In the first three months of 2026, local opposition blocked or delayed at least seventy-five data center projects worth roughly $130 billion. That is not a trend line. That is the entire preceding year compressed into a quarter. Active opposition groups more than doubled, to 833, across forty-nine states, and more than three hundred bills landed in statehouses in the first six weeks alone.

Something moved. It is worth being precise about what.

New Jersey passed a law requiring facilities over fifty megawatts to commit to paying at least eighty-five percent of their projected power costs for a decade. Virginia added a levy of $0.011 per kilowatt-hour. New York passed a one-year moratorium on permits for anything above twenty megawatts. Texas ordered an audit of every data center connecting to its grid, which functioned as a pause on roughly eighteen hundred projects. These are not gestures. These are enforceable numbers attached to specific balance sheets, written into statute in a single legislative season, in states that agree on almost nothing else.

Now notice what they measure.

Water has a gauge. Somebody installed it, somebody reads it, and the reading is admissible. When a Virginia groundwater study concluded in July that new facilities would be unlikely to secure reliable withdrawal permits under current conditions, that conclusion had a unit behind it. Power has a meter, and more to the point, the meter produces a bill that arrives in an envelope at a residential address every thirty days. You do not have to persuade anyone that their electricity bill went up. They are holding it.

That is the whole story of why this moved so fast. Not virtue. Not that communities finally woke up to Big Tech. The two harms that were metered got legislated across forty-nine states in one session. The third harm did not.

The harm with no gauge

Everyone in these hearings knows the jobs arithmetic. It is not hidden and it is not disputed.

Virginia’s own Joint Legislative Audit and Review Commission found that a typical 250,000-square-foot facility employs about fifty full-time workers once it opens, roughly half of them contractors, after employing as many as fifteen hundred during a twelve-to-eighteen-month build. Outside Reno, a 1.1-million-square-foot campus projected seventy-three permanent jobs over a decade against more than four thousand construction positions. In Illinois, a Microsoft facility took more than thirty-eight million dollars in sales tax exemptions and produced twenty permanent jobs.

Communities say this out loud. They say it in the same meetings where they say the water thing and the electricity thing.

And in the last week of August the building trades began saying it back. Unions and construction trade groups have entered the fight from the other side, warning that halted approvals endanger thousands of building jobs and threatening to withhold support from candidates who oppose the projects. They arrived with a number: one electricians’ local reports its members worked twenty-eight million hours last year against fourteen million a decade earlier. That is a gauge, and it is behaving exactly as gauges behave. It is moving votes.

So the jobs argument did not go home. It came back metered.

But look at what that meter counts. Hours are throughput. They measure work passing through bodies, say nothing about what those bodies became, and read zero the day the build ends. The apprentice who topped out and the hand who swept the same floor for the same hours appear in that number identically.

The formation still has no gauge. It has just acquired a very convincing decoy.

Why? Because a construction boom is itself a kind of meter — a bad one, but a real one. Fifteen hundred hard hats on a site is a countable, photographable, ribbon-cuttable fact. Fifty full-time employees three years later is also countable. What is not countable, in any instrument a county commission can enter into the record, is the thing actually being traded: the formation those fifteen hundred people were acquiring, and the formation the next fifteen hundred will not acquire, because the work moved on and nothing accumulated locally except a slab and a substation.

There is a second, larger version of the same invisibility, and it is the one people are most anxious about even though they cannot say it cleanly. The concern is not only that the building employs fifty people. It is that what the building computes may quietly close the first rung of the ladder everywhere else. Stanford’s Digital Economy Lab found a sixteen percent relative decline in employment for workers aged twenty-two to twenty-five in the most AI-exposed occupations. Among software developers in that age band the decline runs closer to twenty percent, while employment for older developers at the same firms grew.

And here is where I have to be honest, because the honesty is the argument.

That finding is contested. David Deming at Harvard points out that the junior hiring decline appears to start about six months before ChatGPT was released, which is an awkward fact for anyone who wants a clean causal story; he suspects remote work. The Economic Innovation Group attributes much of the pattern to the sharpest monetary tightening cycle in four decades. Reasonable economists are genuinely unsure, and anyone who tells you otherwise is selling something.

I used to think that uncertainty was the weakness of the case. It is not. It is the mechanism.

The unmetered harm is always contestable. That is what unmetered means. Water was contestable too, right up until somebody put a gauge on the aquifer and published the number, at which point the contest ended and the statute got written. The absence of a meter does not indicate that the harm is small. It indicates that no one has yet built the instrument that would settle the argument — and in the meantime, every interested party gets to say “you can’t prove that,” and they are technically correct, and the plunder continues on schedule.

Two failures, not one

Two words compete here — “stranded asset” and Hernando de Soto’s “dead capital.” Both are right, about different things.

A stranded asset is one whose value cannot be captured — a $20 million extrusion line running at 30 percent of rated capacity. The asset is present, functional, and producing far below what it could. It also describes a man named Charley.

Charley worked at Saginaw. He could stand next to a piece of turning equipment, listen, and tell you it was going to fail before it failed. Nobody taught him that and nobody wrote it down. I know the capability was real because I spent a doctorate trying to extract it — taking fast Fourier transforms of the vibration off the spindle, hunting for the signature that ran ahead of a breakdown, so that a rule base could do what Charley did without Charley.

I built an instrument aimed at one man’s judgment, and what I learned is that the part of him that mattered would not come out. His job description did not contain that hearing. His pay grade did not reflect it. When he went, it went, and no line on any statement of ours moved by a dollar.

That is the stranded asset. He was not incapable. He was suppressed — a $20 million extrusion line running at 30 percent of rated capacity, except the line at least appears in the fixed asset register.

Dead capital is a different failure. De Soto’s subject was the house outside Lima: solidly built, lived in for twenty years, sheltering a family perfectly well — and worth nothing as capital, because there is no title. His central claim, the one that should interest anyone in finance, is that formalization changes the asset not at all. The house is the same house the day after the deed is issued. Only its visibility changes, and everything that can be built on top of it.

Human capability is both, and the order matters. No registry means no visible return on formation. No visible return means no investment. No investment means the capability was never built — and what was never built cannot be stranded.

So the recognition failure produces the formation failure. That is harsher than either term delivers alone, and it is why the argument cannot stop at better management. You can exhort a plant manager. You cannot exhort a balance sheet.

An asset class already made this crossing

Here is the part that should interest anyone who believes nothing can be done.

For four decades, operating leases sat off the balance sheet. Under the old standard, a company could commit itself to twenty years of payments on hundreds of stores and disclose it in a footnote. The obligation was real, enormous, and legally binding. It just was not on the books.

And every serious analyst knew it. Credit rating agencies and equity analysts spent those four decades manually capitalizing operating leases — building the shadow balance sheet by hand, company by company, because they needed comparability and could not get it from the financials as issued. It was tedious, inconsistent, and universal.

When the standard finally changed and right-of-use assets came onto the balance sheet, the underlying economics of those companies did not change by a dollar. Nothing about the business was different on the day of adoption. Only the visibility changed. And behavior changed with it.

Note who drove that. Not labor. Not legislators. Not a moral argument about the dignity of tenants. The demand came from the capital side, from people whose own self-interest required a number they could compare. The standard-setters ratified a practice that already existed in the market.

That is the instruction. The question is not how to make capital care about human formation. The question is: who is already doing this calculation off the books, badly, and would pay for a better instrument?

They are already paying for it. Blind.

The answer is sitting in plain sight in the acquisition standard.

When one company buys another, the assembled workforce — the fact that there are people there who already know how to do the work, who have been trained, who know where the failure modes are — is explicitly not recognized as a separate asset. It is subsumed into goodwill.

Read that again slowly, because it is a strange sentence. The standard does not say the assembled workforce has no value. It says the opposite. It concedes that value exists, is being purchased, and is being paid for in cash, and then it declines to name it. It goes into the residual bucket with everything else the buyer could not itemize.

So acquirers pay for formation. Every deal. In real money. They simply cannot tell a workforce that took twenty-five thousand hours to build from a workforce assembled last quarter through a staffing agency, because both arrive in the same undifferentiated line. A buyer who overpays for the second and a buyer who acquires the first for nothing both book it identically, and neither learns anything.

That is a live, recurring, high-stakes transaction in which sophisticated parties are systematically mispricing an asset they are already purchasing. It is exactly the condition that preceded the lease reform: a real obligation, a real value, real money moving, and no instrument.

The meter is not automatically on your side

I want to close on the objection I find hardest, because it is the one that decides whether any of this helps.

An asset on the books is an asset that can be impaired. Give a CFO a booked capability asset and you have also given her a clean, defensible, one-time write-off — and boards forgive a one-time charge far more readily than they forgive permanent operating expense. It is entirely possible to build a meter that makes displacement easier to execute and easier to explain than it was in the dark. Measurement is not inherently protective. Measurement is leverage, and leverage has a direction.

What sets the direction is not the existence of the instrument but what the instrument counts.

Measure a task inventory — knows this machine, this process, this line — and you have built a depreciating stock of the exact things automation is best at taking. Every advance impairs the asset. The write-off arrives on schedule, the productivity gain is banked as headcount reduction, and the meter you built to protect people becomes the document that authorizes removing them.

Measure formation depth — the demonstrated capacity to take command of a process you have never met before — and the arithmetic inverts. New capability makes that asset worth more, not less, because a better machine in formed hands returns more than a better machine in unformed ones. The firm’s return on carrying value goes upwhen the technology improves, which means the rational response to an automation advance is to form more people to absorb it, not fewer.

That is not an aspiration. It is Amberg. The most automated electronics plant in Europe, holding one of the most deeply formed workforces in Europe, output multiplied many times over across three decades on a workforce that stayed roughly constant in size and rose steadily in capability. That combination reads as a paradox only if you were measuring the wrong thing.

So I will not tell you that a meter fixes this. I will tell you that the absence of one guarantees the current outcome, that the last asset class to make this crossing did so because the people writing the checks got tired of guessing, and that the design of the instrument — not its existence — is the whole fight.

The water got a gauge. The power got a meter. The question in front of us is what we are going to count next, and whether we are going to count the right thing.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com.

Written with AI assistance. The argument, the judgments, and the floor testimony are the author’s own.

Dear Andy,

This letter has two names on it. The second one spent a career building the machines.

Over the last two days I made an argument to you in two parts: that Amazon holds a fifth pillar it has never named, and that you have already built the instrument to fund it and pointed it at the door.

Both letters left one question standing, and it is the question anyone inside your building would ask first.

Why form people for work the machines are about to take?

It is not a cheap objection. It is the strongest one available, and if the answer is wrong then everything in the first two letters is sentiment. So I am not going to answer it myself.

Dr. Roland Menassa is a co-founder of the Capability Capital Institute. Before that he was Principal and Senior Manager for Advanced Technologies at Amazon, where he led the development and deployment of advanced robotics across the fulfillment network. Before that, the GE Advanced Manufacturing Technology Center. Before that, twenty-four years at General Motors, ending as a GM Fellow in flexible manufacturing systems. He was elected to the National Academy of Engineering in 2022 — the citation reads

for contributions to robotics and automation in aerospace, automotive, and e-commerce. He holds over a hundred patent filings.

He has spent thirty years removing human labour from buildings like yours. I asked him whether the people who remain matter less. He said the opposite, and then he told me about a night shift.

What the robots actually did

Workers once walked miles a day through endless aisles of inventory. Driven by a desire to improve both fulfillment speed and employee safety, Amazon introduced automated robotic storage fields: instead of workers traveling to products, products were brought to workers at ergonomic stations. Projection systems illuminated the correct bins. Vision-based scanning identified barcodes in motion. Fulfillment times shrank from days to hours, and what remained for the employee was final product selection, verification, packaging, and shipment.

Read that last sentence slowly, because it is the whole letter in miniature. The walking went away. The verifying did not. One of those is a task and the other is a judgment, and only one of them was ever going to be absorbed.

The night I walked ten miles

My experience at GM placed me inside one of the highest-volume automotive production environments in the world — a three-shift body shop, roughly 425 robots across 750,000 square feet, welding sheet metal into a complete vehicle body before paint. My first assignment was to lead a third-shift skilled-trades team of forty, maintaining quality and throughput while managing downtime across hundreds of automated stations.

On my first night, I spent the entire shift walking from one issue to the next — robot faults, sensor failures, code execution errors, quality-sensor anomalies. By the end of the shift, I had walked more than ten miles responding to one automation interruption after another. After several months of this routine, two conclusions became obvious. First, you become very fit. Second, and more importantly: the more automation, sensors, programming, and complexity you add, the higher the downtime and the greater the maintenance burden required to keep the line running.

Factories that attempted lights-out operations learned this faster. Fully automated, unattended manufacturing does not work in practice: skilled trades cannot diagnose, repair, or maintain in the dark. The very people required to keep the system alive are removed the moment the lights go off.

Automation is not a substitute for people. It is a system that depends on them. This is why the future of manufacturing is not lights-out. It is lights-on, people-in, automation-with.

Andy, hold those two accounts next to each other. In the first, a human being walks ten miles a day through your aisles because there is no automation. In the second, an engineer walks ten miles in one night because there is. Same distance, same tired body at the end of the shift. The machines did not remove the walking. They changed who was doing it and what it cost to lose him.

The capability that got scarcer, not cheaper

Robots themselves are flexible and can be taught virtually any path. The ecosystem that enables them — end-of-arm tooling, conveyors, fixtures, nests, pallets — is not. These supporting systems often cost up to ten times more than the robot, and tooling programs can stretch two years before a single unit is produced.

Several lines I worked on were locked to specific families of street-lighting fixtures rather than to actual market need. We redesigned the products so key locating features were commonized, then deployed collaborative automation only at the bottleneck stations. On the precision urethane-dispensing operation, the robot performed the high-accuracy dispensing and handed the glass directly to the operator, who installed it and confirmed alignment and no squeeze-out.

The result was a meaningful increase in throughput using the same people, the same footprint, the same process. These robots empowered workers to interact with them directly — repositioning the robot to optimize their station, guiding the arm to refine its path. Robots take precision, repeatability, non-value-added time. People take cognitive work, problem-solving, flow.

That is Agility — the convergence of human capability and robotic precision to create manufacturing systems that adapt as quickly as the market demands.

Note what the operator in that cell is doing. She is not tending the machine. She is judging the machine’s output, correcting the machine’s position, and teaching the machine a better path. Three capabilities, none of which existed in the job before the robot arrived, none of which can be picked up in a ninety-day ramp.

What follows from it

Andy, if Roland is right, the objection inverts.

Automation does not remove the need to form people. It removes the need for unformed people — a different sentence, pointing the opposite direction. Every task a machine absorbs is a task that required no discretion. What remains is the residue: exception, anomaly, the moment the system is confidently wrong and someone has to know it. That is not a smaller job. It is a harder one, and it is the only kind that will exist on your floor in fifteen years.

You cannot staff that with turnover. You cannot hire it in ninety days. It is precisely the population your current model is built not to produce.

Roland gave me one sentence I have not been able to put down, and I want to hand it to you as a capital allocation statement rather than an engineering one.

When complexity scales faster than human maintainability, uptime collapses.

He then took his own sentence apart on your behalf, which is why I trust it.

Amazon has so much redundancy in its fulfillment base that it is unlike anything I have experienced. If one fulfillment center disappears from the map, they fulfill from another. They have created a behemoth of a system that is independent of maintainability.

He is right, and it is the strongest defence of your position available. You did not solve maintainability. You bought your way around it, at network scale, and it works.

But look at what it cost and where it stops. Redundancy is capital held against a problem you chose not to fix. It is the most expensive form of insurance there is, and it does nothing for the quality of the work inside any one building — it only guarantees that another building can cover for it.

And then Roland gave me the natural experiment, inside your own company.

Amazon’s technicians and skilled trades were all contractors at each site, unlike their direct-hire workers. Never made sense. I visited their Japan facilities and the reverse is true. Their skilled trades and technicians were direct hires, and you can see the smooth operation and the advancements they made internally, far better than their US plants.

Same company. Same robots. Same software. One difference: in the United States the people who keep the machines alive are somebody else’s employees, and in Japan they are yours.

The smoother operation is the uptime argument. The internal advancement is mine. Both showed up in the same building, on the same side of the same experiment, and neither of us designed it.

You are scaling complexity at a hundred and thirty-one billion dollars a year, rising toward two hundred. Human maintainability is the denominator in that sentence. There is no line on your books that funds it. There is one that funds departure.

Which means the four moves in yesterday’s letter get more urgent as the capital expenditure climbs, not less. That capex is an enormous bet that the machines will be good. It is also, whether or not it was intended this way, an enormous bet that the remaining humans will be excellent — and you have no instrument that produces excellence.

One more thing falls out of this, and it answers the sharpest version of the objection.

Confirmed capability is portable. That is not a concession to the worker; it is the property that makes the mechanism robust to exactly the future you are building. If the job changes — and it will — a confirmed standard travels with the person into the next configuration of the work. Equipment depreciates against change. People, formed properly, do not.

You are spending two hundred billion dollars a year on the assumption that the work will be unrecognizable in a decade. We agree. That is our argument, not our objection to yours.

The close

Between us we have sixty years on plant floors and one career spent automating them. We do not agree about everything, and the disagreement is worth printing because it is the interesting part.

Roland’s answer to why you should form people under automation is that the system does not run without them. Formation is a reliability requirement: remove the formed human and downtime climbs, quality drifts, and the dark factory stays dark for the wrong reason. He can make that argument from maintenance data, and it requires you to believe nothing about what a worker is owed.

Mine goes further than his. I think the formed human is owed the formation — that capability confirmed on your floor was paid for by work already performed, and that an instrument which funds departure while never funding development has the accounting backwards. Roland makes his case from uptime. I am making mine from a balance sheet that does not yet exist.

We have not settled it between us and we are not going to settle it here. Both roads arrive at the same building.

We agree about this: the machine and the master are not substitutes. One is issued to the other.

The guilds never ran at your scale, and we cannot prove from history that confirmation works across a million people. Nobody can. What we can say is that nothing else has ever produced the person you are about to need, and that you are currently the best-capitalized organization in human history to find out.

You have named Marketplace, Prime, AWS and silicon.

We think there is a fifth, and we think it gets more valuable every time you buy a robot.

What would you call it?

Respectfully,

Venki Padmanabhan

Roland Menassa

Capability Capital Institute

Sources

Kristin Stoller, “No. 1 on the Fortune Global 500: Amazon’s Jeff Bezos on how his garage startup became the largest company in the world by revenue,” Fortune, August/September 2026 issue (published online July 28, 2026). Capital expenditure figures — $131 billion in 2025, an estimated $200 billion in 2026 — are as reported in this profile.

Dr. Menassa’s account of robotics deployment at Amazon, General Motors and GE is original testimony given for this essay.

National Academy of Engineering, Class of 2023 election citation.

Parts one and two of this correspondence: The Fifth Pillar and The Account You Already Built.

Dr. Roland J. Menassa is a co-founder of the Capability Capital Institute and VP of Manufacturing Technology at Champion Homes. He was elected to the National Academy of Engineering in 2022 for contributions to robotics and automation in aerospace, automotive, and e-commerce. He led advanced robotics development across Amazon’s fulfillment enterprise, ran GE’s Advanced Manufacturing Technology Center, and spent twenty-four years at General Motors, ending as a GM Fellow in flexible manufacturing systems. He holds more than a hundred patent filings and a PhD in mechanical engineering from Rensselaer.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com. He has spent thirty-six years in manufacturing leadership across General Motors, Chrysler, Mercedes-Benz, Royal Enfield, Ather Energy, and Cyient DLM, and currently manages a plant for Advanced Drainage Systems in Wooster, Ohio. He holds a PhD in industrial engineering from the University of Pittsburgh.

Written with AI assistance. The argument, the judgments, and the floor testimony are the authors’ own.

Dear Andy,

Yesterday I argued that Amazon holds a fifth pillar it has never named: that innovation and brand equity, the only two assets you own that cannot be bought, are produced by formed people — and that the strand producing them has been harvested for thirty years without replenishment.

Diagnosis is cheap. Here is the mechanism.

You already built it

You have already built the instrument. Career Choice pays the school directly, before the course begins. No reimbursement paperwork. No lifetime cap. Ninety days to eligibility. And — the remarkable part — no service obligation. An associate can finish the degree and leave the next morning and owes you nothing. Over a billion dollars committed. Hundreds of thousands of people through it.

That is a capability account: prepaid, portable, unencumbered by clawback. Nobody else on the list has built one. You did it fourteen years ago and did it well.

And then you pointed it at the door.

The program’s own language is that it prepares people for career success at Amazon or elsewhere. Read that sentence as an operator. It is a concession, printed in a benefits brochure, that the surest way for a person to grow is to stop doing this work. Career Choice is Amazon’s own written admission that the fulfillment center is terminal.

I am not asking you to end it. I am asking you to turn it around.

The accounting is the argument

Under ASC 805, when you acquire a company you recognize the intangibles that came with it. Under ASC 842, when you lease a building you capitalize the right of use. The shell, the racking, the conveyor, the robotics, the sortation equipment — all capitalized, depreciated, carried.

The person who makes all of it productive is an operating expense. Recognized in the period consumed. No basis. No carrying value. No return calculation. When she leaves, nothing is written off, because nothing was ever written on.

Your books record one strand of the Twin Helix and not the other. Every asset on your balance sheet is somebody’s past effort, frozen and counted. The effort still in motion is invisible.

And you are already paying for this.

Turnover in fulfillment is what it is; you know the number better than I do. Every exit is a replacement requisition, a background check, an onboarding cycle, a productivity ramp, and a stretch of elevated error and injury exposure while the new person learns what the last one knew. That money leaves your building every year, already in the P&L — leaving as replacement cost instead of arriving as a deposit.

