ASHFALL INSTITUTE | SUBDUCTION ZONE
Building Bridges — Episode Seven
TWENTY-FIVE OF THIRTY
P. A. Moore
Ashfall Institute | Subduction Zone
Written under the collaboration disclosed in Time Is a Wheel and We Are the Ground. Concept, argument and judgment: P. A. Moore. Research and composition: Claude Opus 5.
There are twenty-five million Americans under the age of thirty-five living in their parents' houses.
That is a record. It is higher than it was during the pandemic. It is close to a third of everyone in that age group.
And here is the number that tells you what it actually means.
Seventy percent of them have jobs.
They are not idle. They are not failing to launch, or refusing to grow up, or any of the other things said about them by people who bought a house on one income. They get up and go to work and come home to the room they slept in as children, because work does not buy a household any more.
That is not a young people problem. That is a country problem, and it is the subject of this episode.
Let me put the rest of it on the table, and I am going to be careful about where these numbers come from, because an argument like this one gets dismantled at its weakest citation.
Housing. The median American home reached about four hundred and thirty thousand dollars — up more than a third since 2019. Median asking rent, sixteen hundred and seventy-three dollars, up nearly eighteen percent over the same stretch. Wages did not do that.
Food. Grocery prices are up roughly a quarter since 2020. For an ordinary household that is somewhere between twelve and fourteen hundred dollars a year, gone, on the same cart of food.
Health. Insurers raised Affordable Care Act marketplace premiums by an average of twenty-six percent for 2026 — the steepest increase since the marketplaces opened. In states using the federal exchange, thirty percent. And that is before subsidies are settled: the Kaiser Family Foundation calculates that if enhanced premium tax credits lapse, what enrollees actually pay more than doubles — from an average of eight hundred and eighty-eight dollars to nineteen hundred and four.
Wealth. The top one percent of American households now hold 31.7 percent of all household wealth. That is the highest concentration recorded since the tracking began in 1989. The bottom half of the country holds 2.5 percent.
One percent of us hold more than the bottom ninety percent combined.
Homelessness. The last full national count recorded 771,480 people without a home on a single night — an eighteen percent jump in one year, the largest ever recorded. The following January’s count came in slightly lower, and I will note that honestly rather than leave it out, because I would rather you trust the rest.
That is the middle class being taken apart. Not in theory. In grocery bills and rent and the spare room.
And I want to be precise about the mechanism, because “the economy is bad” is not an argument and it does not tell anyone what to do.
A middle class is not a natural formation. It does not occur. It was built, deliberately, over about forty years, out of specific things: a minimum wage that moved, unions that could bargain, a mortgage anyone could get, a college education that did not require a lien on your future, and a tax structure that took proportionally more from the top than the bottom.
Every one of those was a policy. Every one of them could be reversed, and most of them have been.
We did not lose the middle class. We stopped paying for it.
Now the part that is harder to talk about, and I am going to talk about it anyway.
Some of what you are paying for instead is war.
Since last September the United States has been conducting strikes on boats in the Caribbean and the eastern Pacific. As of late August, the Associated Press counted at least sixty-eight strikes, on sixty-nine vessels, killing at least two hundred and twenty-seven people.
Every one of them is described as a suspected drug trafficker.
Suspected. Not charged. Not tried. Not convicted. There was no hearing, no evidence presented, no defense offered, and no court involved at any point between the suspicion and the missile.
The AP went to a coastal region of Venezuela where some of these boats had departed, and identified four of the dead. They were laborers and fishermen, paid about five hundred dollars a trip.
The Senate voted on a resolution that would have required Congress to authorize the strikes. It did not pass.
I have no brief for anyone who moves narcotics. I want to be very clear about that. What I object to is the word suspected doing the work that a courtroom is supposed to do — because a country that will kill a fisherman on suspicion has given up something it will find very difficult to get back, and because I have spent this entire series arguing that the difference between a decent arrangement and an indecent one is whether the people affected are allowed to see the terms.
Two hundred and twenty-seven people did not get to see the terms.
Then there is Iran, and I want to do the arithmetic in public.
The United States began military operations in Iran on the twenty-eighth of February. Congress did not authorize them.
The Pentagon told a closed-door congressional briefing that the first six days cost $11.3 billion. By day twelve, the Center for Strategic and International Studies put the running total at $16.5 billion.
Six days. Eleven point three billion dollars. That is roughly one point nine billion dollars a day.
And that is only the shooting. The National Priorities Project — a project of the Institute for Policy Studies — ran a deliberately narrow estimate covering just the operating costs of the major platforms in theater: two carrier strike groups, their naval support, and more than two hundred aircraft. No munitions. No troop deployments beyond those crewing the listed assets. No procurement.
Even counted that narrowly, it comes to $59.39 million a day.
