcement cases, dismissed SEC lawsuits against Coinbase and Kraken without penalty, and disbanded the Justice Department unit responsible for crypto crime regulation. A top DOJ official who shut down enforcement against crypto companies held more than $150,000 in personal crypto investments.

Senators Elizabeth Warren and Jack Reed raised concerns that World Liberty Financial may have sold tokens to buyers linked to North Korean hackers, Russian sanctions evasion networks, and Iranian exchanges. They called for a national security investigation. The company disputed some of the specific claims, and some turned out to be false alarms. But WLF also confirmed it had rejected millions of dollars from purchasers who failed its compliance screening. If the filters caught millions in suspect money trying to get in, the question is not whether the filters worked. The question is who was trying to buy in.

A sitting president is profiting directly from the same unregulated system his administration is responsible for regulating, while that system is being used by the adversaries his government claims to be confronting. This is not an opinion. It is a sequence of events that anyone can verify.

What Brad and Chad think

At this point in the article, the crypto bros are probably ready to burn me at the stake. You have either read this far because you agree or you are holding an objection and waiting for it to be addressed. Or, in the case of Brad and Chad, you are waiting for me to print my address across the bottom of the article so you can come and take up your objections with me in person. (Please don’t.) There are a lot of different arguments people use when they argue for cryptocurrency, but let’s talk about the main ones.

The technology is useful, even if the currency is not. This is true, in a narrow and specific way. The cryptography underneath a blockchain is not particularly advanced. Public-key cryptography, the method that lets you sign and verify a transaction without revealing your identity, dates from the 1970s. What Bitcoin introduced was the removal of the trusted intermediary: the bank, the clearinghouse, the institution that sits between you and the person you are paying. In theory, removing the middleman is elegant. In practice, the middleman being removed is a licensed, regulated institution with consumer-protection obligations, and removing it is not an upgrade. It is the opposite of one. There are legitimate uses for distributed ledger technology, mostly in cross-border settlement and supply-chain tracking, and those applications have been quietly built by banks and logistics firms without using Bitcoin or any cryptocurrency at all. The utility is real. It is also small. The industry has sold the public on it being neither.

Crypto is a hedge against inflation. The daily volatility of Bitcoin, measured against gold or the dollar, has for most of its history been roughly four to five times higher than either. A hedge is something that reduces risk. Something that can lose half its value in a week is not a hedge. It is a risk. What Bitcoin has been, over certain long windows, is an asset that has appreciated. That is a different thing. And it has appreciated because more people have bought it. (So did Beanie Babies).

You are just mad you did not get in early. So is everyone that got involved in a Ponzi scheme. Do you think everyone who “invested” with Bernie Madoff is upset they didn’t get in earlier? The early investors with Bernie Madoff made money too (at first). Nobody would look at them and say the problem with Madoff’s critics is that they were jealous they did not get in sooner. If a structure’s stability depends on newer participants paying older participants, and the older participants’ response to criticism is that the critics should have joined earlier, the argument has quietly admitted what it is. It is not a defense of the system. It is a description of the system. And it matches, almost word for word, the SEC’s working definition of a Ponzi scheme.

The banking system is broken too. Yes. It is. The American financial system is unfair in specific, measurable, documented ways. Black and Latino borrowers pay higher rates for the same risk profile. Overdraft fees extracted twelve billion dollars from low-balance accounts in 2024 alone. Consumer protection structures are weaker than they should be. That is a real problem. It is also the argument for fixing the system, not for replacing it with one that has no protections at all. A broken accountability structure can be rebuilt. A system designed to have no accountability structure cannot be made accountable without rebuilding it from scratch.

Crypto cannot be banned anyway. Correct. Nobody serious is arguing that it should be (that ship has sailed). The European Union has not banned crypto. Japan has not banned it. Singapore has not banned it. The argument I am making is not about prohibition. It is about whether the instruments that make sanctions evasion, nuclear procurement, and political capture possible will be permitted to continue operating without a regulatory floor. That question has been answered in other countries. It has not been answered here.

An important caveat: the crypto industry is not a monolith. There are developers building on blockchain technology who genuinely want regulation. Smaller exchanges have lobbied for licensing standards that would weed out fraud and separate legitimate operations from criminal ones. Compliance-oriented companies have spent years asking for clear rules so they can build businesses without waiting to see if the federal government decides to sue them. The internal fight over what crypto should become is real and ongoing. But the people winning that fight are not the developers asking for clarity. The people winning that fight are the ones writing the checks, and the checks are buying a version of the future that serves the checkwriters.

Share

So what do we do?

The situation is not hopeless. Other countries have already started doing what the United States has not.

The European Union passed a comprehensive crypto regulation that went into full effect in December 2024. It requires crypto companies to obtain licenses, verify who their customers are, maintain reserves, and disclose risks to buyers. In its first months, crypto scam reports dropped fifty-eight percent compared to the same period the year before, and regulators forced hundreds of unregistered crypto firms to shut down. The United Kingdom, Singapore, and Japan have built similar frameworks, all applying the same basic principle: if you handle people’s money, you follow the same rules whether you operate out of a bank branch or a blockchain.

These are not hostile-to-crypto jurisdictions. They are countries that decided clear rules protect everyone, including legitimate businesses that want to operate without being confused with criminals.

The United States, meanwhile, has passed almost nothing. The only major crypto legislation signed into law covers one narrow corner of the market and explicitly exempts it from the same oversight applied to banks and brokerages. Broader regulation has stalled in the Senate. If it does not advance before Congress shifts to campaign mode this summer, the country defaults to what it has been doing for years: no rulebook, just lawsuits.

The single thing the crypto industry fears most is disclosure. Real regulation would require crypto projects to reveal who runs them, how they make money, what risks they carry, and where the money goes. Anonymous founders would have to identify themselves. Projects with no revenue model would have to say so. The concentration of tokens in insiders’ hands would be public information. The reason the industry has spent $288 million on midterm elections is not because it opposes all regulation. It is because it opposes the kind of regulation that would require it to tell you the truth.

There are two things you can do about that.

The first is vote, but the vote only works if you know what you are voting for. Check who is funding the candidates on your ballot. OpenSecrets.org publishes donor records for every federal race. Stand with Crypto, the Coinbase-backed scorecard at standwithcrypto.org, rates politicians on their crypto positions; from their perspective, an “A” means the candidate supports the industry, so read those grades with the full context of what you have just learned. Ask whether the candidate who promises to represent you has already been purchased by an industry that spent more on the last election than every other corporate sector combined.

The second is simpler. You do not have to participate. Remember the structure from earlier: the only way the system works is if new money keeps arriving. Early holders profit when new buyers show up. The price only goes up if someone else buys after you. The entire machine requires recruitment. It requires your money. It requires you. Not participating is not passive. It is a refusal to be the fuel.