One caveat, stated plainly, because your CFO will find it in the first reading if I don’t. The wage steps in item two are real incremental spend in the year they are earned. The deposits and the confirmation infrastructure are not — those are funded out of attrition you stop paying for. Whether the wage steps come out net-neutral depends on how far retention moves, and I will not insult you by claiming to know that from Ohio.

What I will say is that this is the ordinary shape of a capital decision: identifiable outlay, modeled return, testable payback. Your finance organization runs exactly this every time you buy a sorter. The only unusual feature is the asset class.

Four moves

One. Turn the account inward. Keep Career Choice exactly as it is — it does good and should continue. Add a second rail funding mastery of the work itself: mechatronics, flow, quality systems, safety engineering, frontline leadership. Same prepaid terms, same portability, same absence of clawback. Opposite vector.

Two. Three years, guaranteed — and paid as it is proved. Not a single tuition draw but a three-year runway with continuous deposits, so a person who stays can see the shape of what they are becoming. Formation is not an event; it has a duration, and below a certain duration it does not take.

But the deposit alone is not enough. A deposit into someone’s capability that never appears in their compensation is training — and training is something you already do. It asks the associate to accumulate an asset while the entire return accrues to you. People read that arrangement correctly, whatever the brochure says, and they respond by leaving.

So each confirmed step must move the wage. Not negotiated, and not automatic with tenure. Released by evidence. The capability is demonstrated, a master confirms it, the pay grade steps, and the record travels with the person whether they stay or go.

That is the difference between a benefit and a ladder. A benefit is something you give. A ladder is something a person climbs, and the rungs have prices on them.

Three. Confirm through the frontline leader. You already have this layer and you are carrying it as supervisory overhead.

I hired Dennis Boutwell as my first supervisor at Lansing Delta Township. He is now Director of Manufacturing Excellence at the company I work for. What made Dennis worth hiring at twenty-nine was not a credential — it was that he could walk a line and tell you, correctly, which three operators were about to become something and which one was quietly drowning. Nothing in our system asked him for that judgment. Nothing recorded it. We were lucky he offered it anyway.

Your area managers and senior associates have that same judgment right now, on every shift. Credential them. Give them formal standing to confirm progression and let that confirmation release the deposit. Every guild that ever functioned ran on a master who could say this one is ready and be believed.

Four. Work the disclosure. Get your controllers and auditors into a room on whether a defined, confirmable formation program can be recognized as something other than period expense. I know how hard that is. I also know the first company on the Global 500 to move human formation above the line will have done something more consequential than any of the four pillars.

The word someone will use

Let me say plainly what someone in your building will say when this circulates. He’s a labor guy. This is a union by another name.

I have asked you to raise pay, so let me be precise about the mechanism, because the mechanism is the whole distinction. A union raises the price of labor by negotiation — the same increase for everyone in the unit, won at the table, independent of any individual’s capability. What I have described raises pay by confirmation. The standard is published. The associate meets it or does not. A master confirms it. The wage steps for that person and no one else.

That is pay for demonstrated performance against a rigorous bar — word for word, the compensation philosophy Amazon already applies to everyone holding a white badge. I am asking you to extend your own stated principle to the people it was never written to cover.

The institution I am describing is the guild — older than the union, and built to serve the employer at least as much as the worker. Guilds did not bargain. They set the standard of the craft, gated entry, ranked members by proven capability, and guaranteed the buyer that the work would be right. That is a quality system with a human face, and it carried European manufacturing for six hundred years.

Nothing here moves value from your shareholders to your associates. There is no transfer — only a recognition, and a wage that follows the asset because the asset got larger.

The garage

Your founder ends the profile with a line I agree with completely: “all civilizational wealth is driven by invention.”

But invention is crystallized formation. Nobody invents from nothing. There was a formed person in that Bellevue garage in 1995 — formed by a physics education, by a trading floor, by a hundred people who taught him things and were never counted as an input. He kept the door-desk, moved it into his study, and sits at it for photographs thirty years later because he knows exactly what happened at it.

The garage was a formation site.

A fulfillment center is a garage for nobody. Not because the people in it are lesser, but because nothing that happens to them there is designed to accumulate. They arrive able and they leave able, and everything in between is recorded nowhere except your turnover statistics.

I may have the mechanism wrong. Three years may be the wrong runway. The confirmation layer may not survive contact with a million people — the guilds never ran at that scale, and I cannot prove it scales, only that nothing else has ever produced what you need it to produce. If someone inside your building has walked a floor and can show me where this breaks, I would rather know than be right.

What I don’t think is arguable is that the pillar is there, holding up the other four, unnamed and uncounted and therefore free.

You have named Marketplace, Prime, AWS, and silicon.

What would you call the fifth?

Respectfully,

Venki Padmanabhan

Co-founder, Capability Capital Institute

The Long Game for All

Sources

Kristin Stoller, “No. 1 on the Fortune Global 500: Amazon’s Jeff Bezos on how his garage startup became the largest company in the world by revenue,” Fortune, August/September 2026 issue (published online July 28, 2026). The Bezos quotation on invention, the door-desk detail, and the characterization of Amazon’s labor position are drawn from this profile.

Amazon Career Choice program terms — direct prepayment to institutions, 90-day eligibility, no lifetime cap, no post-completion service obligation, and the stated purpose of preparing employees for careers “at Amazon or elsewhere” — as published by Amazon and its partner institutions.

Amazon’s $1 billion Career Choice commitment under the Future Ready 2030 fund (announced October 2025), and the earlier $1.2 billion Upskilling 2025 pledge.

ASC 805 (Business Combinations) and ASC 842 (Leases) are cited as the accounting treatments that capitalize acquired intangibles and right-of-use assets respectively. The proposal to recognize confirmable formation spend as something other than period expense is argued here as a proposal, not asserted as settled accounting treatment.

Guild structure and its wage-by-confirmation mechanics are drawn from the European craft-guild tradition; the claim that such structures have never operated at million-person scale is stated in the essay as an open limitation.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com.

Written with AI assistance. The argument, the judgments, and the floor testimony are the author’s own.

Dear Andy,

Congratulations on the No. 1 spot — past $700 billion, and the end of Walmart’s twelve-year hold on it. That is a real achievement and I don’t intend a word of what follows as a diminishment of it.

Let me deal with the obvious thing first. The largest company I ever ran built about a hundred and thirteen thousand motorcycles a year. You ship that many packages in the time it takes me to walk my plant floor. On scale there is no comparison and I won’t pretend there is.

But I am not writing to you from scale. I am writing from ten feet.

Thirty-six years on plant floors: GM at Lansing Grand River, Lansing Delta Township and Buick City; Chrysler and Mercedes-Benz in Stuttgart; Royal Enfield in Chennai as COO and then CEO; Ather; Cyient DLM. Today I manage a plant in Wooster, Ohio. What all of it has in common is proximity — close enough to watch individual people become capable, or fail to, and to know their names while it happened.

That is not a smaller version of your vantage point. It is a different one, and it is the single thing an organization your size cannot generate on demand, because formation is invisible above a certain altitude. You have to be standing near it.

I have also signed reductions in force and run the P&L that made them necessary. I am not arriving with a grievance, but as an operator who thinks you are carrying an asset you have never counted.

The pillar he didn’t name

In the Fortune profile, your founder lists what holds Amazon up. Marketplace. Prime. AWS. And the silicon business, he says, is lining up to become the next one. Four pillars, and every one of them is a product or a platform.

Then, later in the same conversation, he is asked what actually accounts for the company’s success — and he doesn’t name a single one of them. He points to the leadership principles. To a culture that is good at inventing and good at obsessing over customers.

The thing your founder credits for building the largest company in the world by revenue does not appear on his own list of what the company stands on.

The leadership principles are not a product. They are a formation system — a doctrine for how a person is shaped until they carry an institution’s judgment into rooms where no one is watching. Amazon has one of the most explicit formation cultures in American business: a confirming master in the bar raiser, an apprenticeship artifact in the six-pager, a lineage, an initiation, a standard.

It is a guild. You have simply never called it one, and never put it on the balance sheet.

Two strands

Read the profile once more and you will find it written in two different languages. One for the corporate ranks, entirely developmental — unreasonably high standards, a founder relentless about increasing the capacity of the organization, a company people describe as hard to leave. And one for the warehouse: speed, efficiency, worker tracking, automation, workers organizing.

Increase the capacity of the organization is a phrase Ohno would have recognized on sight.

Two vocabularies, two populations, one badge color between them. That is what it looks like when a company runs one strand of a double helix and lets the other go untended.

I call that structure the Twin Helix, and the axiom underneath all of my work is this: capital is crystallized labor; labor is capital in formation. Two strands, wound together, each becoming the other over time. Neither is prior.

The capital strand — EGIB — carries Earnings, Growth, Innovation, Brand equity.

The labor strand — TLHW — carries Time, Love, Health, Wealth.

Capital’s four are audited. Labor’s four are lived. The claim of the helix is that they convert into one another continuously.

You went to the top of the Global 500 on E and G — earnings and growth at a scale no company has reached — by pointing your labor, blue badge and white badge alike, squarely into I and B: the invention culture, the unreasonably high standard, the two-day promise that became one day that became the same afternoon.

Here is the part that should interest you commercially. Innovation and brand equity are the only two assets you hold that cannot be bought.

Test it. You acquired Whole Foods for $13.7 billion and MGM for $8.5 billion and received four hundred stores and a four-thousand-film library — real assets, fairly priced. Neither transaction transferred a single unit of I or B. Now run it the other way: Walmart has more square footage, comparable capital, and sixty years more retail experience, and could not construct what you constructed. Every input you had was available to them.

The difference was never capital. It was formed people, and nothing else produces I and B.

The four cells

Everything Amazon is famous for therefore sits on the labor strand, drawn out of it for thirty years. And the strand was never sustained. It was harvested. Take the four in turn.

Time. Amazon’s entire proposition is the return of time to the customer. That is the product. Two days, then one, then hours. And the mechanism that produces it is the metering of time on the floor — rate, path, time off task. The company that sells time to a hundred million households accounts for it most severely in the people who generate it. That is not hypocrisy. It is a conversion, running in one direction only.

Love. I use the word deliberately and mean something an operator can act on: the willingness to do the work correctly when no one is watching, and to care what happens to the person next to you. No metric produces it and no incentive buys it, and it is the sole source of the discretionary effort that separates a plant that makes its numbers from one that doesn’t. Your own former executives describe exactly this when they call Amazon hard to leave — which is not a complaint but the definition of a gharana. Nobody finds a jobhard to leave. People find it hard to leave what they love.

Health. This is where your safety investment belongs, and where it stops short. Billions spent so people are not injured is real money doing real good, and it is still only the harm-avoidance half of the cell. A person can finish thirty years uninjured and finish them depleted. Injuries avoided is a floor. Health sustained is an asset.

Wealth. Not wages — wages are consumed in the period earned, the same way you expense the labor that produced them. The symmetry is exact and it is the whole problem. Wealth is what accumulates and can be carried away.

Let me put a name on that.

Wally Vinton worked the trim shop at Lansing Grand River, wiring harnesses, for thirty years. Wally could hear a harness seating wrong. Not see it — hear it, from two stations away, over the line. He could not have written down how he did that if I had given him a month and a desk. Nobody taught it to him and nobody wrote it down, and when Wally retired that hearing left the building and no line on any statement of ours moved by a dollar.

Wally had income for thirty years. He never had wealth — no credential, no confirmed standing, no transferable claim on the one thing he could do that nobody else in the plant could do.

You have tens of thousands of Wallys. They built your brand equity and hold none of their own.

What I want to ask you

You can run this way for a long time — Amazon ran this way to the top of the world, and that is the strongest argument against everything above. It plainly worked. But the two events tucked into the middle of your Fortune profile, warehouse workers organizing and thirty thousand corporate roles cut in a single year, look to me like one story rather than two: a strand drawn on and not replenished, showing at both ends of the badge spectrum in the same year.

So here is the question I would put to you, and I mean it as a question rather than a device.

Earnings and growth are secured in the present by innovation and brand equity. What secures innovation and brand equity in 2036?

If the answer is capital, you already spend $200 billion a year and the answer is easy. If the answer is people, then the four cells above are not a moral argument. They are a maintenance schedule on your only unbuyable asset, and you do not currently have one.

Tomorrow I will make the case that you have already built the instrument for this — funded, running today, and pointed the wrong way.

But I would rather hear where the argument breaks. If you have a mechanism that sustains T, L, H and W at a million-person scale and I have simply missed it, I will publish the correction myself.

Respectfully,

Venki Padmanabhan

Co-founder, Capability Capital Institute

The Long Game for All

Part two tomorrow: the account Amazon already built, and the direction it points.

Sources

Kristin Stoller, “No. 1 on the Fortune Global 500: Amazon’s Jeff Bezos on how his garage startup became the largest company in the world by revenue,” Fortune, August/September 2026 issue (published online July 28, 2026). All quotations from and characterizations of Bezos, Jeff Wilke, and Bill Carr are drawn from this profile.

Fortune Global 500, 2026 edition — Amazon at No. 1, revenue above $700 billion, ending Walmart’s twelve-year hold on the top spot.

Whole Foods acquisition ($13.7B, 2017) and MGM acquisition ($8.5B, 2022) figures as reported in the same profile.

The “unreasonably high standards” characterization originates in The New York Times, “Inside Amazon: Wrestling Big Ideas in a Bruising Workplace” (August 16, 2015), as cited in the Fortune profile. Amazon disputed the article’s framing at the time.

Warehouse organizing activity, Amazon’s safety investment response, and the reported reduction of approximately 30,000 corporate roles are as characterized in the Fortune profile.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com.

Written with AI assistance. The argument, the judgments, and the floor testimony are the author’s own.

During the CTS launch at Lansing Grand River we had a defect that would not go away. The instrument panel rubbed the door opening as it went in. Every shift, on some fraction of the cars, a scuff where no scuff should be.

The easy move on any assembly floor is to blame the operator. The panel is heavy and awkward, the aperture is tight, and somebody is not being careful enough. That explanation is always available and it is almost always wrong.

Ramon Hernandez was my team leader on instrument-panel install, and he did the opposite. He took the fault onto himself and his team, which is not a small thing to do in a plant during a launch, when the whole building is looking for somewhere to put the blame. Then he gathered data. Shift after shift, car after car, until he could trace the scuff back to its origin: a tool mechanism on his own team’s station.

He found the failure point before it became a warranty claim.

Sit with the sequence, because every part of it matters. He noticed. He accepted responsibility before he had established it. He measured rather than argued. He followed it upstream to his own equipment rather than downstream to somebody’s hands. And he did it during a launch, under time pressure, with nobody instructing him to.

No system handed him that. There was no procedure titled trace the chronic defect to its root and absorb the blame while you do it. He had been formed to look — over years, on a floor that had taught him what looking was for.

Now hold that man in your mind and read the rest of this essay, because what follows is the story of an industry that spent four decades converting people like Ramon into a cost line, and then discovered what they had been carrying.

The Welch doctrine

When Jack Welch became CEO of General Electric in 1981, GE made jet engines, locomotives, appliances, and medical equipment. When he left in 2001, GE was effectively a financial company that still happened to make some things.

Welch pioneered what became the dominant model in American manufacturing: relentless cost-cutting, obsession with quarterly earnings, massive stock buybacks, and the treatment of employees as costs rather than assets. His rank-and-yank system — grading people on a forced curve and terminating the bottom 15 percent annually — became the template for a generation of chief executives.

Under Welch the stock soared. Fortune named him Manager of the Century. Every major manufacturing CEO in America tried to copy him.

Then GE collapsed. The financial engineering that inflated the numbers for two decades left the company hollow. By 2024 it had been broken into three separate companies, having destroyed more shareholder value in its decline than it created in its rise.

The Manager of the Century produced a company that could not survive two decades after he left it.

Boeing

Boeing is the Welch doctrine applied to an industry where the product must not fail.

In 1997 Boeing acquired McDonnell Douglas, a company known for aggressive cost-cutting — a merger often described as McDonnell Douglas acquiring Boeing with Boeing’s money. Former McDonnell executive Harry Stonecipher became CEO and announced the intent openly: when people say I changed the culture of Boeing, that was the intent, so that it is run like a business rather than a great engineering firm.

Between 1998 and 2018, Boeing spent $61 billion on stock buybacks — 81.8 percent of its profits. Add dividends and shareholders received 121 percent of profits. The company returned more than it earned.

Between 2013 and 2018 alone it spent $41.5 billion on buybacks. CEO Dennis Muilenburg personally received at least $106 million from 2011 to 2018, primarily in stock. Capital expenditure fell below 2 percent of sales — half what Airbus typically spent.

Boeing could have designed an entirely new airplane with fully modern technology for roughly $7 billion: about what it spent on buybacks in a single year of that period. Instead it chose to re-engineer the aging 737, bolt on larger engines, and add software to compensate for the aerodynamic consequences.

They called it the 737 MAX.

346

On 29 October 2018, Lion Air Flight 610 went into the Java Sea, killing all 189 aboard.

On 10 March 2019, Ethiopian Airlines Flight 302 crashed minutes after takeoff, killing all 157.

Both were 737 MAX aircraft. Both crashed because of MCAS, a software system designed to compensate for instability created by larger engines fitted to a 1960s airframe. The system was a workaround that avoided the cost of designing a new airplane.

Boeing spent $41.5 billion on buybacks in the five years before those crashes. The safety work experts say was necessary would have cost under $7 billion.

In January 2024 a door plug blew out of a 737 MAX 9 in flight. The FAA investigation found systemic quality failures across Boeing’s manufacturing processes.

The industry analyst Richard Aboulafia summarised the culture in one line: crush the workers, share price, share price, share price — financial moves and metrics first, a ruthless effort to cut costs with no realisation of what it could do to capabilities.

Harvard Business School’s Bill George concluded that decisions taken in the name of shareholder value over two decades cost Boeing’s investors $87 billion since 2018. The extraction destroyed more value than it captured.

GM

General Motors ran the same play in a lower key.

Between 2015 and 2023 it spent over $30 billion on stock buybacks. In the same period it closed multiple plants and laid off thousands of people, including the 2019 closure of the Lordstown assembly plant that took an Ohio town’s economy with it.

I watched an earlier version of that closure. Buick City shut in 1999, and the accountants totalled the presses, the conveyors, the tooling — every one of them had a value and a buyer. What the people in that building knew had no line at all. It walked out through the gate inside them and the books recorded nothing, because the books had never recorded a gain.

GM’s stock-based executive compensation created the identical incentive to Boeing’s. Every dollar spent on buybacks lifted earnings per share, which lifted the stock, which lifted executive pay. Every dollar spent forming an operator reduced the metrics that set it.

The company that proved at NUMMI it could match Toyota — when it respected what its workers knew — chose to close plants, buy stock, and reward its executives.

What Toyota did instead

Toyota has never run a massive buyback programme at the expense of manufacturing investment. It has never outsourced its core processes to cut costs. Its capital expenditure as a share of revenue consistently doubles Boeing’s, and its investment in worker development through the Toyota Production System is arguably the most comprehensive frontline training programme in industrial history.

It also consistently outperforms its American competitors on quality, reliability, customer satisfaction, and long-term shareholder returns.

The company that invests in capability outperforms the company that extracts from it. The evidence is not subtle and it is not new.

What manufacturing extraction means

When a retailer is financialized, workers lose jobs and a town loses its shops. When a hospital is financialized, patients lose care. When manufacturing is financialized, people die.

Boeing’s 346 dead are not a management failure in the ordinary sense. They are the logical output of an extraction model applied to an industry where the product must not fail. When financial metrics replace engineering metrics, when buybacks replace investment, the degradation of capability is not a risk. It is the mechanism.

The only variable is time. Boeing had decades of accumulated engineering excellence to burn through before the degradation became visible at 30,000 feet. GE had industrial momentum. GM had NUMMI-era capability to coast on.

Extraction is always on a timer. The question is never whether the bill arrives. Only when.

What I do not know

Here is the part I cannot close, and I would rather name it than perform certainty.

I do not know how you would have found Ramon in a Boeing-sized organisation. He was two levels below me and I knew what he did because I stood on that floor. A company with 170,000 people and a share price to defend has no instrument that surfaces a team leader tracing a scuff to his own tool — and the absence of that instrument is exactly why the capability can be spent without anyone noticing it going.

I do not know what would have stopped it from inside. Every incentive at Boeing pointed one way, and the people who could see the erosion were the people with the least standing to name it. I have been the person on a floor watching a decision made two floors up that I could not reach. I do not have a mechanism for that. I have only ever had luck, and a manager who happened to listen.

And I do not know the number. If a formed operator catches a fault before it becomes a warranty claim — as Ramon did, repeatedly, for years — that is money that never appears on any statement, because the defect never happened. The savings from cutting him are precise and immediate. The cost is diffuse, delayed, and invisible until it is a door plug at altitude. Until somebody can price the fault that did not occur, the ledger will keep recommending the cut.

That is the piece I would most like somebody to build, and I cannot build it alone.

So if you work in a plant, a hangar, a yard, or a line: tell me about a fault that never reached a customer because one person went looking. Then tell me whether anything in your company’s accounts would show that it happened.

Next: The Algorithm That Squeezes — how digital platforms turned extraction into real-time optimization, suppressing wages by the minute while capturing value by the millisecond.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com.

Written with AI assistance. The argument, the judgments, and the floor testimony are the author’s own.