Then they did the thing nobody in Washington does, which is say what the money would otherwise buy.
That $59.39 million a day would cover the daily cost of Medicaid for more than four million Americans. Or SNAP — food stamps — for more than nine and a half million.
Both of those programs were cut last year.
So when you are told there is no money for a jobs program, understand what is being said. There is money. It is being spent at a rate of nearly two billion dollars a day, on an operation Congress never voted for, while the food assistance is reduced.
And now I want to say something about what does not get counted.
The war has four separate organizations tracking its cost. The Pentagon briefed Congress on six days of spending. Somebody, somewhere, is maintaining a spreadsheet of bunker fuel prices per metric ton. The dollars are followed hour by hour, by people who are paid to follow them.
Nobody runs a counter on the other side of the ledger.
It has been measured, at least partly. Economists used the Medicaid expansion as a natural experiment — some states took it, some refused — and compared what happened to people who were otherwise alike. The finding, published in the Quarterly Journal of Economics: mortality among low-income adults fell about a tenth of a percentage point a year in the states that expanded. Roughly nineteen thousand people alive after four years who would not otherwise have been. And in the states that refused, something like fifteen thousand deaths that did not have to happen.
I will be honest about the state of that evidence. A randomized study in Oregon found no significant mortality effect, though it was almost certainly too small to detect one. These numbers are argued over. They arrive years late and they arrive contested.
For people without housing the finding is cruder and worse. They die twenty to thirty years younger than the housed.
But notice the difference in how the two columns are kept.
The cost of the strike is known before the missile lands. The cost of the premium increase is known, if ever, by an economist working with death certificates half a decade later.
One of those has a defendant. The other does not.
Nobody is sued over a woman who did not go in when the pain started, because there is no moment, no incident, no name on a form. It is not an event. It is an absence, distributed across a country, and absences do not generate paperwork.
We count what we can be blamed for. Everything else, we call the way things are.
That is not an accounting problem. It is a decision about what deserves a ledger — and it is the same decision, made the same way, that decided a fisherman could be killed on suspicion and a school could be blamed for a failure it did not cause.
Now here is where it arrives in your kitchen.
The war closed the Strait of Hormuz to practical traffic. About fifteen million barrels a day move through that channel — a fifth of the world’s oil. Tanker traffic all but stopped. Brent crude went above a hundred dollars.
Watch what happens next, because this is the part nobody explains.
Bunker fuel — the heavy oil that moves container ships — went from about five hundred and eighteen dollars a ton in February, the month before the war, to about nine hundred and one dollars a ton across March, April and May. An increase of seventy-four percent.
Every container on every ship got more expensive to move. And so:
Amazon imposed a three and a half percent fuel and logistics surcharge on third-party sellers in the United States and Canada, starting the seventeenth of April. UPS and FedEx raised their fuel surcharges.
And the consumer price index went from 2.4 percent year-over-year in February to 3.3 in March to 3.8 in April.
That is the whole chain, and it is four steps long. A decision made in a room in Washington. A closed strait. A fuel surcharge. Your groceries.
Nobody voted for the last three.
And tariffs.
I am going to use the most conservative number I could find, from a nonpartisan academic source, and I am going to tell you why.
The Budget Lab at Yale — which is independent and which revised its own estimate downward by three hundred and twenty-five dollars in July to bring it into line with the economic literature — puts the cost of current tariff policy at about eleven hundred dollars per household per year. The average statutory tariff rate stands around eleven percent. Over ten years, roughly one point nine trillion dollars.
Eleven hundred dollars a year, from an institution that corrected itself in the direction of a smaller number.
A tariff is a tax. It is collected at the border and paid at the register, and the reason it is politically attractive is precisely that most people do not experience it as a tax. It arrives as a price.
And it circles back. Among the reasons the insurers gave for that twenty-six percent premium increase: hospital costs, expensive new drugs, policy uncertainty — and potential tariffs on pharmaceuticals.
So the tariff raises the price of your groceries, and then raises the price of your insulin, and then raises the premium on the policy that was supposed to pay for the insulin.
Meanwhile the trading relationships that took eighty years to build are being spent. Those were not sentimental arrangements. They were the reason a container could cross an ocean on a handshake and a letter of credit, and they are considerably easier to burn than to rebuild.
The military budget for this year is about eight hundred and ninety-five billion dollars for the Department of Defense alone. Counted properly — nuclear weapons, veterans, homeland security — national security spending passes a trillion.
So let me put the two columns beside each other, because that is all this episode is.
A trillion dollars a year on security. Eleven hundred dollars a household in tariffs. Two hundred and twenty-seven people killed on suspicion. Premiums up twenty-six percent. Groceries up a quarter. Rent up eighteen percent. Twenty-five million adults in their childhood bedrooms. Seven hundred and seventy-one thousand people sleeping outside.