The crypto industry has spent enormous resources convincing people that this is about innovation, financial freedom, and the future of money. It is about money and freedom, but not yours. It is about the freedom of oligarchs, arms dealers, sanctioned nations, and a sitting president to move money without anyone watching. It is sold to regular people as “avoid those pesky banking fees” so that it looks like a consumer product instead of what it is: infrastructure for crime at global scale.

It serves the people who need to move money without anyone watching. It serves the early holders who profit when new buyers arrive. It serves the corporations that took a decentralized philosophy and used it to build centralized political machines. It serves the arms dealers, the sanctioned regimes, the intelligence operatives who embed themselves inside startup companies and drain the treasury. It serves a president who profits from the product he is supposed to regulate while his administration dismantles the agencies that investigate crypto crime.

It does not serve Chad and Brad. And it does not serve you.

The system that crypto was built to replace is broken. The banks failed. The bailouts were real. The anger was justified. But the answer to a broken system is not a system with no accountability at all. Cryptocurrency is the same broken system without a license, without oversight, without the ability to pull the plug when things go wrong, and with a wealth concentration that makes the traditional banking system look egalitarian by comparison. The choice between a flawed system with accountability and a flawed system without it is not a difficult one.

None of this is hypothetical. North Korea already has the money, the politicians are already purchased, the enforcement agencies are already disbanded. The danger isn’t coming. It’s here and it’s operating exactly as designed.

Share The Minuteman

The question to carry forward.

When a technology promises freedom, ask freedom for whom.

Freedom for the person charged overdraft fees by a bank is one thing. Freedom for a sanctioned regime to buy military hardware is another. Freedom for a worker to send money across borders is one thing. Freedom for a president to profit from the market he is supposed to regulate is another.

The same word can hide very different beneficiaries.

If this piece helped you see the machinery underneath the sales pitch, send it to someone who keeps telling you crypto is “the future of money” but has never been told whose future it is making easier.

Share

Sharing is caring :)

Subscribe now

He who controls the logisitcs controls the economy. You don’t need a standing army to conquer a country. You don’t need to fire a single shot. If you own the transportation network and you own the land around it you own the nation.

Share

Read more

A note before we start:

I did not expect to spend this much time thinking about bananas.

There are nobler fruits, I suppose. Apples have gravity and sin. Grapes have wine. Figs have the Bible. The banana mostly has a sticker and a short shelf life.

But that is why I wrote this.

The banana is one of those objects that modern life teaches us not to see. It is cheap. It is everywhere. It arrives in the grocery store looking inevitable, as if nature itself individually wrapped breakfast and put it on sale.

But nothing this cheap, this global, and this ordinary is ever just ordinary.

I wrote this because the banana is a small yellow receipt for a much larger system. It carries the story of corporations that learned how to become governments without calling themselves governments. It carries the story of land, labor, ports, railroads, courts, armies, marketing, and the quiet genius of making power look like convenience.

Read this with one question in mind:

Who owns the road between you and the thing you need?

Because once you start seeing that question in a banana, you start seeing it everywhere.

Subscribe now

The Banana Blueprint for Corporate Sovereignty

There is a good chance there is a banana on your counter right now.

It is not supposed to be there. Bananas do not grow here, and they have never grown here in any quantity that could feed a country, because the plant requires warm tropical air, drenching humidity, and soil that does not freeze. Your banana probably came from Ecuador, or Guatemala, or Colombia, or Costa Rica. It was cut green off a cloned plant, sealed into a refrigerated ship, unloaded at a port like Gulfport or Wilmington, trucked to a ripening chamber in your region, gassed with ethylene to turn it yellow at exactly the moment the grocer needed it yellow, and set on a shelf next to apples for about sixty cents a pound.

It is the cheapest fruit in the store. It might be the cheapest thing in the store.

It is also, by the measure that matters, the most expensive piece of produce on your counter.

The banana is built on four thousand workers dead in twenty-five miles of jungle.

It is built on fifteen hundred men permanently sterilized by chemicals their employers knew were poisoning them.

It is built on the first democratic government in Guatemala, overthrown in 1954 in a CIA operation run by former lawyers for a fruit company, which touched off a civil war that killed two hundred thousand people.

It is built on a century-long racial caste system imported to Central America from the American South, and on payments from an American corporation to a Colombian paramilitary death squad that the United States had designated a terrorist organization.

Some of this has been admitted in federal court. The rest is in the historical record.

None of it is in the price.

That gap, between what a banana costs and what a banana has cost, is the subject of this article. The gap is not an accident. It is the point.

A corporation called the United Fruit Company figured out, over the course of a hundred and fifty years, how to push the cost of a piece of fruit off the balance sheet and into other people’s bodies, into other people’s countries, into other people’s history. The techniques it pioneered are now the standard operating procedure for every global corporation you interact with. Your phone runs on them. Your groceries run on them. Your cloud runs on them.

The banana is a playbook.

A century ago a banana company wrote the book on corporate sovereignty. Own the logistics, enforce the monoculture, privatize the infrastructure, and govern the countries you move through, all packaged into a cheap yellow fruit. Every modern conglomerate, from United Fruit to whoever owns your cloud, is running the same century-old code.

Share

The store clerk from Brooklyn

The story does not begin with a fruit. It begins with a railroad.

In the early 1870s, Costa Rica had a profitable product and a terrible logistics problem. The country ran on coffee, grown in the Central Valley around the capital of San José, where the elevation and climate produced beans that European buyers in London and Paris would pay premium prices for. Getting the coffee to London, however, was a nightmare.

The Panama Canal did not exist yet. To reach Europe, a Costa Rican exporter had to load coffee onto wooden ox carts, haul it for days over mountain passes to the Pacific port of Puntarenas, sail it down the coast of South America, round Cape Horn, and come back up the Atlantic. Shipping ate most of the profit.

The obvious solution, obvious to anyone who looked at a map, was to build a railroad straight east across the country from San José to the Caribbean port of Limón. Coffee on the Atlantic side. Straight shot to Europe.

The Costa Rican government hired a prominent American railroad man named Henry Meiggs to build it. Meiggs, busy with a project in Peru, sent his nephew to manage the Costa Rica side. The nephew was twenty-three years old. He was from Brooklyn. He had worked as a store clerk, then managed a cattle ranch in Texas. He had never seen a tropical rainforest. His name was Minor Cooper Keith, and in 1871 he stepped off a boat into the swamp that was going to make him one of the most powerful private citizens in the Western Hemisphere.

The Atlantic coast of Costa Rica in 1871 was not the tourist destination it is today. It was a vast, humid, perpetually raining lowland forest, cut through by shallow rivers that spilled over into standing water and bred mosquitoes in biblical numbers.

Germ theory was still new. Nobody understood that the mosquitoes were the disease vectors. The prevailing idea was that the swampy air itself was bad, that “miasma” was what made workers sick. What made workers sick was malaria, yellow fever, and dysentery, and what the workers mostly did was die.