Sources: Harvard Business School Working Knowledge, “Why Boeing’s Problems with the 737 MAX Began More Than 25 Years Ago” (January 2024); The CPA Journal, “The Story of Boeing’s Failed Corporate Culture” (August 2025); Seattle Times, “Boeing and Wall Street: How financialization wrecked a great company” (July 2024); Lazonick and Sakinç, Boeing buyback analysis via The American Prospect; William Lazonick / Academic-Industry Research Network, Boeing 10-K analysis; Industrial Equipment News, “The Boeing Supply Chain Gamble.”

The Lansing Grand River and Buick City accounts are the author’s own.

Source: Soren Kaplan, “AI Can Do the Work. It Still Can’t Replace the One Thing Companies Need Most,” Inc., July 27, 2026. Underlying study: Erik Brynjolfsson, Bharat Chandar, and Ruyu Chen, “Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence,” Stanford Digital Economy Lab, revised November 13, 2025, with the authors’ follow-up note of February 9, 2026.

The first supervisor I hired at Lansing Grand River was Dennis Boutwell. He did not arrive formed. He arrived willing.

What formed him was three years on a floor where the work could go wrong in front of him. A bad call at nine in the morning showed up at nine-forty as a stopped line and forty people standing there looking at him. Not a footnote in a quarterly review. Forty people, looking at him, while the cost ran. He had somebody over his shoulder for most of the first year. Less the second. By the third he was the one somebody else stood next to.

Dennis is now Director of Manufacturing Excellence at Advanced Drainage Systems.

Nobody hired that. Nobody could have. It was deposited into him a piece at a time by a system that guaranteed him a runway long enough to be wrong on, and a man standing there who could tell him which wrong mattered. Somebody paid for those three years, and it was not Dennis.

This week a technology recruiter told the Wall Street Journal that one experienced engineer can now deliver the output of an entire team. In Inc., Soren Kaplan asked the right next question. What happens to the inexperienced ones?

Then he answered it with three bullets aimed at a manager. Create stretch experiences. Build coaching into the pipeline. Measure judgment growth. Do it this week — pick one or two people who look at risk and hand them a hard decision.

Every one of those bullets asks a manager to spend real money building an asset no accounting system will record him building, in the same quarter that the technology which closed the gate is booked as a productivity gain on the line above his. He is being told to be generous with something the ledger says does not exist. That is not a coaching problem. That is a bookkeeping problem wearing a coaching problem’s clothes.

The entry-level job was never mainly a job. It was the formation mechanism. The senior engineer everyone wants to hire right now was made on that runway — production failures, bad architecture calls, teams that didn’t perform as promised. Firms are harvesting that capital and defunding the mechanism that produced it, in the same spreadsheet, and nothing requires anyone to notice.

That’s the argument.

The canary

The numbers are real and more careful than the headlines about them.

Stanford’s Digital Economy Lab, working from the payroll records of the largest payroll provider in the country, found roughly a 16 percent relative employment decline for workers aged 22 to 25 in the most AI-exposed occupations — and it holds after controlling for firm-level shocks. Older workers in the same occupations stayed flat or kept growing. The adjustment came out of headcount, not pay. The declines cluster where AI automates rather than augments.

Kaplan quotes 20 percent. That’s the raw peak-to-trough figure for young software developers, not the study’s estimate of the effect. Use 16.

The authors have also been straight about their own limits: interest rates don’t explain the pattern, but under the strictest controls the divergence only turns significant in 2024, and they say plainly they don’t think AI is the sole cause. The gap is still widening. No reversal.

So the causal question is live. Fine. It is also beside the point.

Whatever closed the entry gate — the model, the rate, the layoff, the reorg — the gate is closed, and there is no institution on earth whose job it is to open it.

It isn’t a tech story

Software went first because software is the most AI-native profession there is: structured, documented, gradable as right or wrong. Legal research, financial analysis, accounting, marketing strategy — same shape, same exposure, a couple of years behind.

Think about what the first-year associate was actually doing when she read four hundred documents. She was not producing four hundred documents’ worth of value. She was building the pattern library that lets a partner glance at a contract in 2035 and feel that something is off before she can say what. The staff accountant tying out schedules at eleven at night was not creating audit value that night. He was becoming someone who can smell a bad number.

Automate the output and you have saved the night. You have also closed the only door that person was going to walk through.

Why nobody stops it

Because formation has never once appeared on a balance sheet.

Buy a company and the assembled workforce you acquired — the crews, the judgment, the thirty years of knowing which machine lies — is not a recognizable intangible asset. It is swept into goodwill and never named again. There is no formation account. Nothing shrinks when you close the apprentice bench, so there is nothing for an auditor to flag, nothing for a board to ask about, nothing for a CFO to defend.

We have solved a problem shaped exactly like this before. Operating leases sat off the balance sheet too, until the standard changed and the right-of-use asset and its matching liability came onto the books: capacity consumed now, funded across years, visible. Formation is the same shape. Consumed now. Funded across years. Currently invisible.

You cannot manage what you don’t measure. You cannot even fail to manage what you don’t record.

The ask

Here is where I’ll be specific, and where I’m least sure, and where I’d rather be argued with than agreed with.

The first employer guarantees three years. Not a training budget — a term. Across that runway a meister teaches, assesses, and confirms deposits into a capability account the worker owns outright, portable, verified outside the employer. Three years, not five. Five was too big to sign, and an unsignable cure is not a cure.

This is not theoretical. One profession never let the gate close. Medicine still runs a multi-year, funded, institutionally mandated formation runway, and hospitals did not cut their interns when the tools improved — because the guarantee sits outside any single employer’s discretion and is paid for from outside any single employer’s quarter. That is the whole trick. Put the runway beyond the reach of the person having a bad quarter.

Manufacturing ran the other experiment. We closed the apprentice bench in the nineties for reasons that all looked correct at the time, and thirty years later we are still short of people like Dennis and still surprised about it. The white-collar professions are about to learn what that costs, roughly one profession a year, starting now.

This is a claim about the United States — 4.4 percent of humanity — and I’d like to know whether it holds where you are.

Somewhere right now there is a twenty-three-year-old who would have been Dennis. She is willing. She is not formed, because nobody is born formed. What she has instead of three years and a man over her shoulder is a chat window and a due date, and she will never once be wrong in front of forty people, which means she will never learn what wrong feels like coming.

In fifteen years someone will need her judgment and it will not be there. Nobody will be able to point to the quarter it was lost, because no ledger in the world was keeping the account.

So tell me what it would take, in your shop, to sign a three-year guarantee for one person. Not the objection in principle — the actual number, the actual meister, the actual reason it can’t be this year. Send me that. The objection is worth more to me than the agreement, because the only people who can build this cure are the ones who’d have to fund it, and they’re the only ones who know where it breaks.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com.

Written with AI assistance. The argument, the judgments, and the floor testimony are the author’s own.

My grandfather Guruswami was one of the first civil engineers to come out of Guindy — the College of Engineering, Madras — in a generation when that meant something specific. India was being wired and bridged and tracked by a small cohort of Indian engineers working under British firms, and the work took them wherever the work was. For Guruswami through the late 1930s, the work was the Howrah Bridge. Braithwaite, Burn and Jessop Construction Company — known in the family simply as Jessop — was riveting twenty-six thousand tons of steel into a cantilever span across the Hooghly, in three eight-hour shifts, racing the war. He was on it.

That is why the four children ended up at Rajghat.

Guruswami and my grandmother Vedambal had four to raise — Jayam and Moham, the two older sisters, then Paddu, my father, then Chotu, the youngest, whose home-name simply meant little one — and a profession that placed the family in whatever remote site the bridge or the railway or the dam required. Day school was not possible. Boarding school was the only answer. They placed all four in the same hostel at Rajghat Besant School on the banks of the Ganga at Benaras, and the children came home for summers and returned to Rajghat for the rest of the year, through high school.

I learned this in pieces, the way one learns most family facts — late, by accident, when something else was being discussed. My father had mentioned Rajghat over the years the way people mention the weather, without ever explaining what it was. It was just where he went to school. The way some people went to St. Stephen’s or Doon. It took me decades to understand that Rajghat was not that kind of school at all — and longer still to understand that the choice was not really a choice. It was a logistical answer to a working engineer’s problem. The formation came afterward, as a gift no one had purchased.

There were three older children too — Cheenu, Ramu, and Vishu — who had grown past school age by the time the work travel intensified, and so were never part of the Rajghat cohort. Seven children, raised across two formations by the simple accident of when they had been born. The older three came up through the conventional south Indian route, then went into the major cities as adults — into the modern professional life that post-independence India was building, into accounting and the other services that the new institutions required. The younger four went north to the hostel. Same parents. Same household values. Same Iyer lineage. Two utterly different formations, imposed by the timing of when each child reached school age relative to the bridge work. That fact alone is worth holding onto, because it tells you that formation is something institutions deliver, not something parents transmit by intention.

*

But Rajghat was the second formation, not the first.

Before the train north, the four younger children were raised in Thiruvidaimarudur, the agraharam in Thanjavur district that the family was from — Cauvery delta country, Mahalingaswamy temple, deep Chola lineage, ritual scholarship in the air. They were raised there by their grandmother. Her name was Savithri. Paati to the children. She fed them yoghurt rice that she had set the night before in a mud pot, kept cool overnight by the breath of water moving through clay walls, ready for them in the morning at exactly the temperature a child should eat in the heat of the Cauvery delta. Thayir sadam from a mann paanai — appropriate technology, in the deepest sense, before anyone had a phrase for it. Climate-responsive nourishment delivered by hand, by someone who loved them, in a village that had been doing it this way for two thousand years.

This is the formation that arrived first. Before Krishnamurti. Before the Ganga. The grandmother’s hand and the cool curd rice and the agraharam morning. The temple bells in the distance. The Tamil that was the language of food and prayer and ordinary life. The rhythm of a household where ritual marked the hours.

The village school they attended held its classes outside, in the sand, under the trees. The children carried slates. They sat on the ground. They recited multiplication tables aloud, and worked through math problems aloud, and the price of an error was a quick pinch of sand from the ground the child was sitting on, pressed onto the forearm or the thigh. A small piece of grit pushed into the skin. Then the recitation continued. The ground itself was the disciplinary instrument — the teacher reached down for what was already there. It was harsh but it was not cruel. The error was registered on the body, not on the child’s standing among the other children. There was no ranking. No comparison. No shame in front of the class. Just the pinch, the corrected answer, the next problem. Consequence without humiliation, which is something contemporary education has almost entirely lost. The body learned because the body had felt the cost of the error. By the time my father was eight or nine years old, the multiplication tables were in his fingertips. The Sanskrit recitations were in his throat. His body had been taught how to hold knowledge, even before he had been taught what to do with it.

Both were the village. Paati’s curd rice in the morning. The pinch of sand under the tree in the afternoon. The grandmother held both. The grandmother was both — the same hand that fed him cool yoghurt rice would also, when it came her turn to teach him a verse, expect it back word-perfect. The Cauvery delta had been forming children this way for two thousand years. Loved and disciplined in the same courtyard, fed and corrected by the same hands, the body enrolled as the medium of memory from the earliest age. This is the formation that arrived first. It is not the formation contemporary education would recognize as benevolent. But it produced children who, by the time they were sent north to Rajghat, had bodies that knew how to learn.

The four children were not orphans. Guruswami and Vedambal were not absent parents. They were a working engineer and his wife following the work, and the children stayed with Paati, in the village, because that is how Tamil joint families had handled this for centuries. Paati was not a substitute parent. She was the primary parent for those early years, by design, and that was understood by everyone in the household. What she gave them was warmth, food, language, ritual, and the certainty of being loved. That is the substrate. Everything else got written on top of it.

When the four eventually went north to Rajghat, they did not arrive as blank slates. They arrived as Iyer children from Thiruvidaimarudur, already carrying Paati’s curd rice in their bodies, already carrying the multiplication tables in their fingertips, already shaped by the agraharam in ways they could not have named. The Krishnamurti formation took hold of them at Rajghat because the soil had already been prepared — by a grandmother who had loved them with cool yoghurt rice from a mud pot, and by a school in the sand that had taught them the body is the seat of memory.

*

Rajghat Besant School sits on land that Annie Besant chose and that Jiddu Krishnamurti — the Tamil Brahmin boy the Theosophists believed would be the next World Teacher, who later refused that role and dissolved the order built around him — turned into one of his foundation schools. The school existed for a single purpose: not to prepare children for examinations, not to credential them, not to make them employable. It existed to give them the conditions in which they could, in his phrase, flower in goodness.

He was explicit about what stood in the way. “The purpose of education is not to produce mere scholars, technicians and job hunters, but integrated men and women who are free of fear; for only between such human beings can there be enduring peace.” He believed competition was the disease, not the cure. “We are concerned with the flowering of goodness which cannot possibly flower where there is any competition.” He believed examinations and rewards cultivated jealousy and anxiety in the young. He believed the teacher had to be inwardly free before a student could be. He believed the school had to be a place where, as he put it, one learns about the totality, the wholeness of life.

I read those lines for the first time last week, on a phone, in Ohio. I recognized every one of them. Not as new ideas. As ideas I had grown up inside.

*

The four were not housed together. Rajghat had a boys’ hostel and a girls’ hostel, separated by the convention of the time, and visits between them required crossing the campus. So Paddu and Chotu — the younger brothers, small enough to need supervision and old enough to resist it — would make the walk over to the girls’ hostel to see their sisters. They tended to arrive disheveled. Shirts untucked. Flies often open. Jayam and Moham would see them coming, recognize the state of them at a glance, and take them aside before they could be seen by anyone else — buttoning what needed to be buttoned, tucking what needed to be tucked, fixing the brothers so they could be returned to public view. The sisters made up for the parents in absentia.

*

When my father spoke of Rajghat, which was rarely, two things came up. He swam against the current in the river. And afterward he drank large glasses of milk.

That was the school, distilled into what a boy remembered. Not examinations. Not marks. Not the names of teachers or the prize-giving days. A current and a body learning to move against it, and the simple wage the body collected for the work. The Ganga flows past Benaras toward the Bay of Bengal. He swam the other way. Then he drank the milk.

I do not think he ever told me this as instruction. He told it the way men tell river stories — for the pleasure of the water, the cold, the appetite at the end. But the image stays, because it is the only one I have of him as a boy, and because it turns out to be the whole thing in one frame. A school that teaches a child to swim against the current, and feeds him after, has already done most of what a school is for.

My father never quoted Krishnamurti to me. I do not remember him saying the name with any particular weight. He was an Iyer by birth and lineage, raised on his grandmother’s curd rice, fluent in the Tamil and the Sanskrit names of things. But he was not ritually observant in the way an orthodox south Indian Iyer of his generation would have been expected to be. He did not perform daily puja with the conviction of a man who believed the repetition mattered. He had been taught at Rajghat that ritual itself was a form of mechanical conditioning — Krishnamurti had been as explicit on this as he was on competition or fear. The constant repetition of a ritual, of a word, of a prayer, is a gratifying sensation to which a noble term is given. That was the line, and it had landed in my father somewhere in the hostel years. He kept the Tamil. He kept Paati’s food. He kept the village. He set down the daily mechanical performance of orthodoxy. He worked in coal-town Dhanbad. He read in Hindi as easily as Tamil. He sent me to a Jesuit school named for an Italian missionary who studied Sanskrit and wore saffron robes.

None of this looked like Krishnamurti from the outside. All of it was Krishnamurti, in the sense that Krishnamurti had named the surgical operation that distinguishes the living inheritance of a culture from the mechanical apparatus that calcifies it. My father had made that distinction in himself.

But what got transmitted was not vocabulary. It was temperament.

The household held achievement lightly. Marks were noted, not celebrated. Comparison to other children was almost never used as a motivator — and when it surfaced, usually from a visiting uncle, my father would let it pass without taking it up. Questions were welcome at the dinner table, including the kind that did not resolve. Silence was permitted. The credential economy that organized so many other Indian middle-class households — the IIT-or-bust pressure, the rank-as-identity reflex, the relentless ranking against cousins — did not organize ours. I did not know this was unusual until I left home.

What I now understand is that my father was carrying Rajghat into the household without naming it. The temperament was the air I breathed. By the time I encountered Drucker and Deming and Ohno as an adult — thinkers who would name competition as destructive of systems, who would insist on the dignity of the worker, who would argue that fear is the silent saboteur of any operation — the receptors were already built. Those Western thinkers did not form me. They named what was already in me.

The line runs:

Thiruvidaimarudur and Savithri Paati → Krishnamurti and Rajghat → my father → the Dhanbad household → me

Four of those five links I did not choose. I inherited them in the air of the home. The way questions were asked. The way achievement was held lightly. The way certain kinds of conformity were quietly resisted. The way food was something a grandmother prepared at the right temperature for the right hour.

*

This is the part that matters for everything I am now trying to build.

The Bloom System, the four gates, the Capability Account — none of these are inventions. They are the attempt to translate a private inheritance into a public deposit. What my father got from Rajghat because his parents had no other option, and what I got from him by accident of being his son, most children do not get. They get the household their parents could afford to assemble, in the temperament their parents inherited from theirs, inside an education system that — almost everywhere now — runs on the opposite of Krishnamurti’s principles. Fear is the curriculum. Comparison is the metric. Credentials are the deposit. Formation, if it happens at all, happens by luck.

Krishnamurti understood the diagnosis. He named the disease with a precision no one has improved on. But he refused, on principle, to specify what the school must verifiably deliver. That refusal was a luxury his eight schools could afford. They served a few thousand children. They produced a particular kind of Indian intellectual — including, in a quiet way, my father.

What I am trying to build cannot afford that refusal. If formation only transmits by family lineage and luck of enrollment, it will continue to produce people like me by accident — and continue to fail the hundred million children whose parents did not, and could not, find a Rajghat. The Capability Account is the instrument that turns Krishnamurti’s diagnosis into a societal commitment. It says: the deposit a child receives between birth and thirty does not have to depend on whether their father once boarded a train north and got off in Benaras. It can be made measurable at the home gate, verified at the school gate, compounded at the college gate, and audited at the first-employer gate. Not as a credential. As formation.

*

There is one more thing the family has to disclose before the argument closes.

I am not building the Capability Account alone. Among the founding principals of the Capability Capital Institute is Sridhar Ramamoorti, an accounting professor at the University of Dayton, and Sridhar is my first cousin. His father was Ramu, my father’s older brother — the second of the three older siblings who had grown past school age before Guruswami’s work travel intensified, and who therefore never went to Rajghat. Ramu became an accountant. He raised Sridhar mostly in Bombay, the city where post-independence India was learning how to measure itself — where charter accountancy was professionalizing, where the stock exchange and the auditing firms and the Reserve Bank were building the modern apparatus of how a country counts. Sridhar grew up inside that apparatus. He became one of the country’s serious thinkers on forensic accounting, governance, and how institutions measure what they claim to value. He was born in 1963. So was I.

We work as peer equals — which in Tamil Brahmin terms is itself a small refusal, because Sridhar is the elder cousin in the eldest brother’s line, and tradition would assign him deference rather than partnership. We have set that aside in order to build the work properly.

This matters because of what the Capability Account actually is. It is measurement applied to formation. You cannot build it from formation alone — that is what Krishnamurti tried, and his beautiful refusal to specify what the school must verifiably deliver is exactly why his vision served only a few thousand children. You cannot build it from accounting alone — that is what the credential economy has done, applying measurement to the wrong things and producing a world that audits examinations and ignores flowering. You build the Capability Account by combining the two formations. Which is what is happening now, between Wooster and Dayton, between a manufacturing executive who came up through Paddu’s Rajghat formation and an accountant who came up through Ramu’s Bombay one.

The two cousins inherited not just two different childhoods but the two main trajectories that post-independence Tamil Brahmin families produced for their sons: the professional services route — finance, accounting, law — built in the major cities, oriented toward institutions and the modern bureaucracy of capital; and the industrial route — plant floors, manufacturing, the actual making of things — that built the country’s productive base. The accountants saw what was wrong with how value gets measured. The industrialists saw what was wrong with how capability gets formed. The Capability Account requires both diagnoses fused into a single instrument. Bombay needs Wooster. Wooster needs Bombay. Both need Thiruvidaimarudur as the substrate.

One generation late, two of those children’s sons — born the same year, raised in different cities, separated by the accident of when their fathers had reached school age — are bringing what each lineage carried back together. The accountant from Bombay and the manufacturing executive from the hostel on the Ganga. Both of them grandsons of Savithri Paati, even though only one cohort of cousins was old enough to remember her hands. The Capability Account is what their two fathers’ formations make possible when they meet again in their sons.

This is not a sentimental fact. It is a structural one. The argument the book is making about how formation must be measured rests on the proof that measurement and formation can be brought together — and the proof, in the first instance, is the collaboration of two cousins who inherited the two halves of the same household.

*

I never saw the schoolhouse on the Ganga. I have not been to Rajghat. The four who slept in the hostel through the late 1950s are all gone now — Chotu the youngest first, then Jayam, then my father Paddu, and Moham the last of them. The school today is no longer the place it was when they were boys and girls there. But I have been formed by it, through one remove, in a coal town a thousand miles south. And the thing I owe Krishnamurti — and owe Guruswami and Vedambal for placing their four younger children at the only school the working life of an early Guindy engineer permitted, and owe Savithri Paati who fed them curd rice from a mud pot, and owe the village school that taught them under the trees with slates and sand — is not the language of his books. It is the recognition that what I am trying to do is not original. It is an attempt to make available, by design, to children who will never see the Ganga or the Cauvery, what a village and a school once made available to my father by sheer good fortune.

That is the whole project, stated as plainly as I know how.