And in the other column, the thing we are told there is no money for: anything that would build.
I do not accept that. Not as economics and not as arithmetic.
Which brings me to the thing that was done before, by a country that was in far worse shape than this one.
Roosevelt was sworn in on the fourth of March, 1933. The Civilian Conservation Corps was authorized on the thirty-first of the same month.
Twenty-seven days.
By that summer, three hundred thousand young men were living in camps and working. Across its nine years, around three million passed through it.
There were no computers. No databases, no payroll systems, no way to verify anything about anybody except by asking. A quarter of the workforce was unemployed and the federal government had a fraction of the administrative capacity it has now.
And it stood up a national workforce of three hundred thousand people, housed and fed them, and put them to work, in under four months.
They planted three billion trees. They built more than eight hundred state parks, and bridges and trails and fire towers still standing. You have almost certainly walked on something they made and had no idea, because the work was good enough to read as landscape.
And the pay was thirty dollars a month, of which twenty-five went home.
Not suggested. Required. By allotment, every month, to the family.
Which tells you what the program was actually for. The parks were real. But the engine of it was income reaching households that had none, routed through the labor of their children in a way that let everybody keep their dignity. A man whose family eats because of him stands differently than a man in a relief line.
That is the piece left out whenever this program gets invoked now, because it is the piece that costs money and cannot be photographed.
And it was segregated, and I am not going to soften that, because a romanticized CCC is useless as a model.
After two years of inconsistent practice, the director, Robert Fechner, made complete camp segregation official policy in 1935.
Somewhere between two hundred thousand and a quarter of a million Black men served, across roughly a hundred and fifty segregated camps. Separate quarters. Separate mess halls. Supervised by white Army officers, and never promoted to officer rank, however long or however well they served.
Enrollment was capped near ten percent, pegged to each state’s Black population share in the 1930 census.
Look at what that cap did. Black Americans in 1933 were not unemployed at their share of the population. They were unemployed at far higher rates, having been fired first and hired last across the entire economy. A program for the unemployed that admits people by their share of the population rather than their share of the unemployment is exclusion performed with arithmetic — and it looks fair on paper, which is exactly why it lasted.
The law forbade it. An amendment from Oscar De Priest — the only Black member of Congress at the time — prohibited discrimination on account of race, color or creed in the CCC.
It was never enforced.
There was a separate Indian Division, about eighty-five thousand men working reservation land outside the camp system. And there were women’s camps, set up largely at Eleanor Roosevelt’s insistence and mocked in the press as the “she-she-she” camps.
Eight and a half thousand women. Three million men.
I raise all of this because the first question anyone will ask about a new program is who gets in, and if you have not answered it you will build the same machine and be surprised by the same result.
The CCC is the best evidence we have that this country can decide to build something and then build it. It is also proof that a program can be enormous, popular, effective, legally required to be fair, and discriminatory throughout, for nine years, in the open.
Both are true. The honest version is more useful than the heroic one, because it tells you what to watch.
So what would it do now?
Grid hardening. Undergrounding in fire country, vegetation management, substation work, reconductoring existing lines to carry more on the same towers. I spent a whole episode on why the grid was failing before anyone said the word “AI,” and almost all of the remedy is physical work that can be taught.
Water. There are still millions of lead service lines carrying drinking water into American houses. We largely know where they are. Replacing them is digging, cutting, fitting, backfilling. Not exotic. Just nobody’s job.
Wildfire mitigation. Fuel reduction, defensible space, prescribed burning under supervision, at a scale no fire agency is staffed to reach.
Retrofitting the housing stock. Insulation, air sealing, ventilation, heat pumps. Every one of those permanently lowers a household’s bill. It is the cheapest energy in the country and it is stranded, because it requires millions of small jobs on individual houses and there is no workforce to do them.
And one that ties back. In Episode Five I argued that the obstacle to better building materials is not risk but the absence of data about risk — that a novel wall assembly is uninsurable because it has no claims history. A corps building at scale, monitored, publishing performance data, generates exactly the actuarial record that does not exist. That is not make-work. That is the missing instrument.
Data centers, incidentally, need every one of these things too. But I have deliberately not made this episode about them, because they are a symptom of the arrangement and not the cause of it, and I would rather talk about the arrangement.
Let me put something beside the numbers, because six episodes of figures do not change anyone’s mind.
I wrote a story called The Encounter. It is on my own site, in the section called Shadow Realm, and anyone who wants it can go and read it for nothing.
There is a girl in it, on a farm. The family’s equipment is failing, and one machine in particular is old and patched and past economic sense, and her father has written it off.
She works other people’s land for weeks to buy the part. Her hands blister. Her back aches. She stitches her own palm when a splinter goes deep, and says nothing.