Roughly four thousand people died building the first twenty-five miles of track. Keith’s uncle Henry Meiggs died in 1877. Three of Keith’s own brothers died, all of them in the jungle, all of them from tropical diseases.

The logical response to watching your uncle and three brothers die in a swamp would be to get on a boat and go back to Brooklyn. Keith did not do that. He took over the project and pushed onward.

Which turned out to be a problem, because the project was running out of money.

By 1882, the Costa Rican government, which had financed most of the railroad through British banks, defaulted on its national debt. The London bondholders were not going to get paid. The workers were not going to get paid. The tracks stopped where the money stopped, and the jungle began reclaiming them.

This is the moment the blueprint begins.

Keith did something that had no modern precedent. He did not file suit. He did not petition the American government to intervene. He got on a ship, sailed to London, walked into the offices of the banks holding Costa Rica’s debt, and personally negotiated the sovereign debt of a country he did not run.

He raised one point two million pounds sterling in new financing. He got the interest on the existing loans reduced from seven percent down to two and a half. He saved the Costa Rican treasury from full collapse.

A private American in his twenties, with no official title, acted as the central bank and chief diplomat of a sovereign state, and it worked.

The question, of course, is what he asked for in return.

The contract he got in return is where the blueprint is born. It is called the Soto-Keith Contract. It was signed on April 21, 1884. And it is, without much exaggeration, the single most consequential private agreement in the history of Latin America.

For saving the country’s finances and finishing the railroad, Keith received a ninety-nine-year lease on the only railway line connecting the Costa Rican interior to the world. He received eight hundred thousand acres of land along the tracks, completely tax-free, to do with as he pleased for ninety-nine years.

Eight hundred thousand acres works out to roughly 6.3% of Costa Rica’s total land area. To put that in modern terms, imagine if a single American businessman, in exchange for building a 5G network, received tax-free ownership of something on the order of the state of Delaware inside the United States, plus a hundred-year monopoly on every major highway in and out of it. No property tax. No federal oversight. Corporate sovereign territory, courtesy of the nation that owned it yesterday.

He owned the land. He owned the only way in or out of the land. In a functional sense, he was now the parallel government of a piece of a foreign country.

The railroad was finished on December 7, 1890. And then something almost funny happened.

It turned out the railroad was a financial catastrophe. Coffee exports and passenger traffic did not come close to covering the debt on its construction. The mathematics that had seemed so obvious on paper did not hold up when the trains actually ran. Keith was sitting on the most expensive piece of infrastructure in Central America, and it was losing money. He was staring at bankruptcy for the second time in a decade.

What saved him was an afterthought.

Back in 1873, while construction was still killing workers by the thousand, Keith had faced a mundane problem. The railroad crews needed food, and imported food was expensive, and the jungle the crews were dying in was not exactly teeming with farmable land.

Somewhere along the way Keith got some banana roots from a Frenchman and started planting them alongside the tracks as cheap food for the workers. The plant grew fast, produced calorie-dense fruit in its own packaging, and thrived in exactly the climate that was killing everyone. It was never meant to be a business. It was lunch.

In 1890, with an empty train and a failing railroad, Keith started throwing some of the bananas onto the empty cars going back to the coast and shipping them to New Orleans. He wanted to see if he could make a few extra dollars. Instead he discovered an industrial revolution in fruit.

The United States was in the middle of a massive urbanization boom. Cities were full of people who needed cheap, portable, easy-to-eat food. A banana arrived in its own wrapper, packed with sugar and potassium, and could be peeled and eaten without utensils in thirty seconds. It was a revelation.

The script flipped. The railroad that was supposed to haul Costa Rican coffee to the world started hauling bananas to the United States. Keith bought his own steamboats and formed the Tropical Trading and Transport Company. The train that the Costa Rican government had financed to move its own national economy was now moving Keith’s private fruit business. And that is when Keith, and eventually everyone else, figured out the single most important thing I am trying to show you.

Whoever controls the logistics controls the economy. The product is almost incidental.

The blueprint takes shape

Minor Keith was a builder. He was not a manager. By the end of the 1890s, having bought up more and more land and more and more track, he had overextended himself to the point of bankruptcy again. In 1899 he did what builders do when they run out of money. He merged.

His merger partner was the Boston Fruit Company, run by a hardened sea captain named Lorenzo Baker and a Boston businessman named Andrew Preston. Boston Fruit had mastered the shipping side of the banana business, hauling fruit from Jamaica to the American Northeast, which was a tougher voyage than the run from Costa Rica to New Orleans. Boston Fruit had the ships and the East Coast distribution network and the New England capital. Keith had the land, the railroads, and the political leverage. On March 30, 1899, they combined to form the United Fruit Company.

From the day it was born, the company looked less like a farm and more like a small country.

By the first decade of the twentieth century, United Fruit owned something in the neighborhood of 800,000 acres of land across Central America.

It owned the company towns the workers lived in. It owned the company stores, called commissaries, where the workers were required to buy their food and tools. It paid workers largely in company scrip, a currency that was only redeemable at those stores, which meant that wages left the company’s books and came right back to the company’s books on the same day. It owned the hospitals. It owned the schools. It owned, through a subsidiary called Tropical Radio and Telegraph, the radio and telegraph infrastructure of an entire region, which meant that any message sent out of Central America, whether by a local government or by a journalist wiring a story to New York, passed through United Fruit’s cables. It owned a private steamship fleet called the Great White Fleet, which at its peak carried more tonnage than the navies of most Latin American countries combined.

By one accounting, United Fruit moved 58% of Costa Rica’s total national exports through infrastructure it had built and owned.

A normal industry of that size would act as an economic engine in the country it operates in. It would hire local vendors, buy local materials, pay local taxes, and push wealth outward through the local economy. United Fruit did the opposite. It built sealed corporate enclaves. The money generated inside the enclave stayed inside the enclave, and then the enclave put it on a ship and sent it back to Boston.

You can see this in the railroad maps. If you look at historical maps of Honduras or Costa Rica in the early 1900s, the rail lines do not go where a national planner would put them. They do not connect the capital to the ports. They do not pass through the major existing population centers. They do not go up into the highlands where most of the people live. They loop and curve and reach deep into unpopulated jungles.

The reason is that United Fruit built them to connect United Fruit plantations to United Fruit ports, and if the train stopped in a local town on the way, then the locals might actually benefit from it. The enclave was the product, and the enclave was supposed to be airtight.

Out of this arrangement came a phrase most Americans have heard and almost none of us know the origin of. The phrase is banana republic.

It was not originally the name of a chain of retail clothing stores. It was a diagnosis.

It described a small Central American nation whose economy was so thoroughly dependent on a single exported crop controlled by a single foreign corporation that its political system became a function of the corporation’s needs.

The company picked the president. The company wrote the laws. The company paid the army. The country had a flag and a capital and a constitution, but those things were the paint job.