Formation is not a curriculum. It is what the household, the school, the college, and the first employer either deposit in a child or extract from them. It transmits across generations even when the institution does not. And what an accident of geography did for one Iyer family in Benaras seventy years ago, a Capability Account can do — must do — for a country.

*

Published on 19 August, my father’s birthday. Paddu — G. Padmanabhan — Rajghat 1950s, Indian Railways, Dhanbad.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com.

Written with AI assistance. The argument, the judgments, and the floor testimony are the author’s own.

E2, Foundation track — Part 3 of 3

The previous essay ended on the gap. Profit sharing is a flow that resets. Equity is a stock that requires a sale. What neither does is attribute value to the person who produced it and let him carry it when he goes.

This essay is about what an instrument would have to do to close that gap, what it costs, and the 2 things in it I cannot yet answer.

I. The claim, where a CFO can see it

None of it is worth arguing unless the deposit pays for itself. So here is the falsifiable claim, on the table, where a chief financial officer can see it.

The firm must capture the majority of the return. Not a minority. Not a rounding error to be defended in a values statement. If the employer nets positive after the deposit, this survives a downturn, and it survives the arrival of a CFO who does not care about any of this. If the employer nets negative and does it anyway, it is philanthropy, and philanthropy dies in the first bad quarter. That is not a hypothesis. That is what happened to every training budget in this country in 2009.

In March of this year a research team led by Fidan Ana Kurtulus at the University of Massachusetts Amherst — with Hoyt, Ouimet, Blasi, Kruse, Freeman and Castellano — published the largest study of its kind, joining 3 restricted Census micro-databases to the Department of Labor’s Form 5500 filings. 44,000 American manufacturing establishments across 22,500 companies, observed in 2010 and again in 2015.

They define labor productivity the way a plant does: total value of shipments and inventory divided by total hours worked.

The 600 establishments that adopted an ESOP after 2010 showed labor productivity higher by 5.6 to 6.7 percent when measured in 2015 — the total change across the period, after controlling for management practices, headcount, capital and material inputs, union status, geography and industry trend.

Then the finding that matters most. Establishments that adopted an ESOP and maintained broad-based group performance pay across those same years showed productivity growth of 12.93 percent — an average of 2.6 percent a year.

Roughly double.

The instrument alone is worth something. The instrument paired with the practice is worth twice as much. A deposit is not a payment. It is one component of a formation apparatus, and it badly underperforms when it arrives alone.

And consider what the Census counted as the practices that do the multiplying: performance goal monitoring, remedial action when the production process goes wrong, broad-based information sharing with workers, and group performance bonuses inclusive of non-managerial employees.

That is not a management theory. That is a description of a plant manager’s job, written by the Census Bureau, and it turns out to be the multiplier on the money.

II. Twice, on 2 continents

I have watched it work twice, 18 years apart, and both times it looked the same.

At Lansing Delta Township, the fix that unblocked a bottleneck costing $1M to $1.5M a day was 4 things: line balance on over-cycled stations, training cycles on a third shift where new people had not had enough repetitions, a roof molding station never run at full rate, and defect containment before chassis. Containment first, then irreversible corrective action, engineers and team leaders working it at the station with the people running it.

At Royal Enfield in 2008 we inherited a list of more than 100 customer-facing defects. Dr. Nair pulled us through it problem statement by problem statement, with the repairmen, the drivers, the testers, the production team members — not just containment so defects stopped reaching customers, but correction, so the check-and-repair loop could be dismantled rather than permanently staffed. Those fixes then went into the design of the UCE engine and the Classic 350 and 500.

The Classic is generally credited with the turnaround, and it should be. But the Classic was designed out of knowledge the floor generated while fixing the machine that came before it. The product was the output of formation, not the alternative to it. Volume went from 50,000 units to 113,000 with roughly 20x profit growth, and a rising Indian market does not build you a motorcycle that stays fixed.

III. Why a guarantee, and not a promise

In December 2021, General Motors applied to the city of Lansing for incentives to build a battery cell plant in Delta Township. $2.5 billion. 1,700 full-time jobs “through 2026.” A federal loan of $2.5 billion followed. A United States Senator said the cells built in Lansing would power vehicles built by Michigan autoworkers across the state. In January 2024 a hiring event drew more than 400 people.

As of March this year, that plant belongs to LG Energy Solution. GM sold its stake in 2024. It is now a $4.3 billion facility that will build prismatic cells for Tesla’s grid storage systems, assembled in Houston, beginning in 2027. At the sister plant in Lordstown, 1,334 hourly workers were cut effective January 5 of this year — 850 designated temporary with no return date, 484 indefinite with no callback plan.

Nobody lied. That is the point. The entity changed, the product changed, the customer changed, and the timeline changed, underneath a workforce that had been told to prepare.

Now hold that against the observation window in the Census research. 5 years between measurements. The productivity that an ownership instrument produces when paired with real practice is not a quarterly effect — it is visible only across a span longer than most of these promises survive.

You cannot ask a man to form himself over 3 years against a target that reassigns itself in 2. And you cannot call him short-sighted when he declines. A man who can be walked out in 30 days is pricing his own tenure correctly.

Which is why the 3-year employer guarantee is not generosity and is not a benefit. It is the precondition that makes a wealth horizon thinkable at all.

IV. The instrument

5 moves. The first 2 are available to a plant manager on Monday. The last 3 need a board.

1. Publish the number. He does not know what his improvement was worth. You do. Put it on the board, in dollars, by station, every month.

2. Pair the instrument with the practice. 5.6 to 6.7 percent alone. 12.93 percent together. A deposit arriving without goal monitoring, remedial action at the station, and real information sharing gives up half its return.

3. Index to capability, not to exit. First-pass yield. Scrap. Unplanned downtime. Warranty rate. Changeover time. Turnover replacement cost. Current period. Auditable. Every plant manager reading this has these numbers on a board somewhere in his building already.

4. Make it portable. It travels with him, the way an ESOP balance travels, because the capability travels with him.

5. Guarantee 3 years.

Now the 4 questions a CFO asks in the first 90 seconds, and I am going to answer them plainly, including the 1 I cannot yet close.

What does it cost?

Roughly 8 hours of a 40-hour week in formation during the 3-year window. 20 percent of paid time. To hold line output flat, that implies something like a 25 percent headcount uplift on the covered population. That is the largest number in this proposal and I am not going to soften it.

That 20 percent is all-in. It covers the learner’s hours off the line and the experienced hand who does the forming. There is no separate instructor budget hiding underneath it.

That second part matters more than it sounds. Formation is not a video and a checklist. It is a man who has run the station for 20 years standing next to a man who has run it for 3 months, and the first man’s hours come off the line too. Any proposal that costs the apprentice’s time but not the teacher’s is not costing formation. It is costing attendance.

And one more number, because it is what a plant manager asks the moment he hears the percentage: how many experienced hands does this take, and where do I find them?

Formation partners are not pegged to a position on the org chart. They are pegged to formation load — roughly 1 per 12 people actively inside the window. In a mature plant at steady state that lands near 1 per shift leader. At a launch, a ramp, or the addition of a third shift, it is several times that. Plants that skip it pay the difference in downtime, and I have paid it.

Two more things about the 20 percent. It is approximately the German dual-system ratio, so it is not invented. And more importantly: a good part of it is already being spent, badly. My 2 weeks at Lansing Delta Township were formation under duress — line balance, training cycles, defect containment, engineers and team leaders at the station. Nobody budgeted it. Nobody measured it. Doing it late cost $10M to $15M in margin plus a launch window that could not be bought back at any price. This is not new spending. It is moving unbudgeted emergency spend into a scheduled line.

Where does the account sit?

Alongside the ESOP, in existing employee benefit plan architecture. Trust, trustee, third-party administrator. Not a new legal invention.

And here is a live question I am not going to paper over. Retirement architecture is designed to be illiquid until 59½ — that is the whole intent. If the Capability Account sits exactly where an ESOP sits, a 34-year-old who leaves takes it with him and cannot touch it for 25 years. That is portable in law and useless against a transmission. Whether the deposit should be qualified, non-qualified, or split — with a defined share becoming accessible at the close of the guarantee — is the sharpest open design question in this system. It is a technical question with a technical answer, and it needs a lawyer, not an essayist.

What happens when he leaves?

He takes the record with him, the way he takes an ESOP balance. Fully portable at separation. And the record is worth more than the balance: a balance is money, but a record is a priced claim about capability that any employer can read.

Why would I form someone my competitor will hire?

You win some and you lose some. Once the ecosystem is running, you also inherit what walks in the door already formed — and the firm that hires your man does not pay to form him again. Across a sector, formation is a pooled asset rather than a private one.

That answer is only honest at scale. The first mover eats a pure loss: he forms, he leaks, and he receives nothing back, because nobody else is forming yet. That is a collective action gap, and it is why this is institutional work rather than a memo to a plant manager. Anyone who tells you a single firm can solve it alone has not run one.

V.

Two things in this proposal are unsettled and I am not going to pretend otherwise.

The first is whether the deposit can be made deployable without losing its tax treatment. That is a technical question with a technical answer, and it needs a lawyer rather than an essayist. Until it is closed, the account is portable in law and may still be useless against a transmission.

The second is not technical. Somebody has to form before the ecosystem exists, and that firm eats a pure loss. I do not know who that is. It is not obviously a plant and it is not obviously a policy.

What I am certain of is the wound. A man on a line produces a number his employer knows and he does not, and after 40 years of the best instrument we have ever built for him, he owns nothing.

I am much less certain about the cure. The 5 moves above are what I can defend with evidence today, and no more than that.

So break one. The 20 percent. The custodian. The vesting. The pooling logic. If one of them will not hold, I would rather find out now, while it is an argument on a page, than later, when somebody is running it in a plant with real people on the line.

Sources

Fidan Ana Kurtulus, Hoyt, Ouimet, Joseph Blasi, Douglas Kruse, Richard B. Freeman and Castellano, “Employee Share Ownership, Management Practices, and Labor Productivity: An Analysis Using Establishment Level Micro-Data from the U.S. Census,” SSRN 6490798, 29 March 2026.

National Center for Employee Ownership, summary of the above, 12 May 2026.

City of Lansing incentive filings for the Delta Township battery cell plant, December 2021; U.S. Department of Energy loan announcement.

LG Energy Solution, Lansing facility, March 2026.

Ultium Cells Lordstown, workforce reduction effective 5 January 2026.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com.

Written with AI assistance. The argument, the judgments, and the floor testimony are the author’s own.

Evidence They Can’t Defend — Essay 8 of 13

I have stood at the top of a subcontracting chain. Here is what you can see from there, and what you cannot.

Around 2011 I went looking for land to build a new Royal Enfield plant outside Chennai. The Chennai line had been building motorcycles since 1955 and had run out of room. Finding fifty acres in Tamil Nadu is not a transaction; it is a campaign. I sat in the Chief Minister’s office. Then the Industry Minister’s office. Then SIPCOT, the state corporation that actually holds the land. Then the home offices out in the suburbs, where ministers see people on the days they are not seeing people. It took months.

We got the allotment: fifty acres at the SIPCOT Industrial Growth Centre in Oragadam. Royal Enfield’s announcement recorded it in one clause — the Tamil Nadu Government has allotted us land. Everything above compresses, in the public record, into eight words.

Then the bhoomi pooja, the ceremony that blesses the ground before anything is built on it. It was a small affair — a handful of us from the company, the priests, and nobody else. Fifty empty acres, and perhaps a dozen people standing on them, asking the ground for permission.

We also meant to make the work visible. I flew to London and spent time with a design firm there scoping a living museum to be built into the plant — a walk through Royal Enfield’s history that ended with visitors standing above the line, watching motorcycles being made in front of them. We paid for the design. It was never built.

Construction started in February 2012. The company’s own account of what followed is worth reading closely: civil work continued around the clock, and the plant was finished in a record eleven months.

Around the clock. Fifty acres. Eleven months.

I drove onto that site through most of it. The contractors lived on the land they were building — families on the front fifty acres where the work was, while the back fifty, which we had not yet decided what to do with, stood overrun with brush. Cooking fires in the morning. Children. The ordinary domestic life of people housed by the job because the job is nowhere near anywhere else.

The plant opened in 2013 with about two hundred people, a fifth of them women. That number is in every article written about the place.

Now count what the record holds. It holds the acreage, the allotment, the eleven months, the capacity, the two hundred jobs, and the name of the man who rolled the first motorcycle off the line. I can give you the Chief Minister’s office, the Industry Minister’s office, SIPCOT, my own name on the capital approval, and the colleagues who stood beside me at the ceremony.

And I cannot give you one name from the round-the-clock. Not one. I chose the civil engineering firm — they were near Luz, in Mylapore — and I have gone looking for them and cannot find the name in any record I can reach. Below them were subcontractors I never met. Below those, labour contractors. And at the bottom, the people whose cooking fires I drove past at seven in the morning on the way to my own site office.

That is the argument. Not that anyone on that project was a villain — I was on it, and I was not. It is that the chain is built to end that way: a full record at the top, an empty one at the bottom, and the man who signed the first contract genuinely unable to name the last rung.

Wage theft costs American workers an estimated $50 billion a year, and construction is ground zero. The numbers below are American. The structure is not — I watched it in Tamil Nadu, and it had the identical shape.

The chain

A modern commercial project works like this. An owner hires a general contractor. The GC hires subs. The subs hire sub-subs. The sub-subs hire labour brokers. The brokers hire workers — often classified as independent contractors regardless of what the work actually is.

At every layer someone takes a cut. By the time a dollar of project budget reaches the person swinging the hammer, it has been reduced by margins at four or five levels.

This is not a supply chain. It is an extraction chain. Each layer exists to distance liability and capture margin — and the distancing runs both ways: it protects the owner from the worker’s injury, and from ever knowing the worker’s name.

Misclassification as business model

The Economic Policy Institute estimates 10 to 30 percent of employers misclassify workers as independent contractors. In construction it runs higher: 12 to 21 percent were misclassified or paid off the books in any given month in 2017, and the Century Foundation puts as many as 2.1 million U.S. construction workers in that position.

A construction worker classified as an independent contractor loses as much as $19,526 a year against what he would have earned as an employee — minimum wage protection, overtime, unemployment insurance, workers’ compensation, Social Security and Medicare contributions, anti-discrimination protection. The employer saves roughly 30 percent on labour costs by avoiding all of it.

That 30 percent does not make the project cheaper for the client. It flows upward — to the subcontractor’s margin, to the GC’s profit, and eventually to whoever signed the capital approval. Which, on one project in Oragadam, was me.

The $50 billion

One in five construction workers experiences wage theft: below-prevailing-wage pay, unpaid overtime, denied breaks, off-the-clock work. Those workers are paid $23,000 to $30,000 a year less than they are legally owed.

Payroll fraud in construction alone illegally cuts contractor labour costs by $6.2 to $11.7 billion a year. Workers’ compensation programmes run a $1.7 billion shortfall, unemployment insurance loses up to $725 million, as much as $4.3 billion owed to Social Security and Medicare goes unpaid, and the federal treasury loses roughly $3 billion.

Total wage theft across all industries: $50 billion a year. That is more than all robberies, burglaries, larcenies, and auto thefts combined.

Wage theft is the largest category of theft in America, and it runs from employers to workers rather than the other way.

Who carries the risk

American construction generated roughly $2.1 trillion in 2024. Financialization has concentrated profit at the top of the chain and pushed risk to the bottom, and general contractors with sophisticated legal structures face minimal liability for wage theft by their subs — even when those subs were selected because artificially low bids signalled suppressed labour costs.

A misclassified worker who gets hurt has no workers’ compensation. One laid off between projects has no unemployment insurance. One paid $12 an hour against a $22 prevailing wage files with an understaffed Department of Labor and waits years.

The people who bear the risk have the least power. The people who capture the value bear the least risk. That is not a failure of the design. It is the design.

What a compressed chain looks like

Unionised construction runs the experiment in reverse.

Union contractors hire workers directly. Classification is not ambiguous. Training runs through joint apprenticeship programmes that produce the most skilled workers in the trade. The result is higher quality, lower defect rates, better safety — and a 2017 Illinois study found union projects generate 35 percent more state tax revenue, because the workers are properly classified, paid, and insured.

Investment in workers produces better outcomes and higher public revenue. Extraction through misclassification produces worse outcomes and starves the systems those same workers will need.

Why it stays invisible

Construction extraction does not happen in a boardroom or appear in an SEC filing. It happens jobsite by jobsite, paycheck by paycheck, in an industry where responsibility is diffused through layers of contracts. No single actor is extracting $50 billion; the extraction is structural, built into how work is organised, classified, and compensated.

That makes it harder to see than a leveraged buyout. It does not make it smaller.

What I do not know

Here is where I run out, and I would rather say so than perform certainty.

I do not know what I could have done differently. I could have written labour standards into the contract with the civil firm — flow-down clauses, classification requirements, an audit right. Whether any of that survives a schedule promising eleven months round the clock, I do not know. I have never seen it tested on a project running that hot.

I do not know who audits the bottom of a chain that long. The owner has no relationship with the brokers, the brokers none with the client. Every party can honestly say the workers are somebody else’s, and every one is telling the truth.

And I do not know whether the record can be fixed from the top at all. It may be that only a worker-owned, portable record — one starting with the person rather than the contract — ever reaches the last rung. Construction is the hardest case for that instrument, not the easiest.

What I am certain of is smaller, and it has taken me thirteen years to say. We commissioned a museum so that visitors could watch the work being done. We did not record who did the work of building the place it would have stood in. Fifty acres were blessed by a dozen of us whose names are all recoverable, and built around the clock by people whose names are not — a workforce that appears in no account of that eleven months, including, until now, in mine.

So if you have ever signed first on a project — the capital approval, the ceremony on the empty ground — go and find how far down your own record actually reaches. Then tell me where it stops, and whether anything in your power could have carried it further.

Next: The Factory That Bought Its Own Stock — how Boeing, GE, and GM chose stock buybacks over engineering, safety, and their own workers.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com.

Written with AI assistance. The argument, the judgments, and the floor testimony are the author’s own.

Sources: EPI, “Misclassifying workers as independent contractors” (2025 update); EPI, “Employers Steal Billions from Workers’ Paychecks Each Year”; The Century Foundation, “Up to 2.1 Million U.S. Construction Workers Are Illegally Misclassified” (2023); IIIFFC wage theft data; Fair Contracting, “Wage Theft Facts” (2023); Bureau of Labor Statistics, Contingent and Alternative Employment Arrangements; Royal Enfield / Eicher Motors announcements on the Oragadam facility (land allotment, February 2012 construction start, eleven-month build, 2013 commissioning, workforce at opening).

The Oragadam account — the land allotment, the ceremony, the London museum scope, and the construction period — is the author’s own recollection.

E2, Foundation track — Part 2 of 3

I ended the previous essay by saying that the company had no organ capable of registering what I heard on the floor. No line on the P&L. No field in the system. No metric anyone was held to.

That is a design fault, not a moral one. Design faults have solutions. This essay is about what the solution has to do.

I. The best argument against me

Start with the strongest case that none of this is true.

The frontline worker in an American auto plant has held a formal, transparent, contractual claim on the value he helps create for 4 decades. It is called profit sharing, and it is better built than most people realize.

The GM–UAW formula is mechanical and published: $1,000 for every $1 billion of North American pretax earnings, paid in increments of $250, to any employee with 1,850 or more compensated hours in the year. Ford’s works the same way. There is no discretion in it. There is no committee. A man on the line at Lansing Grand River can read the quarterly earnings release, do the arithmetic on his phone, and know what is coming before the company announces it.

So the objection writes itself. He does have a claim on value. It is auditable. He can see the number.

I want to concede this completely, because the concession is where the argument gets made.

General Motors signed a document decades ago stipulating that the men and women on its floors hold a legitimate claim on value they helped create. That principle is not in dispute. It has never been in dispute. I am not here to argue for a new principle. I am here to argue that the principle was settled in the wrong instrument.

II. 4 properties

Profit sharing fails to build wealth, and it fails for 4 structural reasons, none of which is anyone’s bad faith.

It is collective, not individual. The check is indexed to the company’s regional pretax earnings. Not to his line, his yield, his scrap rate, or the 4 points of first-pass yield he pulled out of a process last quarter. Two men on the same shift — one who transformed his station and one who did not — receive identical checks.

It is a flow, not a stock. A check, not a balance. There is no account. There is no statement. There is nothing that has a value on a Tuesday in March.

It resets annually. Nothing compounds. Nothing accumulates. Nothing transfers to a child. 40 years of participation leaves precisely the asset position of 1 year: none.

It is indexed to what he cannot touch. Tariffs. A fire at a supplier’s aluminum plant in New York. An electric-vehicle strategy decided in a room he will never enter. The formula is honest. The inputs are entirely outside his hands.

III. February, in full

Watch what those 4 properties produce across 3 years.

For 2024 performance, about 47,000 GM workers received up to $14,500 — the largest payout in the program’s history.

For 2025, announced this January, the figure was $10,500. Down 32 percent, the lowest since 2021, because North American pretax profit fell 28 percent to $10.4 billion.

Ford’s workers received up to $6,780, down from $10,208, paid March 12 — hit by roughly $2 billion in tariff costs and a supplier fire that stalled F-Series production.

Stellantis workers received nothing. For the first time since the merger formed the company, the minimum thresholds were not met. The company lost $26.3 billion.

I never received one of these checks. Profit sharing is an hourly instrument and I was salaried management, so what I can report is not what February felt like. It is what February looked like from where I stood.

I usually learned the number the way the floor did. From the papers.

Then I would walk the line and say something about it. Congratulations on the payout. The responses sorted into a small number of shapes, and after enough years I could predict which one was coming.