And when she lays the part on the table, her father looks at her like she’d brought home fire.
He doesn’t ask how. He just nods. They repair the machine together in silence.
Later in the book she leaves to serve. There is a shuttle and a door; her father nods again, her mother smiles tight and brave, and the old machine raises its arm in a slow wave. The door closes and the farm falls away.
“At least they still had Latch.”
I wrote that a while ago without this argument in mind. But it is this argument exactly. A young person goes away to serve. What she worked for stays behind with her family.
Twenty-five of thirty, sent home.
There is a line in that same book I would put on the wall of any new corps. Describing a society that has come out the other side of a catastrophe and rebuilt — every citizen serves, in medicine, engineering, exploration:
“It wasn’t law. It was promise.”
That distinction is the entire design problem.
Obligation produces compliance, and compliance produces a program people leave the moment they can and never mention again.
Promise produces the other thing — the thing the CCC actually produced, which is why men in their eighties were still wearing the pin in the 1990s and could still tell you which trail was theirs.
You get promise by making the offer real. Paid properly. Taught properly. Leading somewhere afterward. And open to everyone, as an enrollment rule with an audit attached — which is precisely what 1933 wrote down and never enforced.
The ask, and it is three lines.
One. A national service corps at a scale that matches the problem. The work above is measured in millions of person-years and essentially none of it is being done. The most recent federal attempt at anything like this was aimed at twenty thousand people, and it has been shut down. Twenty thousand, against three million in 1933, by a country with a hundred times the administrative capacity.
Two. Money that reaches households. The twenty-five of thirty was the mechanism, not the decoration. A service wage that supports a family — or lifts a person out of debt and into their own front door — is what turns a jobs program into an economic one.
Three. Enrollment fairness that is audited, not merely legislated. The De Priest amendment was law for nine years and meant nothing. Write the audit this time, and the penalty, and publish the numbers annually where anyone can read them.
I will finish with a fact and let it sit.
The last time this country was taken apart this thoroughly, it was rebuilt. Not by a committee and not by the market. By a government that decided the situation was intolerable and acted inside a month.
The man who did it was a Democrat, and every Republican president who followed him for forty years kept what he built, because it worked and because taking it apart would have been unthinkable.
I am not going to tell you what to do with that.
You will have noticed I have not had to.
Next time: Federal One, and the two thousand three hundred interviews nobody could have justified on a spreadsheet.
Sources — independent reporting and non-governmental analysis. Mortality and coverage: Miller, Johnson and Wherry, “Medicaid and Mortality: New Evidence from Linked Survey and Administrative Data,” Quarterly Journal of Economics, 2021; Sommers, Baicker and Epstein, New England Journal of Medicine, 2012; the Oregon Health Insurance Experiment null result on mortality, 2013; excess mortality among people experiencing homelessness as reported in the public health literature. Strikes on vessels: Associated Press count as of 25 August 2026, at least 68 strikes on 69 vessels and at least 227 killed; AP field reporting identifying four of the dead as laborers and fishermen paid roughly $500 per trip; NPR and Al Jazeera coverage; Senate war powers resolution failed. Iran and shipping: Brent crude above $100; approximately 15 million barrels daily through the Strait of Hormuz; bunker fuel $518/ton in February 2026 rising to $901/ton March–May, +74%; Amazon 3.5% fuel and logistics surcharge from 17 April 2026; UPS and FedEx fuel surcharge increases. Consumer prices: CPI 2.4% (Feb) → 3.3% (Mar) → 3.8% (Apr) year over year, as reported by CNBC. Tariffs: The Budget Lab at Yale, State of U.S. Tariffs, 24 August 2026 — approximately $1,100 per household annually, 11.0% average statutory rate, ~$1.9 trillion over ten years; estimate revised downward $325 in July 2026. Health insurance: Kaiser Family Foundation and Peterson-KFF Health System Tracker — 26% average marketplace premium increase for 2026, 30% on the federal exchange, and enrollee payments rising from $888 to $1,904 if enhanced tax credits lapse; Commonwealth Fund and Urban Institute analyzes. Wealth concentration and household formation as reported by Pew Research Center and national news coverage. CCC: enrollment, pay structure, trees and parks; Robert Fechner’s 1935 segregation order; approximately 150 segregated camps; the De Priest amendment; CCC Indian Division; women’s camps. Quotations from* The Encounter, *P. A. Moore, in Shadow Realm at the Ashfall Institute.
Building Bridges is a fourteen-part series. Previous: Episode Six — The Grid We’d Have to Build Anyway. Next: Episode Eight — The Last Decade It Was Possible.
P. A. Moore is the pen name of Pamela King, philosopher and artist. Available through the Ashfall Institute.