Throughout Latin America, United Fruit was not called the United Fruit Company. It was called el pulpo. The octopus. Because the tentacles reached into everything.

The octopus had a labor problem, and the way it solved it is worth pausing on, because you will see the solution again later in this article, in different clothes.

The Atlantic coast of Costa Rica, where the original plantations were, was the same lethal swamp that had killed four thousand railroad workers. The majority population of Costa Rica, the cooler-highland Hispanic Ticos, would not go there. They had watched their countrymen die of malaria for twenty years. You could not pay a Tico enough. Which meant the bananas sat on the trees and rotted, and the company faced a labor crisis that threatened the entire operation.

The solution was to import a workforce from somewhere else.

Between 1910 and 1913 alone, around twenty thousand Jamaican workers were brought into the Limón province of Costa Rica to cut bananas. A broader figure for the whole Caribbean workforce imported for regional banana and canal projects is closer to fifty thousand.

United Fruit chose the British West Indies for three reasons. First, the Caribbean sugar industry was in depression, which meant there was a vast pool of labor that was desperate enough to accept dangerous work. Second, the West Indian workforce had generations of experience in tropical agriculture, including banana cultivation. Third, they spoke English, which meant the Boston managers did not have to learn Spanish to run the plantations.

In practice, this meant that the company imported an entire English-speaking, Protestant, African-descended society into a Spanish-speaking Catholic country, and then built a wall around it in a section of the country the majority population already refused to enter.

Limón became an enclave within an enclave. Limón had English-language schools, Protestant churches, Caribbean food, Caribbean music, and the economic and cultural texture of a small piece of Jamaica transplanted into a corner of Central America.

The Costa Rican political elite, meanwhile, was in the middle of constructing a national identity built on a myth of European whiteness. Which meant that the twenty thousand Black English-speaking Protestants in Limón were, from the standpoint of the national project, a problem. They were isolated geographically and politically. Their movements were restricted. They were, legally, not quite citizens.

The company took this arrangement and industrialized it. Inside the plantation system, United Fruit set up an explicit three-tier racial hierarchy.

At the top were white American managers, in segregated company towns with manicured lawns and swimming pools and imported food. In the middle, because they spoke English, were the West Indians, serving as foremen, timekeepers, and clerical staff, often paid in U.S. dollars rather than local currency, and paid at a slightly better rate. At the bottom, doing the most dangerous manual labor, were the Hispanic and indigenous workers. Black workers were paid less than white workers for identical labor.

The hierarchy was not an accident. It was management. The company imported, and then deliberately worsened, the logic of Jim Crow that its Boston executives knew from the American South, because a divided workforce is a workforce that does not strike.

If the Hispanic workers organized, the company could threaten to replace them with West Indians. If the West Indians organized, the company could threaten to replace them with cheaper Hispanic labor. Workers looked sideways at each other instead of upward at the company, which was the entire point.

Everyone was in a dormitory with malaria, and everyone was being paid in scrip, and everyone was being fleeced by the commissary. Nobody could see it, because everyone had been carefully taught to see someone else.

In August 1934, they saw it anyway.

The Great Banana Strike of 1934 ran from August 4 to August 28. It involved more than thirty separate unions and something like a hundred thousand workers across the Costa Rican plantations. It was the largest labor action Central America had ever seen.

United Fruit agreed to a settlement on August 28, branded the strike a communist insurrection in its public relations, and then quietly failed to honor the settlement. It took another four years of fighting to pry a collective agreement out of the company in 1938.

But the more consequential fallout came in the contract the Costa Rican government signed with United Fruit in December 1934, as part of the arrangement to move the bulk of the company’s operations over to the Pacific coast, where the soil was still fresh and the workers had not yet organized.

That contract contained a clause. It explicitly forbade people of colour from working at or moving to the new Pacific plantations.

The company, which had imported twenty thousand people to harvest its fruit, agreed to it instantly. The West Indians were no longer useful. They had started asking for more money, and the Pacific side would be cheaper to staff with local Hispanic labor that had not yet organized.

From 1935 onward, Black Costa Ricans were barred from many public parks. English-language schools were closed. Citizenship was withheld even from West Indians born in the country. The community the company had built and then discarded was written out of the national story.

This is probably a good moment to pause and look at what we have.

By the mid-1930s, United Fruit had been running a continent-scale experiment, and the experiment had converged on a repeatable pattern. Build or take over infrastructure that a country needs. Use that infrastructure to lock up the land around it. Vertically integrate the entire value chain so that no outside competitor or regulator has a place to stand. Engineer a dependent workforce and prevent it from organizing by whatever means necessary, up to and including the importation and later the legal abandonment of entire ethnic groups. Own the communication channels so that you, and only you, shape the story about what is happening inside the enclave. And, when any of that becomes politically inconvenient, use your influence with the metropolitan government to make the problem go away.

The last step on that list is the one the company had not yet needed to use at full strength.

That changed in 1954.

Share The Minuteman

The coup

By the early 1950s, the United Fruit Company was the largest landowner in Guatemala.

It owned something like 550,000 acres, or roughly 42 percent of the country’s arable land, and it was using only about fifteen percent of that land. The other eighty-five percent, the vast majority of it, was held idle.

This was partly as a hedge against plant disease, which was starting to ravage the company’s Atlantic operations, and partly as insurance against potential competitors, who could not enter the market if United Fruit was sitting on all the best soil.

This became a problem when Guatemala held democratic elections in 1950 and elected a reformist army colonel named Jacobo Árbenz.

Árbenz was not a communist. He was a nationalist with a common-sense observation. In a country where the majority of the population were landless and lived in poverty, a foreign corporation should not be allowed to leave 85 percent of the country’s best land sitting empty as a business hedge.

His response, passed in 1952 and implemented starting in 1953, was a land reform law called Decree 900. Decree 900 authorized the government to expropriate uncultivated land from large landowners, including domestic Guatemalan elites, and redistribute it to landless farmers.

The key detail, the detail that matters, is that Árbenz was not seizing the land. He was buying it. Decree 900 promised compensation to the landowners, paid in government bonds. And he announced that the price paid for the land would be the same price the landowners had declared on their own tax returns in 1952.

It was a clever move. United Fruit had been dramatically undervaluing its Guatemalan land holdings on its tax filings for years, in order to minimize the taxes it owed to the Guatemalan government. Árbenz called the bluff.

If you say it is only worth this much, he said, I will pay you exactly that much. The government offered United Fruit $627,572 for the expropriated land, because that was what United Fruit had said the land was worth. United Fruit suddenly insisted the land was actually worth $15.8 million, which was a number it had never paid taxes on. The United States State Department took up the demand.

At this point, the company made one of the more successful bets in the history of corporate public relations. It hired Edward Bernays.

Bernays is now understood to be the father of modern public relations. He was also Sigmund Freud’s nephew, which is the kind of biographical detail that used to sound random and now seems almost inevitable.