When the figure landed below the maximum, I heard about the goalposts. Management had set targets nobody on that line could influence — electric vehicle volume was a favorite example, and a fair one — and then paid out against them. The word that came up was dishonest.

When the maximum hit, I got a smile.

If I probed past the smile it went one of 2 ways. Either I deserve a lot more than this for what I do for you, or I’m using it to pay down my debt.

What I almost never heard was a plan. Very rarely a vacation. Very rarely a purchase somebody had been waiting on. Almost never anything that would still exist in 5 years. I am reporting an absence across many Februaries rather than counting anything, and I want that stated plainly.

Now hold those 3 responses against the 4 properties.

The goalpost complaint is the fourth property, stated by the people it lands on. They were not wrong. EV volume was decided in rooms none of them would ever enter, and their February depended on it.

I’m using it to pay down my debt is the wage arriving already spoken for, one more time, at a larger number.

And I deserve a lot more for what I do for you is the closest thing to a claim I ever heard on a factory floor in 36 years.

Notice what it is missing. It has direction and no magnitude. It is a sense of desert with no figure attached — made, every February, to the one man in the building who had the figure on his desk and never once thought to bring it down to the line.

That is the best instrument the American frontline has ever been given. Transparent, contractual, negotiated by a union at the height of its leverage. And after 40 years of it, a 55-year-old in Lansing owns exactly as much of anything as he did the day he hired in.

IV. Somebody already solved half of it

Here I have to be fair, because the most serious answer to this problem did not come from a union, or a policy institute, or from me. It came from private equity.

In 2015 KKR bought C.H.I. Overhead Doors, a garage door manufacturer in Arthur, Illinois. At acquisition, all 800 employees — people in the factory, people driving trucks, people in the corporate office — were made owners of the business. It was free. It was incremental. It was explicitly not traded against wages or benefits, and wages rose 7 percent in 2020 and 12.5 percent in 2021 anyway.

Over 7 years, EBITDA rose nearly fourfold organically. Margin climbed more than 1,400 basis points, from 21 percent to well over 30. Revenue grew roughly 120 percent. The improvements came from procurement, scrap reduction, labor productivity, working capital.

Read that list again. That is not a finance story. That is a plant.

In 2022 KKR sold C.H.I. to Nucor for $3 billion, roughly 10 times invested capital. Hourly employees and truck drivers averaged about $175,000 on their equity, on top of some $9,000 in dividends over the holding period. The most tenured cleared more than $750,000. An office manager with 17 years took home 5.5 times her annual salary.

Pete Stavros, who runs KKR’s Americas private equity platform and drove the program, was asked why a buyout firm would do this.

“This isn’t charity, it’s not a gift.”

He went on to say that the workers drove an enormous amount of productivity in the business.

I have been making that argument for years. He was making it from the other side of the table, with a 9.8x multiple to prove it. The strongest current proponent of the capability argument is a private equity co-head, not a labor advocate. If you want to know whether this is a business strategy or a moral appeal, that is your answer.

V. And why it still isn**’**t the answer

Now the hard part, and I have to hold myself to the standard I set in the previous essay.

That office manager received 5.5 times her salary because KKR exited at nearly 10 times invested capital in the first quarter of 2022. Same woman. Same 17 years. Same work. Move the exit to 2009 and she receives nothing.

That is the Mracek problem in a better suit. I spent an essay establishing that wealth produced by timing is not earned wealth. I cannot arrive here and applaud wealth produced by timing because this time the beneficiary was on the floor.

Broad-based equity fixes the most important of the 4 properties. It is a genuine stock. It compounds. It transfers to a child. That is real and it is not a small thing.

But it keeps 2 of the others and adds a new one.

Still collective. You are paid for the enterprise’s outcome, not for the capability you formed.

Still timing-contingent, and now event-contingent. Ownership Works — the nonprofit Stavros founded, now working with Apollo, Ares, Silver Lake, TPG and others — describes payouts as arriving within roughly 5 years, tied to events such as a sale. Which means the frontline worker’s wealth now requires that his employer be sold. Consider what that asks a man to hope for.

And the base rate is not the headline. Across the 41 liquidity events in the Ownership Works portfolio to date, the 20 completed exits have averaged about $55,000 per employee-owner. That is real money and I do not want to diminish it. It is not $175,000. C.H.I. is the ceiling, not the median.

Concentration. His employer now holds his job and his savings. That is the objection the ESOP literature has fought over for 40 years and it has not gone away.

So: profit sharing is a flow that resets. Equity is a stock that requires a sale. Neither is indexed to the thing the man actually did.

What is missing from both is attribution and portability. A claim tied to the capability he formed rather than to his employer’s exit multiple. An account that travels with him, because the capability travels with him.

VI.

I have been looking for an instrument that does both, and I have not found one in the field.

Profit sharing solved transparency 40 years ago and stopped there. Broad-based equity solved the stock problem and tied it to a sale. The ESOP solved ownership and locked it until 59½. Every one of them is better than nothing and not one of them puts an asset in the hands of a 38-year-old who improved a process last quarter.

The next essay describes what I think such an instrument has to do, what it costs, and the 2 questions in it I cannot yet answer.

I would genuinely prefer to be told it already exists. If you know of one — any sector, any country — that attributes to the individual, holds as a stock, and travels at separation, write and tell me. I would rather adopt than build.

Part 3 of 3: **A Deposit Is Not a Payment** — what the instrument has to do, and what it costs.

Sources

UAW–General Motors national agreement, profit-sharing formula.

General Motors, Ford Motor Company and Stellantis 2025 results and profit-sharing announcements, January–March 2026.

KKR, sale of C.H.I. Overhead Doors to Nucor Corporation, 2022.

Pete Stavros, interview, CNBC, May 2022.

Ownership Works, impact reporting, 2026.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com.

Written with AI assistance. The argument, the judgments, and the floor testimony are the author’s own.

I.

This summer the median American home sold for $440,600. Prices are up more than 50 percent in 6 years. 77 percent of listings are out of reach for a middle-income earner, and nearly half of all renters pay more than a third of their income for shelter.

NPR reported this on July 30. Read it for what it documents rather than what it says.

Matt and Amanda Mracek live outside Orlando. He bought a foreclosed house in college after the 2008 crash, on a federal tax credit. She bought a 900-square-foot fixer-upper for a little over $100,000 with $5,000 down. A few years later their realtor told them it had gained $50,000. They traded up, then traded up again. In 2021 they locked at 2.6 percent, weeks before rates spiked.

They are decent people and they worked hard. But look at the sequence. No good was made. No process was improved. No customer was served. The wealth arrived because they were standing in a particular place at a particular moment, and most of it arrived while they slept.

Then read what the President said at a Cabinet meeting on January 29 of this year.

“People that own their homes, we’re gonna keep them wealthy. We’re gonna keep those prices up. We’re not gonna destroy the value of their homes so that somebody who didn’t work very hard can buy a home.”

It would be easy, and wrong, to make this about one administration. Protecting homeowner equity is neither recent nor partisan — the mortgage interest deduction, the 1997 capital gains exclusion, Fannie and Freddie, and local zoning defended with equal energy by municipalities of every political color. What is unusual is not the position. It is the plainness.

But hold that last clause.

Somebody who didn’t work very hard.

Paul Blumberg described this in 1982. He called it the Las Vegas syndrome: citizens observing that security no longer rests on the old virtues of work and saving but on inflationary currents nobody controls. He wrote it in Inequality in an Age of Decline, 44 years ago.

So the ledger reads as follows. The largest source of middle-class wealth in America is produced by timing, defended as federal policy, and described as desert.

Hold all 3.

II. 1984

In 1984 a woman in Augusta, Georgia bought a house.

No college degree. A single income. A job at a local manufacturing plant. Her realtor told her that in about 5 years she would call to thank him, and she did. She is still in that house. It is worth close to $300,000.

Her son is 38. He graduated into the 2009 job market with a degree from a for-profit college later sued over deceptive practices, cycled through underemployment, went back for an associate’s degree, then a second bachelor’s. He has a good job in information technology and cannot assemble a down payment. He says he feels like he is surviving. He is not certain he will ever own a house.

This is not a story about a lucky mother and an unlucky son. It is proof of feasibility. A single frontline manufacturing wage, in 1984, with no credential attached to it, converted labor into a transferable appreciating asset. The plant did that. Not a degree, not an inheritance, not a windfall. A job on a floor.

Which forecloses the first objection anyone raises to what follows — that wages are simply what the market will bear, that the arithmetic does not permit more. The arithmetic permitted it once.

III. The serious objection

There is a second objection and it deserves a straight answer.

The Wall Street Journal ran a column the same morning as the NPR report, making the point that this complaint recurs. In August 1982, New York magazine put “Downward Mobility” on its cover: You Thought You’d Live Better Than Your Parents Did. Wrong. Fran Schumer interviewed people in their 20s and 30s about their financial distress. Those people are the baby boomers now accused of holding all the wealth. They did fine. So, the argument runs, discount the current alarm.

The pattern is real. So read the 1982 article.

Her subjects: a Manhattan couple on $70,000 joint — roughly $240,000 today — a tenured professor and a manager at Morgan Guaranty, unable to buy. A lawyer in an East Side studio who had expected a second home by 36.

And this, from the husband: “I’m not saying we’re pressed or even badly off.”

Schumer is honest about it. She itemizes his camera and his two IBM Selectrics, and then writes the sentence that ends the argument. Comparing her subjects to the blue-collar unemployed in sunset industries and to the chronically poor, she concludes that the young middle class clearly are not deprived.

Sunset industries. In 1982 that meant steel and auto. It meant Buick City, 6 years before it started dying.

She names the people this essay is about in a subordinate clause and moves on.

So the 1982 anxiety was overblown. Ask why.

Washington’s mother bought her house 18 months after that issue went to press.

One panic, two groups. The professionals were rescued by 40 years of asset appreciation. The plant worker was rescued by a wage that still converted into an asset. Only one of those mechanisms is still running.

IV.

The frontline worker in this country is not asking for the Mraceks’ kind of wealth.

He is being refused the other kind.

V. The floor

Ken Knight taught me the practice at Lansing Grand River, and I did it for the rest of my working life.

At the start of shift you begin at Trim 1, Station 1, and you walk the line to Trim 5. About 160 stations. It takes 90 minutes. You shake every hand. There is no version of it where you skip a station because you are busy.

Most of it takes 4 seconds. Hey, Joe. All good? Move on. Some picks up a thread from the day before — the childcare thing, how did that go. Do it for years and the people on that line come to know the difference between a plant manager who walks for effect and one who walks to hear. So they tell you things.

Remember I told you about my mother in the hospital? I got paid yesterday. I cleared the card. Now I don’t know how I’m covering food this week.

Remember the transmission? I’ve been riding in with a guy down the line. They read the code. It’s $1,500. I don’t have $1,500 and I don’t know if I’ve got a ride tomorrow.

Remember the bathroom? Contractor came and quoted close to $800. I don’t have $800, so I’ll do it myself. I don’t know when. We’ve got one that works and the whole family to get out the door in the morning.

Those 3 are composites, drawn from many conversations over many years. The particulars are changed. The shape is exact.

Now here is the thing I want to report, and it is the only claim in this essay that rests entirely on me.

I made that walk for 36 years, on 2 continents. Thousands of conversations about money.

I do not recall one man ever putting a number on what his work was worth.

Not once. In 36 years of listening, nobody ever told me a figure.

VI. Already spoken for

Read those 3 conversations again for what they share.

A mother’s hospital bill on a credit card. A transmission. A bathroom with one working fixture and a family to get out the door.

Not one of them is an indulgence. Every one is a shock landing on a household with no buffer.

And notice what 2 of the 3 are about. Getting to work. A $1,500 repair he cannot make threatens his ability to keep earning at all. The wage cannot defend the means of producing it.

The wage was not mismanaged. It was pre-committed — spoken for before it arrived, against shocks that had not happened yet and certainly would.

A wage that meets shocks never becomes a stock. It cannot. Every dollar has a claim on it before it clears.

VII. February

Which brings me to the check.

In February about 47,000 GM workers received profit sharing of up to $10,500 for 2025 performance. Ford’s people received up to $6,780 in March. Stellantis workers received nothing, for the first time since the merger formed the company.

I take that instrument apart in the next essay. Here, one observation.

$10,500, arriving once, into a household where every dollar was committed before it landed.

That is not wealth formation. That is triage funding, and calling it anything else is how we have all agreed not to look at this.

VIII. The admission

Now the part I have avoided writing.

A few years ago I was part of the launch of the Chevrolet Traverse and its Buick and GMC sisters. A few weeks in, the trim shop was the bottleneck for the whole plant. We were losing 2 to 3 hours a day, which starved chassis and stopped everything. At roughly 50 units an hour and industry-typical margins for a full-size crossover, every day cost over $1M. Across 2 weeks, $10M to $15M.

And that was the recoverable part. The launch itself was not — advertising, dealer incentives, the press cycle, a sales plan already committed against a date. That money is spent against a window, and the window closes whether or not trim can feed chassis. The true cost of unformed capability was never the units. It was every other function‘s budget, spent against a date that depended on people who had not yet had enough cycles.

We fixed it in 2 weeks. Line balance on the over-cycled stations. Training cycles on third shift. The roof molding station, never run at full rate. Defect containment before chassis — containment first, then irreversible corrective action, engineers and team leaders working it at the station with the people running it.

Every one of those fixes had an address. A shift. A station. Specific people.

I knew what the fix was worth. The number was on my desk.

I could have walked to the roof molding station on a Tuesday morning — I was going to be there anyway, shaking their hands — and told the people standing there what their 2 weeks had been worth against a $10,000-a-unit launch curve.

I never did. Not once, in 36 years.

Not because I was withholding it. The comfortable version of this story is that somebody was indifferent, and indifference can be corrected by hiring better people.

There was no place to put it. No line on the P&L. No field in any system. No agenda item in any meeting I attended in 36 years. No metric anyone was held to. The company was not indifferent to what I heard on those walks — it had no organ capable of registering it. An institution with no receptor cannot be repaired by staffing it with kinder people.

I had the yield numbers on my desk and the hardship in my doorway, every morning, for 36 years.

I never connected the two.

IX. One thing, Monday

If you run a plant, there is one thing you can do on Monday, and it costs nothing.

Publish the number.

The operator who took 4 points out of first-pass yield last quarter does not know what that was worth in dollars. You do. Finance does. He has never seen the figure.

Tell him. Tell the crew. Put it on the board next to the safety cross — in dollars, by station, every month.

It will not build him an asset. It will not survive his next transmission. But it establishes the one thing that was missing from every conversation I had on that line for 36 years: a number he could make a claim against.

I never ran that experiment. I had 36 years and it did not occur to me, which is the confession this essay has been building toward.

So I do not know what happens next. I do not know whether a man who is told what his 4 points were worth says nothing, or says thanks, or finally asks the question that was never once put to me in a plant.

If you run a line and you try it, I would like to know what he says. That answer is not mine to give, and I have no way left to get it.

Part 2 of 3: **The Check That Resets** — why the instrument he already has cannot hold the claim.

Sources

Jennifer Ludden, “The renter-owner wealth gap is wider than ever, as many are priced out of buying,” NPR, 30 July 2026.

Remarks at a Cabinet meeting, 29 January 2026.

Fran R. Schumer, “Downward Mobility,” New York, 16 August 1982.

Jane Shaw Stroup, “The ‘Downward Mobility’ of Struggling Young Baby Boomers,” The Wall Street Journal, 30 July 2026.

Paul Blumberg, Inequality in an Age of Decline, Oxford University Press, 1980.

Harvard Joint Center for Housing Studies, The State of the Nation’s Housing 2026.

National Association of Realtors, median existing-home sale price, summer 2026.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com.

Written with AI assistance. The argument, the judgments, and the floor testimony are the author’s own.

Evidence They Can’t Defend — Essay 7 of 13

In 2007 we stayed at the Oberoi Rajvilas outside Jaipur, in the tented rooms. The night before, at a chowki at the town fair, the five of us had eaten sitting on the floor. By morning it was clear the food had gone badly wrong for all three children, coming out both ends, and my wife and I stood in a tent on the grounds of a hotel in a country I had not lived in for twenty years, doing the arithmetic every parent does. Two weeks of vacation left. A flight back to the United States at the end of it.

An attendant came to the tent. We showed her the children curled on the beds and told her what the night had been.

She said, “Don’t worry.”

I cannot fully describe the look that came with it. I have spent thirty-six years reading faces on factory floors, and I know the difference between a person saying the reassuring thing and one who has already started solving your problem. She was the second kind.

A doctor arrived shortly after, with the same disposition — unhurried, certain, speaking first to us and then to the children. He examined them and got them medication. We cancelled the day’s plans. And then the attendant did the thing I have thought about ever since: she rebuilt the day. There would be a puppet dance on the grounds that evening, she said, and the children should see it. She would make something for them and send it to the tent herself.

We healed. The two weeks were not lost.

Now consider what she actually did, because none of it was in a procedure.

She read that the injury was not the illness. Three children with food poisoning is a medical problem, and the doctor handled it in twenty minutes. The injury was a family a long way from home watching a vacation come apart. She diagnosed that and treated that, and no script told her to — the script ends at the doctor.

Nor did she escalate. She had the standing to summon a physician, alter our itinerary, and commit the kitchen, and she used all three inside ten minutes without asking anyone.

And the doctor carried the same disposition. That is the detail that gives it away. One extraordinary employee is luck. Two in a row, under the same pressure, is a house that forms people on purpose.

I never got her name. Nineteen years later I can describe the look on her face and I cannot tell you what she was called, and that is not only my failure. Nobody wrote it down. The most valuable thing that hotel sold anyone that morning was produced by a woman whose name appears in no record of the transaction — not in mine, and not, I would wager, in theirs.

What the industry spends instead

In 2024, U.S. companies spent an average of $774 per learner on training — down from $954 the year before. Training hours dropped from 57 to 47 per employee. At companies with more than 10,000 employees the figure is lower still: $398 a year to form the human being on whom the entire value proposition depends.

These are not recession numbers. The cuts came during record corporate profits, stock buybacks exceeding $900 billion, and CEO compensation averaging $23 million.

Meanwhile, spending on outside products and services — third-party content, consultants, platforms — surged 23 percent to $12.4 billion. Companies weren’t developing their own people. They were buying off-the-shelf content deliverable at scale with minimal interaction.

Hospitality runs a structural paradox: its entire value proposition depends on human interaction, and it chronically underinvests in the humans. Training is the first line cut under budget pressure, because its absence doesn’t appear on a quarterly earnings report. Eliminate the programme in Q1 and service quality holds until Q3 — by which time the cut has already been reported as margin.

That’s the argument. The attendant at Rajvilas was not a nicer person than an American front-desk clerk. She was a formed person, and forming her cost money that somebody decided to spend.

What the money buys

The Oberoi group runs its own school, and has since 1966. The Oberoi Centre of Learning and Development in Delhi takes roughly eighty to a hundred people a year. Its undergraduate programme runs three years, combining on-the-job training at Oberoi properties with a degree; its postgraduate management programme runs eighteen to twenty-four months. Entry is by group discussion and interview. And the tell is where those graduates end up: the majority of the group’s heads of department and general managers came through it.

Set that against $398 a head and the gap stops being a matter of degree. One organisation decided judgment under load is the product and must be manufactured deliberately, over years, at its own expense. The other decided judgment is a personality trait it can hire for at market rate, and bought a compliance video.

The extraction dimension is what happens to the difference.

Consider the math. A chain with 50,000 employees that cuts training by $200 a head saves $10 million. That flows straight to operating margin, margin drives stock price, and stock price drives executive compensation. The CEO who made the cut collects hundreds of thousands — sometimes millions — through stock-based pay. The workers who lost the training collect nothing.

When Gallup reports that only 31 percent of U.S. employees were engaged in 2024 — the lowest in a decade — the training cuts are part of the explanation. Employees who receive no development perceive no investment. Workers who perceive no investment give no discretionary effort. The engagement crisis is a training crisis is an extraction decision.

The turnover tax

Here is what extraction through training cuts actually costs.

The average cost of replacing an employee is 33.3 percent of base salary. In hospitality, where annual turnover frequently exceeds 70 percent, the math is catastrophic. A hotel with 500 employees earning an average of $35,000, experiencing 70 percent turnover, spends approximately $8.2 million per year just replacing people.

The training that might cut that turnover by 20 points would cost a fraction of it. But training appears as an expense line, while replacement costs are scattered across recruiting, overtime, productivity loss, and quality decline — invisible in quarterly reporting. The extraction model favors visible cost cuts over invisible capability investment. It is not that executives don’t understand the math. It is that the pay structure doesn’t reward it.

Training disinvestment compounds the way investment compounds, in reverse. Five years of progressive development arrives at judgment. Five years of the same compliance video arrives at the exit. Multiply that across an industry and you get American hospitality — high turnover, chronic understaffing, an endless cycle of hiring and replacement — which the extraction model calls labor market dynamics.

It is not an India story

The obvious objection is that Rajvilas is a different country — a market where hotel work is a career and formation is cheap to fund. That objection would be more comfortable if an American chain had not already proved otherwise.

Marriott has run for more than ninety years on a sentence from its founder, J. Willard Marriott: take care of the associates and the associates will take care of the guests, and the guests will come back again and again. In March 2020, with worldwide revenue per room down about ninety percent, Arne Sorenson announced on video that he and Bill Marriott would take no salary for the rest of the year and his executive team would take fifty percent cuts.

Be honest about what followed, because the extraction case is not that simple. Marriott still furloughed the majority of its above-property staff. Sorenson called that heart-wrenching and did it anyway. What he did not do was treat the associate as the first line to cut and the last to explain to — and when he died the following February, the company’s response was to fund a hospitality centre in his name aimed squarely at building leadership talent in the industry.