Bernays had, for decades, been working out how to manufacture mass opinion on behalf of corporate clients. He ran a front organization for United Fruit called the Middle America Information Bureau, founded in 1943, which quietly fed “background information” on Guatemala to something like twenty-five thousand American journalists. He sent a weekly Guatemala Newsletter to two hundred and fifty working reporters.

In January 1952 he flew a cohort of prominent American journalists, including a New York Times reporter named Will Lissner and the columnist Walter Winchell, on an all-expenses-paid junket to Guatemala, where Bernays controlled what they saw and whom they spoke to.

The narrative Bernays manufactured, in that moment, was the most useful narrative available in the early 1950s: communism.

The Korean War was underway. Senator McCarthy was in full cry. The American public had been trained, through years of cultural priming, to respond to the word “communist” the way a Pavlov’s dog responds to a bell.

Bernays framed Árbenz’s modest land reform law not as a national development project, which it was, but as a Soviet beachhead in the Western Hemisphere, which it was not. Major American publications including the New York Times, the New York Herald Tribune, Time, Newsweek, and the Atlantic Monthly picked up the framing. The story that Guatemala, a Catholic country with a reformist colonel for a president, was falling to Moscow reached the American public as established fact.

United Fruit had also, as a matter of long-standing corporate relationships, been a client for decades of the New York law firm Sullivan & Cromwell.

Sullivan & Cromwell’s senior partner for much of that period was John Foster Dulles, who personally negotiated major United Fruit land concessions in Central America during the 1930s. His brother Allen had also worked at the firm and had documented financial ties to United Fruit of his own.

In January 1953, John Foster Dulles became the United States Secretary of State. His brother Allen became the Director of the Central Intelligence Agency. Neither disclosed the full extent of the relationship when the Guatemala question came across their desks.

The Secretary of State and the Director of the CIA, looking at the Bernays-shaped public opinion in American newspapers, decided that Guatemala was a national security problem. (are we clearly seeing the modern parrallel yet???)

The operation authorized by President Eisenhower in early 1954 was called Operation PBSUCCESS, and it is worth a short paragraph in any honest account of American twentieth-century history.

The CIA did not send troops into Guatemala. The CIA sent radio transmitters. A clandestine broadcasting operation called Radio Liberación launched on May 1, 1954, beaming signals into Guatemala from a station in Opa-locka, Florida. The broadcasts described, in dramatic detail, a rebel army of five thousand men marching on the capital. The actual invasion force the CIA had backed was fewer than two hundred Guatemalan exiles. American pilots bombed empty fields to create the impression of a military campaign. The Guatemalan army, hearing what it believed to be overwhelming American-backed force approaching and seeing no reliable signal that the story was fake, lost confidence in the president. Árbenz resigned on June 27 and fled the country.

The military dictatorship that replaced him reversed Decree 900 almost immediately. The expropriated land was handed back to United Fruit.

What came after is a fact that most Americans will live and die without ever knowing, and should know.

Guatemala did not go back to normal. Guatemala plunged into a sequence of military regimes and a civil war that ran, depending on how you count, from 1960 or 1962 until the peace accords of 1996.

When the war ended, the United Nations-backed Commission for Historical Clarification documented more than two hundred thousand people killed or disappeared. Eighty-three percent of the identified victims were indigenous Maya. The Guatemalan army destroyed six hundred and twenty-six villages. One and a half million people were displaced inside the country. More than a hundred and fifty thousand fled as refugees to Mexico. The Commission concluded that the Guatemalan state, over the course of the war, had committed acts of genocide against its own Mayan population. Ninety-three percent of the documented human rights violations were attributed to state forces and their paramilitary allies.

Two hundred thousand people dead and a generation of Maya communities destroyed, in part, because a fruit company wanted to keep its unused land.

What the coup showed, to anyone who was paying close attention, was the last and highest move in the blueprint.

True corporate power, in its most developed form, is not the ability to hire lawyers. It is not the ability to lobby Congress. It is the ability to intertwine your private profits so completely with the geopolitical interests of a superpower that the superpower will overthrow a democracy on your behalf and call it a national security operation.

Subscribe now

The monoculture

To understand why United Fruit was so willing to overthrow a government in order to protect 85 percent of its land from being redistributed to poor farmers, you have to understand the plant it was growing.

The banana you eat is not, in any meaningful biological sense, a fruit. It is a genetically cornered plant.

Commercial bananas are triploid. Instead of having two sets of chromosomes they have three, which means the chromosomes cannot pair up properly during reproduction. The plant produces fruit without fertilization. That is what the condition called parthenocarpy refers to. No pollination, no seeds, no sexual reproduction. Those little black dots running down the center of a banana when you cut it in half are the remains of ovules that never developed. A commercial banana is a reproductive dead end.

The way you plant a new banana, therefore, is not by planting a seed. You dig down to the underground stem, called the corm, of an existing banana plant, cut off a piece, and stick it in the ground somewhere else. The new plant grows from the cutting. It is an exact genetic clone of the plant you took it from.

Which means that the variety of banana United Fruit built its entire empire on, a variety called the Gros Michel, was not a population of related plants. It was one plant, replicated millions of times across eight hundred thousand acres.

Every banana in every crate in every cargo hold of every Great White Fleet ship was genetically identical to the banana next to it. It was the largest continuous monoculture in the history of agriculture. Zero genetic variation. One plant, repeated forever.

This is an extraordinary business arrangement, and it has one obvious vulnerability that is visible to anyone who has ever taken a middle school biology class.

In a natural forest, if a pathogen attacks a plant species, some of the plants happen to have a genetic mutation that makes them resistant. They survive. They pass the resistance on. The species adapts. In a monoculture, there is no mutation. If a pathogen figures out how to kill one plant, it has the key to every plant on the continent.

Nature, given enough time, always figures it out.

In this case, nature was a fungus.

Its name is Fusarium oxysporum f. sp. cubense, Race 1. It became known as Panama Disease, because the first outbreak that got anyone’s attention happened in Panama. It works by entering the banana plant through its roots, colonizing the vascular system, and clogging the internal tubes the plant uses to pull water up out of the soil. The plant, sitting in a swamp full of water, dies of thirst.

Once the fungus is in the soil, it cannot be removed. Its dormant cells, called chlamydospores, survive in the ground for decades, immune to every fungicide ever marketed against them. A field infected with Race 1 cannot grow a Race 1-susceptible banana. Ever.

The company’s response to this was, in retrospect, its most revealing single habit. It did not invest in breeding resistant varieties. It did not try to fix the soil. It did not reduce the scale of the monoculture. It ran.

The practice is called shifting plantation agriculture, and it worked like this. When Panama Disease infested a plantation, the company abandoned the plantation. It moved twenty miles down the road into a patch of virgin rainforest, clear-cut the forest, drained the wetlands, built a new railroad spur to the coast, and planted the same Gros Michel clones all over again.

The fungus, in a season or two, would find the new plantation. The company would abandon that plantation and move another twenty miles down the road. For decades, the banana industry tried to outrun Panama Disease by destroying virgin rainforest at a pace the pathogen could not match.

By the 1950s it had stopped working.

The company ran out of accessible rainforest. The Gros Michel was, as a commercial product, functionally extinct. Every American grandparent who complains that bananas used to taste better is, in most cases, not wrong. That banana is gone.

The industry solved this problem the way it solved every other problem. It did not change the model. It found another clone.

The banana you are eating today is almost certainly a variety called the Cavendish. United Fruit initially hated the Cavendish. It had a thinner skin, bruised easily during shipping, and tasted, by most accounts, blander than the Gros Michel. But the Cavendish had one useful property, which was natural genetic resistance to Race 1 Panama Disease. You could plant it directly into infected soil and it grew fine. The company retooled its entire logistics pipeline, invented specialized cardboard boxes and refrigerated shipping, and replanted the global banana supply in Cavendish clones. Today, the Cavendish accounts for roughly half of all bananas grown in the world and over ninety percent of bananas exported to Europe and North America.

The industry replaced one genetic monoculture with another genetic monoculture. It learned nothing. It just found a new clone.

This is the moment where I have to stop and say the obvious thing out loud.

The corporate model is itself a monoculture. One crop. One supply chain. One mode of extraction, repeated everywhere. The response to failure, in both cases, is identical. Do not fix the underlying vulnerability. Find an ecosystem that has not yet been exhausted. Replicate.

When bananas fail, the company slash-and-burns its way to new rainforest. When a plantation workforce organizes, the company slash-and-burns its way to a new workforce that has not yet organized. When a country’s courts get ambitious, the company slash-and-burns its way to a country whose courts are more cooperative. The biological monoculture and the corporate monoculture are not analogies to each other. They are the same structure, in different materials.

There is a chemical footnote to the monoculture story that is worth knowing, because it shows up on the counter.

Because the Cavendish is a clone, and because a monoculture has no natural resistance, the Cavendish can only be kept alive through relentless chemical intervention. Synthetic fertilizers. Herbicides. Insecticides. Fungicides sprayed from airplanes up to forty times a year. And in the 1970s, the industry’s solution to a microscopic worm, called a nematode, that was eating the roots of banana plants, was a chemical called DBCP, short for 1,2-dibromo-3-chloropropane.

Roughly fifteen hundred Costa Rican plantation workers were exposed to DBCP, and nearly all of them became permanently sterile. The chemical destroyed their capacity to father children.

The chemical’s manufacturers, Dow and Shell, had internal toxicology data going back years that showed severe reproductive hazards in mammals. The data was concealed. Workers were handed the chemical, told to inject it into the soil, and given no meaningful protective equipment and no warnings. DBCP was banned in the United States in 1977, but American manufacturers continued to sell it in Central America for years after. Fifteen hundred men were permanently unable to have children, and their families were permanently wrecked, so that a grocery store in Ohio could have a perfect yellow banana at fifty-nine cents a pound.

This is the idea I keep circling, and it is the idea the price hides. A cheap banana is not actually cheap. The cost does not disappear. The cost is offloaded. It is paid by environmental destruction in the tropics, by the bodies of workers in the global south, by the violent histories of the countries the fruit is grown in. It is just paid somewhere else, somewhere you cannot see from the checkout counter.

The rebrand

The United Fruit Company does not exist anymore, and that is not an accident either.

In 1970, United Fruit merged with an American meatpacking holding company called AMK Corporation and became United Brands. In 1984, a Cincinnati financier named Carl Lindner Jr. took over and renamed the whole thing Chiquita Brands International. The name was the old marketing mascot, Miss Chiquita, a cartoon woman with a basket of fruit on her head. Light blue sticker. Bright smile. Friendly, harmless, entirely disconnected from the blood in the soil.

Underneath the sticker, the business kept doing what it had always done.

In the late 1990s and the early 2000s, Chiquita’s Colombian subsidiary, a company called Banadex, operated plantations in the banana regions of the department of Antioquia. Colombia at that time was in the middle of a long, horrific civil conflict involving left-wing guerrilla groups like the FARC and ELN and, by the mid-1990s, a sprawling right-wing paramilitary umbrella organization called the AUC, the United Self-Defense Forces of Colombia.

The AUC committed massacres. It tortured. It disappeared people. Its specific targets included trade union leaders, human rights workers, indigenous advocates, and political opponents. It was a death squad that operated, in practice, as private security for wealthy landowners and corporations, clearing the countryside of anyone who might organize labor or advocate for land reform.

On September 10, 2001, the United States government formally designated the AUC a Foreign Terrorist Organization. As of that date, it became a federal crime for any American corporation to engage in any financial transaction with the group.

Between 1997 and early 2004, Chiquita made more than a hundred payments to the AUC, totaling over $1.7 million. The payments continued for more than two and a half years after the terrorist designation. They were written down in Chiquita’s corporate books as “security services” and “security payments.” The Department of Justice’s factual proffer, filed in federal court when Chiquita later pleaded guilty, established that no actual security services were received in exchange. The payments funded AUC operations. Those operations suppressed labor organizing in the banana region.

In 2007, Chiquita pleaded guilty in federal court in Washington, D.C., to engaging in transactions with a designated terrorist organization. It paid a $25 million criminal fine. The fine went to the United States Treasury. The families of the Colombians killed during the period Chiquita was writing checks got nothing.

But the families did not stop. Represented by a group called EarthRights International and a handful of other human rights organizations, the victims’ families filed a series of civil suits in the United States using an old and strange legal tool called the Alien Tort Statute, a law that has existed in more or less its current form since 1789. The Alien Tort Statute allows foreign nationals to sue in American courts for human rights violations committed abroad, provided the defendant is sufficiently present in the United States.

For seventeen years, Chiquita fought the cases. Motions to dismiss. Appeals on jurisdiction. Arguments that no American court had the right to hear the claims. The company’s primary substantive defense, when it came to the substance, was a technicality about causation. Yes, it said, we gave the AUC millions of dollars. No, you cannot prove that a specific dollar paid for a specific bullet that killed a specific person, in a civil war with thousands of armed actors. We cannot be held civilly liable for the violence they committed.

On June 10, 2024, a federal jury in West Palm Beach, Florida, rejected the defense. After weeks of testimony, the jury found Chiquita civilly liable for the murders of eight of the nine victims whose cases were before the court. It awarded the families $38,300,000 in damages.

It was the first time in history a United States jury had held a major American multinational civilly liable for human rights abuses committed by a foreign paramilitary group that the corporation had financed. The jury specifically concluded that Chiquita had “knowingly provided substantial assistance” to the AUC, rejecting the extortion defense entirely.

In October 2024, the presiding judge denied Chiquita’s motion to reduce the damages and entered full judgment. Chiquita has appealed to the Eleventh Circuit, and the appeal is likely to take two years. Thousands of additional plaintiffs wait behind the first bellwether case for trials of their own.