That is not Jaipur, and it is not charity. It is an American public company under the worst conditions in its history deciding that the person nearest the guest is the product. The extraction model reverses the order: cut the people, capture the savings, report the margin, collect the bonus — and then wonder, three quarters later, where the service went.

What I do not know

I want to be honest about the edge of this, because the diagnosis is the easy half.

I do not know what her formation actually contained. I watched the output and I can describe the school in outline, but I cannot tell you which part of those years produced the woman who understood that the vacation was the injury. Nobody has taken that apart. If somebody had, we could build it elsewhere.

I do not know whether it survives a market running 70 percent annual turnover. Whether you can form a person faster than that door revolves is a real question, and anyone who tells you they know is selling something.

And I do not know who pays for the first one. A single American hotel that funds real formation trains people its competitors hire for free. That is the arithmetic that killed the apprentice bench in manufacturing, and I have never seen a firm beat it alone.

What I am certain of is narrower and harder to argue with. That morning in Jaipur, the thing that saved our two weeks was not a policy, a script, or a technology platform. It was a formed human being with the standing to act, and somebody paid to make her.

So if you run a floor, a property, a ward, or a shift — tell me about the person on your team who did the thing no procedure specified. What was in them that put it there, and who paid for it?

And if you can remember what they did but not what they were called, ask yourself the same question I have been asking since 2007: who was supposed to write it down?

Sources: Training Magazine, “2024 Training Industry Report” (November 2024); Statista, U.S. Training Expenditures 2012-2024; Gallup, “State of the Global Workplace 2024”; EPI, CEO Pay data page (2025); SHRM, employee replacement cost data; Research.com, “2026 Training Industry Statistics” (January 2026); The Oberoi Centre of Learning and Development — programme structure, intake and admissions process, OCLD/Oberoi Group; Marriott International, 2020 Letter to Stockholders (founder’s principle; Arne M. Sorenson Hospitality Fund); Marriott International, video address to associates, March 19, 2020 (salary and executive pay reductions); Marriott International Q1 2020 Form 8-K (RevPAR decline); Forbes, “Marriott CEO Arne Sorenson On The Future Of The Hospitality Industry, Masks, And Furloughs,” July 13, 2020.

The Rajvilas account is the author’s own, from a stay in 2007.

Next: The Jobsite That Got Financialized — how subcontracting chains, worker misclassification, and $50 billion in annual wage theft turned construction into an extraction architecture.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com.

Written with AI assistance. The argument, the judgments, and the floor testimony are the author’s own.

In response to: “Software engineers are facing an ‘identity crisis bordering on depression,’ Menlo Ventures partner says” — Thibault Spirlet, Business Insider, June 22, 2026.

Deedy Das, a partner at Menlo Ventures, looked at what AI coding tools are doing to engineering teams and saw two kinds of people. The first he calls lazy — engineers who let the model write the code, answer the questions, file the updates, and engage as little as the system will tolerate. The second he calls craftsmen — the experienced ones who now spend their days reading, reviewing, and repairing the flood of machine-written code the first group ships. Most engineers, he says, are living through an identity crisis bordering on depression. The craft they loved is dead.

Keep craftsmen. Put down lazy.

Here is the argument. The divide Das describes is real, but he has named half of it wrong, and the wrong name hides who is actually on the hook. The so-called lazy engineer is not lazy. He was never formed — and then he was handed a power tool and paid to pull the trigger. He is doing, precisely and rationally, what his employer measures. Somewhere upstream a company decided that the way to capture AI’s promised productivity was to count adoption — tokens spent, tickets closed with the assistant, the percentage of commits that touched the model. And the moment you measure the tool instead of the work, you manufacture the crutch. You don’t get craftsmen at scale. You get a building full of people optimizing the number on the wall. The craftsman doesn’t survive that system; he subsidizes it. He eats the review burden, the defect hunt, the silent cleanup, unpaid and unnamed, while the dashboard upstairs turns green. The craft is not dead. The unformed relationship to the craft is collapsing — and nobody built these people the other kind. That’s the argument.

I watched this exact movie play out in steel and torque, twenty years before it arrived in pull requests. I was running General Motors’ Lansing Grand River plant.

GM had done the expensive, sensible thing across all its plants: computer-controlled DC electric nutrunners, every critical fastener run by a tool that monitored its own torque and reported up into an electronic error-proofing system. State of the art. Corporate sent torque auditors to confirm each plant was using the tools, and using them correctly. The machines were genuinely good. And if you wanted a number that proved the capital had paid off, it was right there on the wall: compliance, green across the board.

Here is what the green did not tell you. The monitor watched the tool, not the joint. Under the wrong run of conditions it would fail to record a missed fastener — and worse, if a tool could reach a little too far, an operator could run a bolt down on the wrong vehicle entirely, and the system, satisfied that a rundown had happened inside its window, would call it good and release the line. The car that actually rolled out the door short a bolt was logged as fine. Months later a dealer would call about that very car, you would pull its build ticket to see what went wrong, and the ticket would swear nothing had — because the fault had quietly landed on the record of the car ahead of it or behind it in the sequence. The system was confidently, auditably green about a defect it had already shipped to a customer.

So who actually caught those? Not the system. The formed operator — the one who had spent years learning what a seated joint feels like in the hand before any light confirmed it — would look at a green light and still know something was wrong, and stop. The operator who had never been formed trusted the light, because trusting the light was the whole of what anyone had taught him. Same tool. Same green light. Opposite outcome. The variable was never the tool. The variable was the human the tool had been handed to, and whether the institution had bothered to form him before it armed him.

And the floor reorganized itself around that gap without anyone deciding it should. We measured compliance at the gun and caught the real defects downstream — which is to say, we caught them in the hands of our best people, who pulled bad work back all shift and mostly said nothing, because catching what the system missed had quietly become the definition of being good. We called the green dashboard a productivity win. It was a win. It was being paid for, in full and in silence, by the most formed workers in the building.

Now read Das’s two columns again. The lazy engineer ships. The craftsman cleans. The dashboard is green. The craftsman is depressed.

The white-collar version is worse in one specific way, and it is the way that produces the depression Das names. On the factory floor, scrap is visible. A cracked casting is a cracked casting; eventually it falls on the floor and somebody trips on it. Software defects are quieter and they compound — the bad abstraction that runs, the plausible function that is subtly wrong, the test that passes for the wrong reason. So the craftsman in code carries a heavier and lonelier load than the craftsman in steel: he is the last line between a confident machine and a customer, and his employer has not named that job, does not measure that job, and in many shops is actively measuring the opposite of that job. He was hired to build. He has been quietly conscripted into verification — into being the human who can tell when the machine is wrong — and nobody asked him, promoted him, or paid him for the reassignment. That is not a craft dying. That is a craft being looted to cover for the absence of formation in the person at the next desk.

So the word matters. If you call the unformed engineer lazy, you have located the failure in his character, and you have let the institution off entirely. You have written a morality tale: some people have grit, some people don’t, AI just sorted them. It is a comfortable story for everyone who runs the place, because it requires them to change nothing. The truer story is harder and lands on the people with power: we never formed him, we measured the tool, and then we acted surprised. Formation is not a virtue the worker either possesses or lacks. Formation is a thing an institution does to a person, deliberately, over time — or fails to do, and then blames the person for the gap.

Which points the way out, and I want to be honest about how far down that way I can actually see.

The direction is not subtle, and Das half-named it himself. Stop measuring the tool. Measure the work the tool was supposed to serve, and measure whether the person is being formed into someone who can wield it — which means someone who can tell, fast and unaided, when the machine is confidently wrong. Pay for that judgment instead of strip-mining it from whoever happens to have it.

And notice the word Das reached for when he wanted to name the people worth protecting. Not experts. Not seniors. Craftsmen. He reached, by instinct, for a guild word — and the word carries its whole world in with it. A craftsman is not a personality type you either have or lack. A craftsman is an output. He is what comes out the far end of an institution built to produce him: the apprentice, the journeyman, the master, and the standard held above all three. You cannot have craftsmen and skip the thing that makes craftsmen, any more than you can have a harvest and skip the field. The Germans called that thing the Zunft and kept it, in one form or another, for six hundred years — a power tool placed only into a hand that had been prepared to hold it, inside a relationship that did the preparing. We did not lose the craftsmen first. We dismantled the field, and then expressed surprise at the harvest. We retired the word the moment software made everyone feel they could skip straight to master.

But here is the genuine open edge, and I don’t have it solved. A guild could form an apprentice because the chisel did not change. You could spend seven years learning a tool that would still be the tool when you finished. The model under your hands now is reinvented quarterly; the thing you formed someone into using last spring is deprecated by autumn. So what, exactly, is the durable formation when the tool itself will not sit still long enough to be mastered? What is the apprenticeship for a craft whose instruments are replaced before the apprentice’s hands stop shaking? I have one floor where I think I see the shape of an answer, and no proof at all that it scales. That is not false modesty. That is the actual state of my knowledge, and if you are an engineer living inside this right now, your view from the floor is better than mine.

So I’ll put the chisel down and hand it over. Das gave us the right word for the people we should be protecting. He gave the wrong word to the people we failed. Fix the second word, and the question that’s left is the one worth arguing about — not who is lazy, but who is supposed to do the forming, and what does the forming even contain when the ground keeps moving.

Tell me where I have it wrong. I mean that as a request, not a flourish.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com.

Written with AI assistance. The argument, the judgments, and the floor testimony are the author’s own.

I sat down with a group leader once to talk about his downtime numbers. His name was Ethan.

He was soft with his people in a way I had not seen much of on a floor — patient, unhurried, the kind of man who would stand next to somebody having a bad shift instead of walking past. He was slowly learning to lead, and slowly is the only speed that works for that. But the line was not running and had not been running for a while. There is a surgery that job requires — the hard conversation, the unpopular call at two in the afternoon, the standing of ground — and he would not do it.

Ten minutes in, it stopped being a conversation about the line. It was a conversation about his heart not being in it.

So I asked him where his heart was.

He told me about the weekends. Every weekend we had asked him to work was a weekend he had planned to fly, and he had been counting them — the hours he needed for a certificate, and how many of them we had taken.

That is when I said stop.

We spent the rest of the meeting on the other side of the table. What it would cost. That he would have to quit outright. That he would move back to his parents’ place in Chicago to cover his living while he flew. What the certification sequence was and what order to take it in. We charted the whole thing out.

He’s on his way to becoming a pilot today. As he earns the wages to paying for lessons.

Here is the argument. Ethan was keeping a capability account the entire time he worked for me. A real one — hours logged, signed by an instructor, building toward a rating issued by somebody who was not his employer. My plant was drawing that account down, weekend by weekend, without ever knowing it existed. We were not taking his time off. We were taking the only hours in which he was becoming something, and we could do it without argument because those hours had no standing in our accounts. Meanwhile the account we should have been keeping — the leadership he was slowly and genuinely building on my floor — did not exist at all. Two ledgers in one man. He kept the first himself, in a book with his name in it, because his trade requires that. Nobody kept the second, because mine does not. Aviation solved the ledger and never solved the financing. Medicine solved the financing and never solved the ledger. The floor and the ward have neither. Everything I am proposing has already been built — just never in the same building.

That’s the argument.

What a logbook is

It is a better artifact than anything industry has built for this.

It is individually owned. Not held by the employer, not by the school. The pilot carries it. When he changes airlines he loses nothing — the exact opposite of what happens when an operator changes plants.

It is continuously confirmed. Every hour is logged and the significant ones are endorsed by a certificated instructor who signs his certificate number next to his name. The endorsement is a person putting his standing behind an assessment — a Meister function, whatever aviation calls it.

It is issued through a firewall. The rating comes from the FAA, or a designated examiner acting under it. The employer trains and benefits, but does not certify. The party that captures the value does not get to grade the paper. Medicine has the same firewall — the board is not the hospital — and neither trade would function without it.

It has a half-life. Three takeoffs and landings in the last ninety days or you don’t carry passengers. Six instrument approaches in six months or you lose currency. Medical exam on a clock. Recurrent training, line checks. Capability is treated as perishable, because it is. No other American credential admits this out loud. A degree from 1994 still reads as a degree.

And it is modular. Private, then instrument, then commercial, then multi-engine, then a type rating — capability accreting in increments that can each be signed for.

A three-year guarantee reads to an employer as a monolith — three years of somebody else’s risk, signed blind. Broken into confirmable increments, it becomes a sequence of things an operations manager can agree to on a Tuesday. Nobody signs a monolith. People sign ratings.

The honest ledger on aviation

The weakness in the aviation model is precisely the strength of the medical one.

Civil aviation finances formation on the trainee’s back. The path to the right seat runs through fifteen hundred hours the pilot largely pays for himself, and the debt keeps capable people out of the trade for reasons that have nothing to do with capability. Worse, the person signing those hours is often a young instructor building time toward the job he actually wants — the least experienced person in the building, teaching. A greenbeard, not a graybeard. Medicine avoided both with a public payer and attendings who are senior practitioners, not people passing through.

So the synthesis is: aviation’s ledger, medicine’s financing, and a Meister from neither — from the floor, from the guild tradition industry had and let go.

The version at the desk

If you work at a desk, you may be reading this as somebody else’s problem. It isn’t.

Your credential is a résumé and a profile page, both self-attested. No examiner, no endorsement, no certificate number, nobody with standing signing anything. When a firm cuts three hundred people, the twenty-year project manager and the two-year one reach the market carrying the same document, distinguishable only by the name of the last employer — a proxy for a company’s reputation, not a record of a person.

The desk has a logbook problem too. It just has a nicer font on it. The real difference between the desk and the floor is not whether formation gets recorded. It is how long you can go before anyone finds out it wasn’t.

Three columns

Line them up and the pattern is plain.

The resident is financed and not recorded. Three guaranteed years, a public payer, real teachers — and what he carries out is a certificate, while the milestones live in systems he doesn’t own.

The pilot is recorded and not financed. He owns everything, he can prove everything, and he paid for most of it himself.

The aide and the operator are neither. No guaranteed runway, no curriculum, no signature, no account, no half-life, because there is nothing to have a half-life.

Ethan crossed from the third column to the second. He did not escape the guild model — he escaped intoone. He left the building where his formation was invisible for the building where formation is the entire basis of employment. He paid his own way across, and the only institutional help he got was a plant manager who happened to ask a second question.

What he left behind

Here is the part I have not been able to put down.

Ethan’s logbook has his hours in it. It does not have the two years he spent learning to stand next to a struggling operator instead of walking past. That was the harder thing, and it was the thing I watched him build, week over week — the frontline judgment that decides whether a team gets used or gets ground down, the scarcest capability on any floor in the country.

I read him right in the end. It made no difference to that. Reading a man correctly and having some way to write down what he has become are two different problems, and I only solved the first one, in one conversation, by luck of temperament. None of it followed him. He started the new trade at zero. The plant recorded nothing either, so the loss appears in nobody’s accounts. It just quietly costs us, forever, in a line item no one has ever written.

Wally Vinton in the trim shop could find a fault in a wiring harness by feel before the test station found it. Ramon Hernandez was the best team leader I ever had. Ask me what he knew and I could talk for an hour and not hand you a document. Those two men’s formation is on somebody’s balance sheet right now, swept into goodwill under ASC 805, carried at a number, in nobody’s name. The accounting profession already decided, under ASC 842, that an asset a company controls and benefits from but does not own belongs on the books. We have the template. We have simply never pointed it at the person.

The blank page

What actually saved that conversation is that I happened to ask a second question.

Nothing required me to. No policy, no process, no field on any form asks where a man’s heart is, and nothing would have caught it if I had let the meeting stay about downtime. A different manager on a worse day ends it at minute ten and is entirely within his rights. Ethan goes out the gate with a bad number behind him and no plan in front of him, and maybe finds his way to the trade five years later and maybe doesn’t. What worked for Ethan was a manager going off-script. That is not a system. That is weather.

An account would have made it a system. If his hours had been visible — if formation toward a trade were something an employer could see rather than something a young man mentions at the worst possible moment — the weekends become a scheduling conversation in year one instead of an exit conversation in year three. And if the leadership he was building had been confirmed by someone with standing to confirm it, he would have carried it out with him whether he stayed or went.

I know the shape of the cure. First employer, three guaranteed years, a curriculum across the runway, a graybeard who teaches and confirms, an account the worker owns with a half-life on it. Aviation proves each piece works, because each piece is already running somewhere.

What I don’t know is who signs. Who is the certificated instructor for a group leader learning to lead? What would that endorsement even say? What is the equivalent of ninety days and three landings for judgment — and if capability has a half-life, what exactly decays and what doesn’t?

If you have ever formed somebody on a floor and had no way to write it down, you have thought about this longer than I have. Tell me what you would have put in Ethan’s logbook.

That’s the Long Game.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com.

Written with AI assistance. The argument, the judgments, and the floor testimony are the author’s own.

Evidence They Can’t Defend — Essay 6 of 13

The doctors knew exactly what to do. That was never the problem.

A woman was bleeding to death from her liver, one day after giving birth. Her heart had already stopped once, and the team had brought her back. They had a plan — an embolization coil, a small device that goes in and closes the bleed. Standard. The kind of thing a hospital that delivers babies simply has on the shelf.

The hospital did not have it. Weeks earlier, the manufacturer had come and taken its coils back, because the hospital’s owner had not paid the bill. Some of the staff had feared exactly this and said so, out loud, to the people above them. The warning did not reach the operating room in time. She was moved to a second hospital, arrested again about an hour later, and could not be brought back.

Her name was Sungida Rashid. She was thirty-nine. Her daughter was one day old.

The people who took the coil away were not doctors. They had never penned a note or walked a patient. They owned the hospital the way a strip-miner owns a mountain — for what could be pulled out of it. The firm was Cerberus Capital Management. The hospital was St. Elizabeth’s in Boston. And by the time Sungida Rashid bled out, Cerberus and the men who ran the system it built had already pulled roughly $1.3 billion out of it. One of them bought a $40 million yacht.

That’s the argument. Here is how it was done — and here is the one thing they could never load onto the boat.

The Extraction Architecture

In 2010, Cerberus bought a nonprofit Catholic hospital network in Massachusetts — Caritas Christi, a mission system that existed to treat the poor. They converted it to for-profit, renamed it Steward, and promised to expand care.

Then they engineered the extraction. In 2016, Steward sold the land under its hospitals to a real-estate trust, Medical Properties Trust, at an inflated price, and leased the buildings back at rents it could never sustainably carry. The sale threw off a wave of cash. That cash did not go into the hospitals. Steward’s own audited statement shows $789 million paid out in dividends in 2016 alone — roughly $682 million to Cerberus, about $73 million to CEO Ralph de la Torre, the rest split among insiders. Nearly $800 million, in a single year, pulled out of a system serving low-income communities.

The yacht came after. Two private jets. A private suite at the arena in Dallas. And the hospitals, meanwhile, could not pay for embolization coils.

The Human Price

The coil that would have saved Sungida Rashid was made by a company called Penumbra. Court filings show Steward owed Penumbra about $2.5 million, and the company had sued over the unpaid bills weeks before she died. She was not an accident of the system. She was the system working as designed — returns flowing up, risk flowing down, until the risk reached a delivery room.

Senator Elizabeth Warren later walked Cerberus’s founder through what his ownership had done to one hospital, Quincy Medical Center: before Cerberus, a full medical center — surgery, specialty care, urgent care, a VA clinic; after, an emergency room and nothing else. Warren’s arithmetic on the decade was that Cerberus investors took something like 23 percent a year while nurses’ pay barely moved. Two years after the takeover, nurses in Massachusetts had filed more than a thousand unsafe-staffing complaints. When the Senate subpoenaed de la Torre to answer for it, he refused to appear, and was held in criminal contempt — the first time in decades.

What They Could Not Take

Here is the one thing none of them could put on the yacht, because it was never theirs and never sat on any page they could sell.

My sister is an ophthalmologist — Dr. Padma Paul, a professor at Christian Medical College in Vellore, a hospital founded in 1900 by Ida Scudder to treat people the world had written off. Her department still stands on a road named for its founder. In 1948, a CMC surgeon named Victor Rambo performed sixty-nine cataract operations in a single day in a field near Vellore, and the model of carrying sight out to the villages spread across India. Seventy-five years on, my sister drives that same mission down the same roads — a fully equipped ophthalmic van going out into the hill villages to catch disease early and operate on cataracts in place, at a fraction of the cost, so that a farmer does not go blind for want of a bus fare. I watched that formation happen over decades. It is thirty-odd papers and hundreds of citations and a van in the Jawadhi Hills, and none of it appears as an asset on any ledger anywhere.

And I am watching it happen again. Our son is a haematology-oncology fellow. When we call him at night, we ask the small things first — did you eat, did you get to the gym. He tells us how many notes he still has to write before he can sleep, and then he walks us through his patients, and we understand almost none of it. We don’t need to. We can hear that he wants to live inside it — the agony and the ecstasy of becoming the doctor who will one day be the coil in the room. That is what formation sounds like while it is still happening: a young man too tired to eat, unwilling to put down the patients he is learning to carry.

This is the accounting fact underneath the whole series. When one company buys another, the standard — ASC 805 — folds the assembled workforce into goodwill. The formed people are not permitted to stand as a named asset on the books. So my sister’s forty years, and the nurses who filed a thousand complaints, and every hour my son spends tonight learning a patient — that capability lives in the hands and shows up on no balance sheet. Which is precisely why the extractors could take it for free. You cannot loot the real estate; it is booked, and worth selling. You cannot loot the coils; someone will come repossess them. You loot the formation, because it was never on the books to defend.