Then, in July 2025, the story moved again, this time not in a Florida courthouse but in a Colombian one. A court in the department of Antioquia handed down criminal convictions against seven former Chiquita and Banadex executives, including two American nationals named Charles Dennis Keiser, who had directed Chiquita’s Colombia operations from 1990 to 2000, and John Paul Olivo, the controller for Banadex and Chiquita’s North American operations from 1996 to 2001. The sentence was eleven years and three months in prison, plus a collective fine of thirteen point eight billion Colombian pesos, roughly three and a half million dollars. Arrest orders were issued immediately. It was the first time, in the entire history of the banana industry, that the human architects of the paramilitary relationship had been criminally sentenced.

It is a remarkable moment. It is also worth being very careful about calling it a reckoning.

A $38 million verdict against a global fruit company is a rounding error. The $25 million criminal fine in 2007 went to the United States Treasury and never reached the victims. The July 2025 Colombian convictions, historic as they are, involve two American nationals who are currently on American soil, where extradition is a long and uncertain process, and whose sentences will, at best, punish a handful of managers for a set of decisions that the company itself, in the aggregate, is clearly designed to keep making.

The banana trade continues. Chiquita still buys, ships, and sells bananas. The supply chain it sits on top of still operates. If Chiquita itself were to collapse tomorrow, another multinational would buy the structure and the supply chain would keep running.

The blueprint is durable in a way that the companies that implement it are not. United Fruit did not die. It molted. United Brands. Then Chiquita. Whatever comes next will have different lawyers and a different sticker and the same operating model.

Subscribe now

The fungus comes back

The other thing the industry does not want you to notice is that it is, right now, watching Panama Disease play out again in real time, and is responding to it exactly the way it responded the first time.

The current pathogen is called Tropical Race 4, or TR4. It is another strain of the same family of Fusarium fungus that killed the Gros Michel. TR4 was first identified in Southeast Asia in the 1960s. For decades it stayed in the Eastern Hemisphere. That is no longer true.

TR4 was confirmed in Colombia in 2019, in Peru in 2021. Venezuela declared a national phytosanitary emergency over TR4 in 2023. Peru’s agricultural health authority eradicated more than four hundred separate outbreaks in the early months of 2024 alone.

As of 2024, TR4 had spread to roughly twenty-three countries. In September 2025, the fungus was officially confirmed in Ecuador, the largest banana exporting country on Earth, which by itself accounts for about one third of the global banana trade.

There is no cure. There is no fungicide that works on it. The current response, when TR4 is found in a field, is to burn the field. The chlamydospores, as in the case of Race 1, live in the ground afterward for decades.

And the thing that is spreading TR4 the fastest, the mechanism by which it is leaping from continent to continent, is the supply chain itself.

TR4 travels on the soil clinging to a worker’s boots, to a tractor tire, to a shared irrigation pipe, to the underside of a shipping container. The global logistics infrastructure that Minor Keith and Andrew Preston built out of a coffee railroad and a Boston steamship company, the infrastructure that made the modern banana possible, is now the infrastructure that is delivering the pathogen that may end the modern banana.

The industry’s single largest vulnerability is a direct consequence of the structural choice it made a century ago to substitute scale for diversity. The industry’s response, at this moment in 2026, is to continue planting Cavendish monocultures and hope that the next clone is found before the current one collapses.

On the labor side, the picture is equally familiar. The most recent field research on Ecuadorian banana plantations, published in 2024 and surveying workers at 140 plantations in the province of El Oro, found that the national minimum wage for agricultural workers is about $460 a month, that workers report routine violations of basic labor standards, that paid holidays and social security enrollment are frequently absent, and that workers who attempt to organize are dismissed. In Ecuador’s El Oro province, where most of the country’s export bananas are grown, the documented unionization rate is roughly one percent.

One percent, in the largest banana exporter in the world, ninety years after the Great Banana Strike of 1934, in a country whose neighbors have spent a century litigating the legacy of the industry it runs on.

The banana on your counter, as of this writing, still costs somewhere around sixty cents a pound. That has not meaningfully changed in years. Nothing that has happened in the last hundred and fifty years, and nothing that is currently happening in Ecuador or Colombia or the soil of an infected plantation, is visible in the price.

The point is not that we should feel terrible every time we eat a banana. The point is that the price is a lie, and it is a specific and engineered kind of lie, and the engineering is the point I’ve been trying to make through this whole article.

Share

The blueprint runs everywhere now

If the only story here were the banana, it would still be worth knowing. But it is not the only story here, and this is the part where, if you have been reading carefully, you probably already see the shape of what is coming.

The United Fruit Company figured out how to do six specific things in sequence, and those six things, performed in that order, constitute a repeatable method for how a corporation can replace a state without ever bothering to declare itself one.

Extract a concession on the infrastructure the country depends on. Use the infrastructure concession to lock up the land and the resources around it. Integrate vertically so that no outside competitor or regulator has a place to stand. Engineer labor dependency, and suppress any attempt at solidarity with whatever tools are handy, from racial segregation to the importation and later the abandonment of entire ethnic groups. Control the flow of information so that the outside world sees only the version of events you authorize. And, when any of the above is threatened, align your corporate interests so tightly with the foreign policy of a larger power that the larger power intervenes on your behalf and calls it the national interest.

Every one of those steps, without exception, is operating in the world right now. In many cases, in the same literal form.

The most direct modern example is a port in southern Sri Lanka called Hambantota. It was financed and built beginning in the 2000s with Chinese state loans as part of the Belt and Road Initiative. Sri Lanka, facing acute pressure on its foreign currency reserves, entered a renegotiation with China. The resulting agreement was signed on July 29, 2017, and under its terms, China Merchants Port Holdings received a ninety-nine year lease on the Hambantota port and a controlling equity stake in it.

A ninety-nine year lease. The lease Minor Keith got on the Limón railroad in 1884 was also a ninety-nine year lease. The structural similarity is not an accident. Sri Lanka’s government chose a lease-and-equity deal because it needed foreign exchange, and China was in a position to offer it. That is precisely the position Minor Keith was in when he sailed to London in 1882.

Amazon, on the American side, has received something in the neighborhood of $4.25 billion in various local and state subsidies for its fulfillment and distribution centers. North Andover, Massachusetts, kicked in $27 million for a single warehouse. Glenwillow, Ohio, granted a distribution center property-tax reductions of nearly half for fifteen years. The subsidies are pitched as economic development. The research, when it is done, suggests that the counties that host Amazon fulfillment centers do not, in fact, see broad economic growth; they spend roughly $43k in public money per job created. Not an economic engine for the community. An enclave that ships value out.