Cerberus took a mission hospital and asked how much can we pull out before it falls. Vellore took the same kind of mission and asked how far into the hills can we carry it. Same institution at the start. Opposite question. One of them ends with a van full of restored sight. The other ends with a woman bleeding out for want of a device that had been carried away.

Not an Accident — a Model

Steward was the loudest case, not the only one. Research in JAMA has found more adverse events at hospitals after private equity buys them, and the reason is structural: PE runs on a three-to-seven-year exit, and a hospital is meant to serve a community indefinitely. Optimize a hospital for a five-year sale and you cut the things whose absence won’t show for seven — training, maintenance, staffing depth. By the time the harm surfaces, you have already gone.

It does not have to run this way, and the best hospital in the country proves it. Mayo Clinic is a nonprofit. It reinvests its surplus into care, research, and its people; its nursing retention is among the highest in the industry; its outcomes lead nearly every category. Mayo has never been taken by private equity, never sold its buildings to fund a dividend, never bought its CEO a yacht. And it beats every extraction-owned chain on every measure that matters.

The Fix

The reason extraction keeps winning is not that it is smarter than care. It is that care is invisible on a balance sheet and cash is not. As long as a nurse’s judgment and a surgeon’s formation sit on no ledger, they will always be the cheapest thing in the building to cut and the last thing anyone is punished for cutting.

So the fix is not sentiment. It is to make the asset legible — to give a worker’s formation an account that follows them and registers when it is destroyed — so that stripping it finally costs something on the one page these men actually read.

I will not pretend I know the finished shape of that ledger. How you value a nurse’s tenth year against her first, who is fit to sign, what happens to the balance when a hospital closes and forty years of formation disperses into other people’s buildings — those are open questions, and the people who can answer them are working nights in those buildings right now, not sitting in think tanks. What is not open is the direction. As long as formation is free to take, it will be taken. Name the asset, and the free lunch ends.

What Extraction Costs

Cerberus made its 23 percent a year. The system it built left behind $9.2 billion in liabilities, $290 million in wages and benefits it never paid its own workers, five closed hospitals, and twenty-four hundred people out of a job.

A mission hospital was converted and drained for the one asset the rules refuse to name, by men who never penned a note or walked a patient — and a new mother died in the gap they opened. That’s the argument.

One yacht.

Next: The Hotel That Stopped Training — how the hospitality industry’s race to cut training budgets created a service-quality crisis while executive compensation compounded.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com.

Written with AI assistance. The argument, the judgments, and the floor testimony are the author's own.

Sources: Boston Globe, Jessica Bartlett, “Steward’s Medical Devices Were Repossessed. Weeks Later, a New Mother Died” (January 2024); CBS News, “A new mom died after giving birth at a Boston hospital” (February 2024); WBUR, “Mass. alleges Steward jeopardized patient safety while paying off investors” (May 2024); Private Equity Stakeholder Project, “The Pillaging of Steward Health Care” (2024) and “One Year Later” (2025); OCCRP / Boston Globe Steward investigation (Pulitzer entry); SHC Creditor Litigation Trust $3.4B filing (November 2025); Sen. Elizabeth Warren, questioning of Stephen Feinberg (February 2025); Senate HELP Committee proceedings; Paul P, Kuriakose T, et al., “Prevalence and Visual Outcomes of Cataract Surgery in Rural South India,” Indian J Ophthalmol 2019;67(3):386–390; Friends of Vellore, CMC mobile ophthalmic unit.

Source: “Introducing the Fortune 500,” Matt Heimer, Fortune, June 3, 2026. Ranking and figures: fortune.com/ranking/fortune500/2026/.

The shareholders get a column. The workforce does not get a clause.

Every June, Fortune publishes the closest thing American capitalism has to scripture, and this year it opens on a coronation. After thirteen years Walmart has been knocked off the top, and Amazon sits at number one with revenue of $717 billion — up twelve percent in a single year, from a company that entered this list at number 492 barely two decades ago. Only four names have held the top spot in seventy-two years.

Matt Heimer, introducing the list, draws the lesson plainly: you cannot succeed at that scale without persuading many, many customers that you have got the goods. What follows is a hymn to that idea. The customers are the heroes. The CEOs are the heroes. And the people who actually make the goods appear nowhere in the chorus.

I want to start with the smallest thing in the magazine, because it gives the game away.

For each company it spotlights, Fortune prints a little box of vital statistics. Four numbers. Revenue. Profit. Year founded. And — this is the literal fourth line, not my editorializing — twelve-month return to shareholders. Four numbers chosen to tell you what a company is, and a full quarter of that definition is reserved for how well its owners did. There is no line for how many people it employs. No line for what it paid them, what it taught them, how many it kept, how much capability walked out the door or was built inside it. The shareholders get a column. The workforce does not get a clause.

That box is the argument in miniature. We have built a magnificent instrument for measuring what an enterprise captured and no instrument at all for measuring what it formed — and an economy grows whatever it agrees to count. So capture compounds, year after year, while the capability that produced it stays off the page, until the day it has worn thin enough that a door plug leaves an airplane over Oregon. The answer is not a better magazine. It is a second ledger: a Capability Account that ranks a company by the people it formed, not the dollars it took. That’s the argument.

The pattern holds one level up. On its own ranking page, Fortune sums up all five hundred companies in a single sentence: combined revenue of $21 trillion, profits of $2.1 trillion — two-thirds of American GDP — while employing 30.5 million people worldwide. Read where the grammar puts the human beings. Revenue is the subject. Profit is the subject. The 30.5 million are a while clause — a thing happening in the background while the real action took place.

Here is the question the Fortune 500 is built not to ask: where did the $717 billion come from?

Not the customers. They brought the dollars, yes, but a dollar is only the receipt. Revenue is crystallized labor cashed out — capability formed in people over years nobody invoiced. By the time it shows up as $717 billion in June, it has already been paid for. The forming happened upstream, off the page, uncounted.

Two people, both real, both on the record, stand in for the 30.5 million.

The first is Emily Guendelsberger, a journalist who took a holiday picking job at an Amazon fulfillment center outside Louisville and wrote it down in On the Clock. She walked up to sixteen miles a shift, and the scanner in her palm counted down the seconds she had left for each item before she fell behind the rate. That device is the Fortune 500 stat box strapped to one human being. It measures her seconds. It cannot measure what makes a good picker — the slowly built map of a twenty-five-acre warehouse, the body that learns where thirty million items live and the shortest line between them. It counts the easy thing and is blind to the thing that took weeks to form.

The second is Abe Collier, a delivery driver who walked through a single stop in an open letter to Jeff Bezos. A house with no number, so he doubles back to read it off the curb. A whistle at the gate to check for a dog. A spot on the porch chosen so it cannot be seen from the street. A dozen small judgments per stop that the route software hands him as a dot on a map and never makes for him. Multiply Guendelsberger and Collier by 30.5 million and you have the asset that built the list — and the one asset it declines to measure.

I do not have to take their word for it. I spent two decades on floors where that asset was the whole game. At Lansing Grand River, a team leader named Ramon Hernandez could feel a line drift out of true before any gauge on it moved — could hear it, almost — and pull it back before a single bad part reached the next station. No stat box ever held what Ramon knew. It lived in his hands, and it walked in and out the door with him every shift. That is the thing the Fortune 500 cannot see: formed in one man, and multiplied by thirty million.

Now turn to Boeing, ranked 47th, and watch what happens when that asset is allowed to rot.

Fortune tells Boeing’s story as a comeback and, in its own account, names the disease exactly: the company, it writes, increasingly put profits over quality. That is the formation layer collapsing, in four words. Two 737 Max crashes killed 346 people. A door plug blew out over Oregon. None of that was a pricing failure or a marketing failure. It was a capability failure — decades of letting the floor knowledge that builds a safe airplane erode while the financial number was optimized, until the number and the airplane came apart in mid-air.

And how does the magazine frame the recovery? As the energy of a new CEO. The same move runs through every comeback in the issue — Intel, Macy’s, all of them “tapping the energy of a new CEO,” as if the capability that saves a company were a trait of the man at the top rather than something painstakingly rebuilt, or not, in ten thousand people below him.

Intel’s own stat box closes the case. Profit last year: negative $267 million. Twelve-month return to shareholders: plus 494.9 percent. The company lost money and the one number Fortune tracks about people went up fivefold. That is the entire worldview in two figures.

So I read the new Fortune 500 the way I read every one: a magnificent, immaculately reported answer to a question we should stop asking first. Who captured the most? is a real question. It is not the same question as who is the economy for? — and we have let the first stand in for the second so long that most readers no longer notice the swap.

Ask the second question and you need a different instrument. Call it the Capability Account: a ledger that ranks an enterprise not by the revenue it extracted last year but by the human capability it formed — the skills built, the apprentices carried from raw to journeyman, the tacit knowledge deposited into people who keep it whether they stay or go. By that measure the leaderboard reshuffles completely. Some companies near the top of Fortune’s list would sink, having harvested capability that other institutions formed. Some firms too small to make the 500 at all would lead it, because forming people is what they are actually for.

What I do not have is the unit of account. Nobody does yet. What a person can do the day they walk out is one candidate. The answer may be something no one has proposed. That work needs people who run plants and payrolls and residency programs more than it needs people who write about them.

None of this makes Fortune a villain, and that is the harder point. It is a precise instrument doing exactly what instruments do — making visible only what it was built to measure. The fault is not dishonesty. It is that we have spent seventy-two years sharpening the one lens and never grinding the other, so that the people who form the capability — the only asset that finally decides whether the plane flies — show up, if at all, in a subordinate clause.

The customers are already counted. The shareholders have a column of their own. The 30.5 million are right there in Fortune’s own sentence, formed and paid for and waiting for a line that does not yet exist.

Until we build that line, we will keep mistaking the receipt for the work, and keep optimizing the one number that can never tell us the only thing worth knowing: who the economy is actually for.

It is for the people in it. Or it is for nothing.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com.

Written with AI assistance. The argument, the judgments, and the floor testimony are the author's own.

Source: “Yale asked the right question. Now the rest of higher education owes an answer,” Steve Beard, Fortune, April 22, 2026.

Steve Beard published a commentary in Fortune that does something rare in higher education discourse: it takes Yale’s year-long faculty investigation seriously and asks the rest of the sector to answer it. He is right to. Ten tenured Yale professors spent a year diagnosing why public confidence in higher education has collapsed from 57% to 36% in a decade, and their answer — that the sector has tried to be all things to all people and lost the plot — deserves engagement rather than deflection.

Beard’s answer is that higher education must be measured on outcomes, not intentions. He points to his own institutions — Chamberlain, Walden, and the rest of the Covista system — which graduate 24,000 healthcare professionals a year and post a 97% first-time residency match rate. He invokes the Carnegie Opportunity Colleges designation. He makes the case that access plus transparent outcomes can rebuild trust.

He is right about access. He is right about transparency. And he is still one layer away from the actual problem.

Because a residency match rate measures whether the credential cleared. It does not measure whether the person became capable. These are not the same thing, and the conflation of the two is the rot at the center of American higher education. We have built a system in which the institution carries the brand and the student carries the debt, when the account should run the other way. Every individual should have a Capability Account — a lifetime ledger, parallel to a Social Security Number, into which verified formation is deposited by schools, apprenticeships, and employers over the full arc of a working life. The credential is a claim made by the institution about itself. The Capability Account is an asset held by the person, about the person, signed by the people who helped form them. That’s the argument.

I spent the first two decades of my manufacturing career watching the gap between credential and capability up close. At General Motors Lansing Grand River in the early 2000s, we won the J.D. Power Gold Award — the first American plant ever to do so on a new launch — and we did not win it because of the degrees on the wall. We won it because Wally Vinton in the Trim Shop knew more about how a wiring harness wanted to be routed than any engineer with a diploma, and because Dennis Boutwell, my first supervisor hire, had been taught to see a line the way a physician sees a patient. Neither man had been issued a credential that captured what they actually knew. The plant ran on formation the accounting system refused to recognize.

When I moved to Chennai to run Royal Enfield, the gap widened. We had engineers from the IITs who could solve any textbook problem and could not, in their first six months, tell you why a weld was cracking at the heat-affected zone. We also had shop-floor technicians who had never cleared Class 10 and could diagnose a crankshaft imbalance by sound. The credential said one thing. The capability said another. I say this with some personal weight — I did not get into the IITs the way my uncles had, and my mother cried the way she might have if her son had died. The credential was that weighty in our family’s accounting. Two decades later I was running a motorcycle company, and the men who were teaching me how the machine actually wanted to be built had never been admitted to any institution at all. We grew the company from 50,000 units to 113,000 and twentyfolded profit — and we did it by building parallel formation systems inside the plant because the outside system couldn’t be trusted to deliver.

I tell you this not as nostalgia but as evidence. The problem Yale named and Beard is responding to is older and structurally deeper than either acknowledges.

Here is what Beard’s frame misses.

When Yale’s cost of attendance hits $94,425 a year against an American median family income under $84,000, and when a quarter of federal student loan holders are in default, we are not looking at a pricing problem. We are looking at an accounting failure. The institution has capitalized a credential onto the student’s personal balance sheet — at full sticker — without underwriting whether the cash flows that credential is supposed to generate will ever arrive. The nursing, public health, and environmental science graduates Yale singled out are not victims of a market miscalculation. They are carrying a liability the institution booked as its own asset.

A 97% residency match rate is a better number than most of higher education can produce. I want to say that plainly. Covista’s medical schools are doing something real. But the residency match is the credential clearing the credential. It tells you the student passed through the gate. It tells you nothing about what they can do at the bedside in month six of intern year, when the chief resident is asleep and the patient is crashing and the question is not what did you learn but what have you been formed to notice. That is a different register. That is Vocational Value and Contribution Value, not Accreditation Value. The American system measures the third and pretends it has measured the first two.

This is the move Beard stops short of. He is right that access without outcomes is a broken promise. But outcomes measured as did the credential clear is itself a broken measurement. The honest question is whether the person is capable — verified by people who would stake their name on the verification — and whether that capability compounds through their working life or atrophies.

Which is why the Capability Account matters.

Imagine every American receives, at birth, a ledger. Not a score. Not a transcript. A capability ledger — an asset account in their own name, structured the way the Germans structure their Ausbildung system and the way the medieval guilds ran before industrialization severed apprenticeship from accreditation. Into this ledger, verified formation gets deposited. A high school that teaches a student to braze deposits a verified capability, signed by the instructor and countersigned by a chamber. A community college that certifies a phlebotomy technician deposits another. An apprenticeship with a master electrician deposits a third. These are not course credits. They are attested capabilities, reviewed by bodies that would lose their standing if they signed falsely.

The Ausbildung threshold is the floor. Below it, the labor market does not open. You cannot be hired into formation-sensitive work without having cleared the basic capability bar. This is not credentialism. It is the opposite of credentialism. A credential says the institution vouches for itself. An Ausbildung says a master vouches for the person, and the chamber vouches for the master.

Employers then hire against the account balance. And here is the move that gives the whole structure teeth: the first employer guarantees three years.

Not three years of a job. Three years of formation — a guaranteed runway with a curriculum laid across the whole of it, taught by a master who is answerable for what the person can actually do at the end. The master teaches. The master assesses. The master signs the deposits into the account as they are earned, continuously, across the runway — not at milestones, because capability does not arrive on anniversaries. And the wage rises against the capability, not against the tenure, so that for the first time the ledger and the paycheck point in the same direction.

Three years, not five. Five is the number I used to use, and five was wrong — not because formation is faster than that, but because five is an ask no real employer signs. A guarantee that cannot be signed forms nobody. Three is short enough to be signable and long enough to be real, and I would rather have a runway that exists than a runway that is correct on paper and refused in every room I take it into.

If the employer takes the labor and skips the deposits, the account makes the omission visible. The ledger tells the truth the accounting statements refuse to tell.

I can hear the objections. Who certifies deposits? Chambers, guilds, accredited verifiers — the infrastructure the Germans already run through the Handwerkskammer and the Indians approximate through traditional ustad-shagirdlineages. What is the unit of account? Not dollars and not hours, but verified competencies clustered by domain — clinical, mechanical, analytical, relational — and building that taxonomy is exactly the work a serious country would take on. How do you prevent grade inflation? The same way you prevent it in any ledger: independent examination, skin in the game for the verifier, loss of standing for fraud. How do you handle capability that decays? You depreciate it honestly, the way no American institution currently depreciates the human capital on its books.

These are design problems. They are not conceptual problems. The conceptual problem was solved by guilds eight centuries ago, by the Germans a century ago, and by every apprenticeship tradition that has ever produced a capable generation. We dismantled it in the American twentieth century and replaced it with a credential economy that has now lost the public’s trust because it deserves to.

Yale asked the right question. Beard is right that the rest of us owe an answer. But the answer is not better credentials, wider access to credentials, or more transparent credential outcomes. The answer is to stop confusing credentials for capability, and to give the person — not the institution — the ledger.

The bottleneck, as Beard himself writes, is not talent. It is design.

Design the account.

Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of the forthcoming Built to Extract and Already Paid For (Capability Capital Press). He writes at thelonggameforall.substack.com.

Written with AI assistance. The argument, the judgments, and the floor testimony are the author's own.

Maryjane Williams learned her life was over on a conference call.

Twenty years on the floor. Five kids. She carried the medical benefits for the whole family, and the life insurance, and she had just turned fifty. Somebody she had never met, in a room she would never see, had done arithmetic that ended her — and scheduled a call so she could hear the result with everyone else.

She had given those twenty years to a company that, the year the arithmetic started, sold one out of every five toys in America.

The story you have heard about Toys “R” Us is that Amazon killed it. That retail changed and it couldn’t keep up. That is not what happened. The company was profitable. Its operating income was growing. It was selling one in five toys in the country. What killed it was $5.3 billion in debt it never borrowed to build anything — debt loaded onto it, by people who never worked a shift, to buy it using its own body as collateral. And the thing they were really stripping was never on any balance sheet at all.

That’s the argument. Here are the receipts.

The Leveraged Buyout

In July 2005, three firms — KKR, Bain Capital, and Vornado Realty Trust — bought Toys “R” Us for $6.6 billion. They put in $1.3 billion of their own money. The other $5.3 billion was debt, and here is the move that makes a leveraged buyout what it is: the company borrows the money used to buy itself. The buyers don’t carry the loan. The bought does.

Overnight, Toys “R” Us owed $450 to $500 million a year in debt service. Half a billion a year that could have gone to the website, the stores, the supply chain, worker training, lower prices — gone to creditors instead. By 2007, debt payments ate 97 percent of the company’s operating income.

The company was profitable. The debt was not survivable. Those are two different facts, and the gap between them is where 33,000 people fell.

Extraction by Design

They didn’t only load the debt. They extracted while it bled. The advisory agreement signed at the buyout paid Bain, KKR, and Vornado $15 million a year in fees, rising 5 percent annually — and the contract said, in writing, that no minimum number of hours was required of the advisors. Fees for nothing.

The Private Equity Stakeholder Project counted $470 million in interest and fees paid to the three firms before the bankruptcy. Extracted from a company whose floor workers earned $8 to $14 an hour. That $470 million would have covered more than $14,000 in severance for every one of the 33,000 — the severance the company swore it could not afford. John Eyler, the CEO who ran the sale, took home $65.3 million when the deal closed.

Debbie Beard gave the company 29 years, an assistant manager in Chandler, Arizona. “The company makes $11 billion a year,” she said when it ended. “You kind of wonder. It must be an awful big debt if we can’t bring ourselves out of it.” It was. And it was placed there on purpose, by people who never learned her name.

What They Were Really Taking

Here is where the Amazon story falls apart, and I have to tell you something I watched with my own eyes.

I spent ten years, 1989 to 1999, on GM’s Quality Network maintenance team, walking component plants across North America. One of them was Saginaw. Saginaw had invented something nobody else in the world could do as well: cold-extruding hardened steering steel at room temperature, the metal’s own grain flowing along the part so it came out stronger than anything machined from bar. Steel does not want to do that — at those pressures it seizes and welds itself to the tool. It worked because of a piece of process chemistry developed on that floor, a phosphate coating that carried lubricant down into the die. Get it wrong, wreck a die. Get it right, make a part the competition couldn’t touch.

That secret sauce was not written down anywhere. It lived in the hands of the men who ran the tanks and the presses, and in the head of the man who coordinated my work there, Dave Hitz — a local expert of the kind every real plant has and no org chart ever captures. The floor that Roger Smith was trying to automate until it needed no one was, at that very moment, quietly forming a whole cohort of engineers: Ravi Dugiralla, with a doctorate from Ohio State; Joe and Roland; several of us later named outstanding young engineers of the year by the Engineering Society of Detroit. The most valuable thing in that building was the one thing the plan was built to remove.

I tell you this because it is the same theft, in a suit instead of a jumpsuit. What Saginaw had, and what Toys “R” Us had, and what every company in this essay had, was formed capability — decades of it, crystallized in people. And here is the accounting fact underneath the whole series: that capability appears on no balance sheet. When one company buys another, the standard — ASC 805 — folds the assembled workforce into goodwill. It is literally not permitted to stand as its own named asset. So when the extractors arrived, the thing they could strip for free was the thing no ledger was guarding. You can’t loot the buildings; those are booked. You can’t loot the inventory; that is booked. You loot the people, because the people were never on the books.

That is why it is always the workers. Not because they are incidental to the extraction. Because they are the only asset the rules leave undefended.