On the resource monopoly side, consider cobalt. Cobalt is essential for the batteries in your phone, your laptop, your electric car. Roughly 70% of the world’s cobalt comes out of one country, the Democratic Republic of the Congo. The vast majority of the DRC’s cobalt production, in turn, is controlled by three foreign corporations, principally the Chinese firm CMOC and the Swiss firm Glencore. The refining capacity that turns raw cobalt into battery-grade material is dominated by China. The supply chain between a Congolese mine and your phone is deliberately fragmented across intermediaries, refiners, and manufacturers, in a way that obscures the origin of any given gram of cobalt and fragments the legal responsibility for how it was extracted.

That is the same structural logic as UFCO’s railroad network that skipped over local towns, just digitized. The supply chain is designed to make the extraction invisible.

At the production-enclave level, consider Foxconn’s plant in Zhengzhou, China, known informally as iPhone City. The facility covers 5.6 square kilometers, about a tenth of the size of Manhattan. At full capacity it employs around two hundred thousand workers, who live in company dormitories, eat in company cafeterias, and produce, by most estimates, about half of the world’s iPhones. Wages at Foxconn average in the range of $1 to $2 an hour including overtime. Fifty-hour weeks and twelve-hour shifts are typical. In peak production periods, fourteen-hour days, seven days a week, have been documented by workers and labor researchers. The company has, for more than a decade, been shadowed by a series of worker suicides. The structure, from orbit, is a company town. The same structure the United Fruit Company ran in 1920s Limón, with better electronics.

On the information control side, the mechanism is no longer a telegraph cable. It is an algorithm. Meta, the parent company of Facebook and Instagram, changed its algorithm in 2018 to deprioritize news content in favor of content from friends and family. Between 2021 and 2024, user engagement with news on Facebook dropped by roughly 78%. In 2023, in response to a Canadian law called the Online News Act, which required platforms to pay publishers for hosting their content, Meta pulled news entirely from its Canadian platforms and, as of late 2025, had not restored it. A corporation that operates the primary means by which a majority of citizens in several countries receive information about current events can, and does, unilaterally decide how much news its users will see.

The analogue, for Americans, is United Fruit’s Tropical Radio and Telegraph Company, which in the 1920s owned every wire by which news left Central America. Today the cables are digital, and the cables are owned by three or four corporations.

On the political capture side, the revolving door is now public record. The firm Palantir, which holds billion-dollar contracts with the Department of Defense and more than two hundred million dollars in ICE contracts, has systematically hired former senior officials from the Pentagon office responsible for integrating commercial technology into military operations.

Specific individuals are publicly traceable. Greg Little, one of the first officials to run the Pentagon’s Chief Data and AI Office, joined Palantir as a senior counselor in August 2023. David Spirk, who was the Pentagon’s chief data officer, joined Palantir as a senior counselor in July 2022.

Big Tech as a whole spent more than $100 million on federal lobbying in 2025, the first time the industry crossed that threshold, with Meta alone spending over $26 million. Amazon, Meta, Google, and Microsoft each gave a million dollars to the inaugural committee of the current administration.

Amazon Web Services, meanwhile, has for more than a decade been the primary cloud infrastructure provider to the United States federal government. It runs the classified infrastructure used by the Central Intelligence Agency.

In December 2022, Amazon was one of four companies awarded shares of the Pentagon’s Joint Warfighting Cloud Capability contract, worth up to nine billion dollars through 2028. The Pentagon, at this point, could not easily switch cloud providers. Doing so would require rebuilding, from the ground up, the computing layer of significant portions of the American military. That is not a relationship of vendor and client. That is, by the functional definition, infrastructure dependency.

John Foster Dulles, in 1954, felt the need to hide his United Fruit stock holdings. His modern counterparts file quarterly disclosure forms, and nobody has to overthrow a country for their interests to be served, because the interests are already, legally, part of how the state operates.

El pulpo did not die. It digitized.

What is on your counter

After a long article, it is worth coming back to where we started, which is the banana itself.

That banana is fifty cents, or sixty cents, or maybe seventy cents a pound. It is the cheapest piece of produce in the store. It has been the cheapest piece of produce in the store for most of your life, and for most of your parents’ lives, and for most of your grandparents’ lives.

It did not become cheap because the industry figured out how to produce it more efficiently. It became cheap because the industry figured out how to offload the cost onto people who could not fight back.

A railroad in a swamp paid for in four thousand lives. A civil war in Guatemala paid for with two hundred thousand lives. A labor force in Limón raised up and then legally abandoned. Fifteen hundred men in Costa Rica whose children were never born. A decade of payments to a death squad in Colombia. An industry-wide monoculture that is now watching a second extinction play out in real time because it could not stand to rethink the model.

That is what is on your counter. Not produce. A receipt. A blueprint.

And the receipt is not only for the banana. That is the point I’ve been working my way toward in this article.

The receipt, if you read it carefully, is for a pattern, and the pattern is everywhere, and the pattern did not stop when United Fruit became Chiquita or when Chiquita lost in federal court.

The pattern lives in Sri Lankan ports and Congolese mines and Chinese factory cities and American warehouses and Pentagon data centers. It lives in the boxes it is not visible from.

When you can see the pattern, you cannot stop seeing it. You start to notice, for instance, that the question “who owns the infrastructure between me and the thing I need” is not an abstract question in a policy paper. It is the question. It has been the question, in one form or another, since 1884 (and longer, really).

There is a useful way to hold all of this, and it is the thing I want you to walk out of this article carrying.

We are accustomed to thinking about power as something that wears a uniform and carries a flag. We think about elections. We think about armies. We think about the big, visible organs of the state. Those things are still real, and they still matter.

But what a banana company learned, and what every corporation that learned from it has since refined, is that the deepest form of power does not have to wear a uniform. It does not have to pass a law. It just has to own the road and the port and the wire.

When you own the road and the port and the wire, the government on paper continues to exist, and the sovereignty on paper continues to exist, and the citizenship on paper continues to exist, and all of those things become a little bit thinner than they were before. Not by much, at first. Over a century, by a lot.

The banana on your counter is not a metaphor for this. It is the oldest running example of it.

If the article ends with a question, which is the form I usually take with these, the question would be this: What kind of country, what kind of citizenship, what kind of life is actually available to a person when the critical infrastructure of their country, the logistics and the communication and the computing and the food, is owned by a handful of private entities whose legal obligations are to their shareholders, and not to them?

This has been the question for a hundred and fifty years.

It is the question Costa Rican’s asked themselves in the 1920s and the Guatemalan farmers asked themselves in 1953 and the West Indian workers in Limón asked themselves in 1935. The answer they got, each time, was written in someone else’s ledger.

The banana is on your counter, whatever you decide to do about it. The history is in the banana. So is the future, if you know where to look.

Share The Minuteman

If this piece helped you see something ordinary a little differently, send it to someone who has a banana on their counter and a suspicion that cheap things are rarely as cheap as they look.

The Minuteman grows when readers pass the frame along, one person at a time.

Sharing is caring :)