Not an Accident — a Template

Toys “R” Us shows the theft run from the outside. Sears shows it run from the inside. Eddie Lampert merged Kmart and Sears in 2005 and spent a decade pulling value out through the front door — selling the profitable brands, spinning the real estate into his own vehicles, starving the stores while his hedge fund drew fees. Sears went from 300,000 employees to under 70,000, roofs leaking onto the merchandise, and when it filed in 2018 the man who had run the extraction controlled the bankruptcy. Same asset stripped. Different hand on the knife.

And it is not two companies. From 2015 to 2020, private-equity-owned retailers were 56 percent of all retail bankruptcies — Payless, The Limited, Gymboree, RadioShack, Sports Authority, Claire’s — shedding roughly 542,000 jobs and shuttering about 18,000 stores over two decades, at a bankruptcy rate ten times that of firms private equity never touched. A pattern, not an accident.

The Shelf Proves It

It doesn’t have to run this way, and a company on the same shelves proves it. Costco has never been taken private, never loaded itself with debt to fund a buyback, never stopped reinvesting in the floor. Starting wages well above the industry. Benefits for part-timers. Turnover under 10 percent — a fifth of the industry norm. Its stock has beaten the retail sector over every long window measured, its customer satisfaction leads, its theft rates are among the lowest in retail. Investment outperforms extraction. The evidence is sitting right there on the shelf next to the toys.

The reason extraction keeps winning isn’t that it is smarter. It’s that formation is invisible and cash is not. As long as the capability inside a workforce sits on no ledger, it will always be the cheapest thing in the building to take and the last thing anyone is punished for taking. The fix is not sentiment — it is to make the asset legible. Give a worker’s formation an account that follows them and registers when it is destroyed, and stripping it finally costs something on a page where costs are counted. Name the asset, and you take away the free lunch.

What Extraction Costs

KKR and Bain walked away with at least $470 million in fees and interest. Later, under pressure, they put $20 million into a fund for the workers they had displaced. Twenty million, for 33,000 families. About $600 a family. The advisory fees for doing nothing ran $15 million a year.

Thirty-three thousand people. No severance, no pension, no retraining. A conference call telling them their family was over.

They were told Amazon did it. Amazon did not do it. A profitable company was killed for the one asset the rules forgot to protect, by people who never worked a shift and never learned a name. That’s the argument.

Next: The Hospital They Looted — how private equity pulled $1.3 billion out of a healthcare system while a new mother died because a medical device was repossessed.

Sources: Americans for Financial Reform, “Stop Private Equity from Driving Retailers into Bankruptcy” (2025); Private Equity Stakeholder Project, “KKR, Bain Capital, Vornado repeatedly rewarded themselves” (2018); The American Prospect, “Private Equity: Looting ‘R’ Us” (2018); In These Times, “How Private Equity Killed Toys ‘R’ Us” (2017); The Nation, “Toys ‘R’ Us Workers Take on Private-Equity Barons” (2018); Cal Poly leveraged buyout study (2019).

Responding to: “The American E.V. Has Been Crushed. Will It Take the U.S. Auto Industry With It?”

Matthew Shaer · *The New York Times Magazine* · July 15, 2026

https://www.nytimes.com/2026/07/15/magazine/electric-cars-american-evs.html

When Matthew Shaer asked Scott Case about the three-row electric SUV that Ford unveiled in Dearborn in 2023, Case laughed.

Case runs an EV analytics company, and he had reason. The single hottest category in American electric vehicles right now is the three-row, full-size electric SUV. Kia’s EV9 is crushing it. Toyota and Subaru are rushing theirs out. Hyundai sold twice as many EVs in the United States last year as Ford did.

Doug Field stood on a stage in a Michigan factory town in May 2023 and described that exact vehicle. Fast. Quiet. Three rows. Three hundred fifty miles. He called it a personal bullet train. Ford pushed it back two years in April 2024, killed it that August, and let Field go this spring in a restructuring.

He was right. The market arrived precisely where he said it would — about two years after his employer decided he was wrong, and one quarter after it decided he was surplus.

So here is what I think this article proves, against its own conclusion and against forty-three years of American consensus.

Detroit did not fail to see. Detroit saw — in a named man, on a stage, with the product in the room — and could not hold what it saw for twenty-four months. Seeing is an event; you can buy it by Friday. Formation is a stock; it takes fifteen years and it lives in bodies. Every book ever written about this industry has diagnosed the eyes. The failure is in the hands.

That’s the argument.

The genre

There is a shelf of these books. Yates in 1983. Ingrassia and White in 1994. Maynard in 2003. Murray and Schwartz in 2019. Whyte in 2021. Shaer jokes the next will be called Should Have Seen It Coming, and he’s right about that too.

They are all about the same verb. Yates found Detroit cocky and flat-footed, waiting in corner offices for Americans to come back to gas guzzlers. Susan Helper describes an industry moving through the stages of grief, managers grown over-reliant on the mental models that made them successful — a well-grooved track that traffic turns into a rut. Padgett calls it blinkered thinking.

Detroit couldn’t see. Detroit wouldn’t look. Detroit looked and refused to believe.

Murray and Schwartz get closest — their subtitle says the industry destroyed its own capacity to compete. Capacity is one shelf over from the right word. It’s still not the right word.

And notice what that diagnosis buys you. A perception failure is a tragedy: cognition, frailty, unfixable. Write the book, sell the book, wait eleven years, write the next one. A formation failure is a design failure, and design failures get fixed by people who decide to fix them.

What the crusher couldn’t reach

In 1995 my wife drove a GM EV1 home.

She was the design release engineer for its HTCM — the module that regulates heat in the cabin. Not the battery. Not the motor. The part that keeps a human being warm, which in an electric car is the problem nobody outside the program thinks about, because there is no engine throwing off waste heat to steal. Every degree of comfort comes out of the pack and out of the range. Her module is where the car’s physics met a person’s body.

She parked it in the driveway and asked if I wanted to take it around the subdivision.

So that’s how I rode in an EV1. Not at a press event. On a residential street in Michigan, in a car GM would later hunt down and crush in the desert, because my wife drove it home from work and offered me the keys.

GM formed a few hundred engineers on that car. Real formation: a new architecture, no precedent, everything wrong the first time. Then it crushed them. Literally crushed them, which is the part everyone remembers, and it made a documentary, and the documentary was about the cars.

The cars were never the asset.

By the time the crushing started, my wife wasn’t on the program. She had moved to Lansing to follow me to Grand River and taken a job as an electrical quality engineer, on the floor.

Sit with that, because it is the entire argument standing inside one biography. GM destroyed the vehicles. The judgment those vehicles built did not go into the crusher — it went to Lansing and spent years walking a plant floor, in a body, doing electrical quality on Cadillacs. GM kept the asset. It just never knew it had it, never booked it, never named it, and never once asked what else it had formed and thrown away.

The other half went to California. Alan Cocconi worked the Impact program that became the EV1. He walked out and founded AC Propulsion, which built the tzero, which Martin Eberhard drove before Tesla existed. The crusher reached the aluminum. It could not reach what was in Cocconi’s head, and what was in Cocconi’s head became a company now worth more than the one that formed him.

So GM ran the experiment and published the result: formation is the only asset that survives a program cancellation. Some of it stayed and got used without being counted. The rest drove to Palo Alto. GM noticed neither.

Thirty-one years later Doug Field walks in from Tesla carrying formation the other direction, and Detroit dissolves it in a restructuring. Same script, backwards. The industry has now proven it can neither hold formation nor absorb it.

Two ledgers, identical cash

The Tesla Roadster did not really work. Musk has said so. Cantankerous software, abysmal reliability, a price out of reach for nearly everyone. It sold almost nothing. By every measure a program is judged by, it failed.

Tesla used what it learned building it to build the Model S.

Ford wrote down nineteen billion dollars on EVs. Stellantis twenty-six. Forty-five billion dollars of precisely the same substance — expensive lessons about batteries, software, cost and customers, bought at full retail by engineers who now know things they didn’t know in 2021.

Same cash. Same learning. Opposite ledgers. The difference isn’t the technology or the market. Tesla kept the people. Ford let Field go.

A failed program that forms people is an investment. A failed program that dissolves them is a loss.

No balance sheet in Dearborn or Auburn Hills has a line called engineers who now know how to do this. So $45 billion books as destruction, a Lotus with a battery in it books as a company, and the entire difference is who was still employed on the far side.

The experiment already ran

Now the part of the record nobody reads correctly.

In the early 1980s, under pressure from Detroit, Reagan negotiated a “voluntary” quota limiting Japanese imports and let Toyota and Honda build American plants staffed by local workers. Shaer states the purpose plainly, because it was stated plainly at the time: buy time for the domestic industry — let executives study, and ideally copy, how the Japanese built cars so efficiently.

That is not a jobs program. That is a formation grant. Washington purchased a protected market for the express purpose of forming American engineers on somebody else’s method.

Look what got done with it. Detroit cloned a Civic and called it the Neon, poured capital into pickups, hit record profits in 2000 — and entered the new millennium, in Shaer’s phrase, in a defensive crouch, never recovering anything like its former clout.

The barrier didn’t fail because barriers don’t work. It bought exactly what it promised: time. And time is worth precisely what you form during it. Detroit took a formation grant and converted it to margin.

That trial is closed. We’re running it a second time, on the same institution, with a tariff on Chinese EVs — and the argument you’ll hear this week is that a wall can’t hold a better product back forever. True, and worth nothing: it’s been true since 1983 and changed nobody’s behavior.

Here’s the version nobody is making. The tariff doesn’t protect Detroit from BYD. It protects Detroit from having to form anyone.

The gauge already exists

Stephen Ezell says China takes a new EV from blueprint to launch about 33 percent faster than an American company, and that the gap widens until it’s close to impossible to close. Every reader files that under cost, or scale, or subsidy. It is none of them.

Blueprint-to-launch is cycle time. Cycle time is the pure output of formation with nothing else in it — no capital, no policy, no branding. It measures how fast formed people can decide. You cannot buy 33 percent. You cannot tariff it. You form it or you don’t have it, and forty-five billion dollars is what not having it costs.

Ezell already published the gauge. The trade debate isn’t reading it.

The book nobody has written

Case told Shaer the whole globe is “aimed in one direction.” He’s right, Detroit is aimed the other way, and in about eleven years someone will publish the sixth book about how nobody saw it coming.

It will be wrong exactly the way the first five were wrong. Not because its facts will be bad — the facts have been excellent since 1983, meticulous and completely useless. It will be wrong because it will be about seeing. Forty-three years of diagnosis, and the whole literature is a literature of perception, which is precisely why it keeps needing another volume. You cannot fix what you have misnamed.

So stop denominating this in jobs, price, share, units, writedowns. Denominate it in unformed engineers and it resolves in an afternoon.

That question belongs to anyone. Anyone running a plant, a program, a line, a policy, a classroom can ask it Monday about any protected year, any killed program, any restructuring:

Who did we form?

It has an answer. The answer is on the record either way.

Ask it about 1983. Ask it about the EV1 and the desert. Ask it about the last four years.

Then ask it about the next four, while they’re still yours to spend. That’s the long game. Nobody has written that book yet, and I would rather somebody beat me to it than watch us need a seventh.

Dr. Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of Built to Extract and Already Paid For, forthcoming from Capability Capital Press.

Written with an AI research and editing partner — the augment-not-replace thesis practiced, not just argued. The tool supplied speed, recall, and arrangement; the experience and every judgment are the author’s own.

Vanilla Ice’s real name is Robert Van Winkle.

The man who gave America its most inescapable song about ice is legally named for a character whose entire contribution to literature was sleeping through the thing that mattered and waking to a country he no longer recognized.

He is the right patron saint for this, because withdrawal is never announced. Nobody holds a press conference. It happens while you are asleep, and it gets named afterward, by historians, in books called Decline and Fall.

There are two ICEs in American industrial life. The first is the internal combustion engine — the asset Detroit already owns and has now retreated onto. The second is the cold that a formed engineer meets when they consider coming here. Different reporters, different pages, same story.

Here is the story.

A country can get formed engineers exactly two ways. It can buy them — import people another country paid to raise, at the moment they turn useful, for the price of a plane ticket and an open door. Or it can make them — take its own eighteen-year-olds and stand beside them for fifteen years until they can judge without supervision.

That is the entire menu. Germany makes. Japan makes. China has made, for twenty years, at a scale with no precedent.

America bought. Brilliantly, at enormous margin, for sixty years, and never wrote it down anywhere. Now it is withdrawing from buying, and it has not started making, and both withdrawals have the same cause — which is not immigration and is not batteries. This country has developed an allergy to holding any asset that pays out later than the term of the person who authorized it. Buying formation pays back in a decade. Making formation pays back in fifteen years. Extraction pays Friday.

That’s the argument.

The best business this country ever ran

In the late 1980s I flew into Pittsburgh to start a PhD in industrial engineering. Arjun Jayaraman met me at the airport. Today he is a co-founder of the institute I helped start. That day he was a man who showed up for a stranger.

Somebody met me. That’s not sentiment. That’s a business model.

India paid for my first eighteen years — the schooling, the food, the electricity, the whole crushing subsidy of raising a human to the point of usefulness. My family paid the rest. America paid for a doctorate and collected thirty-five years: Buick City, Lansing Grand River, Stuttgart, Chennai, and now a plant in Wooster, Ohio.

Thirty-five years of formed judgment, and this country never made the deposit.

That was not charity. It was the highest-margin arrangement in the history of industrial policy. America did not have to form its engineers. It only had to be worth coming to. Be the place where the work is real and the door is open, and finished people arrive on their own — funded by another treasury, showing up at exactly the age when they start producing, never appearing on any balance sheet because nobody bought them.

Every rival had to actually raise its engineers. America just had to be attractive.

We are closing that business now, by two mechanisms, and only one of them is written down.

Withdrawal One: from buying

New F-1 student visa issuances are down roughly 36 percent year over year. From India — the country that has been paying the first eighteen years of America’s engineering bill since before I got on the plane — summer issuances fell 60 percent. NAFSA surveyed 149 American institutions this spring: new international graduate enrollment down an average of 24 percent.

The University of New Haven lost around three thousand graduate students in two years, opened a $35 million hole that was 17 percent of its budget, cut ten academic programs, and stopped contributing to its employees’ retirement accounts.

Those are American jobs, gone, because foreign students stopped coming. Sit with that before you pick a side.

Now the machinery, which is where this stops being a trend and becomes a statement.

DHS finalized a weighted selection rule, first applied to this March’s H-1B cap. The lottery is no longer random. It is weighted by wage level, and there are four. Level 4 gets four entries. Level 3 gets three. Level 2 gets two. Level 1 gets one.

Read that as a formation document, because that is what it is. Wage level is a formation proxy — the closest thing the federal government has ever built to a capability gauge, and it built it by accident. Level 1 is a person at year one. Level 4 is a person at year fifteen.

So the United States has written into regulation, with an effective date, a four-rung formation ladder — and inverted the weights. Four tickets for the person somebody else finished. One ticket for the beginner. We will bid for the formed. We will not participate in the forming.

Then the fee, which has the best story in the piece.

In September a presidential proclamation put a $100,000 charge on new H-1B petitions for people abroad. Twenty state attorneys general sued. On June 8, Judge Leo Sorokin vacated it — not on compassion, on taxonomy. He held the $100,000 was not an immigration restriction at all. It was a tax, which a president cannot impose. And he reasoned his way there from the Supreme Court’s tariff case.

A federal court looked at the price America set on a formed human being arriving at the border and ruled that the correct legal category was tariff.

There is a tariff on Chinese electric vehicles. There was, until a judge said otherwise, a tariff on Indian engineers. Same instrument, two imports, same defect. A tariff on cars protects you from having to build a car. A tariff on engineers protects you from having to build an engineer. Both work exactly as long as the wall holds and not one hour longer.

The part that isn’t written down

Everything above was signed by someone. It can be litigated, and is being.

This next part can’t, and that’s what makes it work.

ICE’s Student Criminal Alien Initiative ran the names of 1.3 million international students through a federal criminal history database. Under Catch and Revoke, a State Department revocation alone can trigger termination of a student’s SEVIS record and the start of removal proceedings. The attorneys tracking it list the risk factors: a past encounter with law enforcement even if charges were dismissed or never filed; an old misdemeanor; attendance at a protest. Many students got termination notices with no discernible reason at all.

The point is not that this is cruel. The point is that it is unpredictable. Only the second claim is load-bearing.

A twenty-two-year-old deciding where to spend the formation their family spent eighteen years accumulating is making a twenty-year bet — degree, visa, green card, career, mortgage, children. You cannot price a twenty-year bet against a posture that changes without notice and counts charges that were never filed among its risk factors. There is no compliance strategy. There is no way to be safe.

A rational person facing an unpriceable bet does not negotiate. They go somewhere else.

Which means you can support strict enforcement, genuinely, on the merits, and still lose this argument. Ambient deterrence has no targeting mechanism. It lands on the person you wanted exactly as hard as on the person you didn’t, and the person you wanted is the one with options.

So look where they went. Same survey: 82 percent of Asia-Pacific institutions outside Australia, and 47 percent of European ones, reported international undergraduate enrollment growing.

The forming didn’t stop. It relocated.

Be honest about the rest or the piece isn’t worth reading. Canada, Britain and Australia are all down too. China’s youth unemployment is above 16 percent and Chinese families are increasingly unconvinced a foreign degree pays. Not all of this is America being cold.

Enough of it is.

Withdrawal Two: from making

Meanwhile, the other engine. Stellantis wrote down $26 billion on EVs, Ford $19 billion. Lines dormant, battery plants repurposed to industrial storage, the whole industry backing onto trucks while one in four vehicles sold on earth runs on a battery and China builds three-quarters of them. I take that story apart on Thursday.

The one line that matters here: it is not confusion, it is alignment. A truck pays this quarter. A formed EV engineering organization pays in about fifteen years. There is not one compensation committee in America that pays anybody on fifteen years.

If you want the tell, it isn’t in Dearborn — it’s in Texas, where a regulatory loophole let thousands of new fossil-burning power sources onto the grid to run AI data centers and the neighbors found out from the dust. The same country de-electrified its cars and is burning gas to run its inference. Nobody holding a coherent view about carbon arrives there. That’s not a position on climate. That’s a position on capital.

Detroit retreats onto trucks: we already have an engine.

America shuts the door: we already have engineers.

Same sentence. Same bet.

The best argument against everything I just wrote

In its strongest form, which is not the version you hear on television:

The visa is the reason America stopped forming its own. Every H-1B was a reason not to fix a high school. Every imported master’s was an apprenticeship that never got funded, a kid in Flint or Akron who never got the seat — because the seat could be filled instantly by someone who arrived pre-formed. The pipeline was a painkiller. It let the country skip the surgery for two generations. Cut the supply and America will finally have to make its own.

That argument is correct.

It is also a formation argument. The restrictionist case, at its best, isn’t about foreigners at all. It’s a claim that America outsourced its formation and atrophied — which is precisely my claim, arrived at through the opposite door. Strip the politics off and it’s a sentence about deposits.

Both sides of this fight are making formation arguments. Neither side knows it. So they argue about people — how many, from where, at what wage — and the only question that matters goes unasked for another decade.

Here’s where it fails, and it fails the way Detroit failed.

Withdrawal is not surgery. Taking away the painkiller does not summon a surgeon. And no surgeon arrived. There is no national apprenticeship. There is no capability account. There is no deposit at year one for an American kid either. We closed the import business and went to sleep.

A country that neither buys formation nor makes it isn’t running a strategy. It’s drawing down an inventory. Detroit’s trucks are inventory. America’s aging engineers are inventory. Neither is being replenished, and inventories do exactly one thing.

The part that should give you hope

The weighted lottery works.

Somebody sat in a room and built a national mechanism that identifies where a human being sits in their formation, assigns it a number, attaches real money and real odds, survives notice-and-comment, and administers it across hundreds of thousands of people with an effective date.

For thirty years every objection to a capability account has been the same: you can’t measure it, you can’t administer it, it’s too soft, it’ll never survive the rulemaking. It isn’t, you can, and it did. We built one this March. We aimed it backwards.

And here is the part I don’t have finished, which you should know before you take any of this on faith.

France has been running my argument since 2015. The Compte Personnel de Formation — the Personal Formation Account. Twenty-five million workers, euros accruing annually, worker-controlled and portable. Exactly the instrument I’ve spent four years describing.

It got looted. Training providers cold-called workers to harvest their accounts — providers verifying their own value, selling into a ledger nobody was guarding. I have a firewall drawn for that: the people who train can never be the people who certify. On paper it holds. France ran the experiment and broke on the precise joint I designed for, and I do not yet know whether my joint is stronger than theirs or whether I simply haven’t been tested at twenty-five million.

I’d rather hand you that question than pretend I closed it.

So — Monday morning, your plant, your program, your line, your last four years:

Who did we form?

Not who did we hire, recruit, retain, or import. Who did we take at year one and stand beside until they could judge without us.

If the answer is nobody — and it was nobody at GM in 1996, nobody at Ford in 2024, nobody at the consulate this March — then this is not a talent shortage, not a trade problem, not a China problem.

It’s a country that stopped being worth coming to before it ever got around to becoming a place that makes its own.

Van Winkle slept twenty years and woke to a republic he didn’t recognize. Detroit has slept forty-three and is still going. The consolation is that inventories take a while to run out, and the quiet is what lets you keep sleeping. That’s the short game. We are extremely good at it.

The long game is the one where somebody wakes up and forms a person.

Ice, ice, baby.

Dr. Venki Padmanabhan is a co-founder of the Capability Capital Institute and the author of Built to Extract and Already Paid For, forthcoming from Capability Capital Press.

Written with an AI research and editing partner — the augment-not-replace thesis practiced, not just argued. The tool supplied speed, recall, and arrangement; the experience and every judgment are the author’s own.