Four Pounds of Debt for One Pound of Growth

Every rescue has to be bigger than the last one because each rescue adds to the very problem it’s treating. A rescue is new money and credit, and new credit is new debt. That debt joins the pile that has to be kept serviceable the next time trouble comes. Meanwhile technology keeps cutting the cost of making things, and the push behind that fall keeps getting stronger, because the tools are improving faster than they did, so what policy has to cancel out keeps getting bigger. And each new pound of debt buys less growth than the one before it. So the gap to fill is wider, the pile to defend is taller, the medicine is weaker, and the economy is more fragile. Multiply those together and the next rescue has to be larger just to keep the picture looking stable. Nothing in that loop settles, because every part of it feeds the others.

The cost of keeping prices up is the abundance that technology created and that never reached you as lower prices. Because the loop never settles, this system changes, one way or another. What’s open is how, and who pays on the way.

What it takes to hold prices up

Most money is loaned into existence.¹ When a bank lends, it creates the deposit it lends to you.² Most of those debts are nominal [fixed in pound terms, so the amount you owe doesn’t change when prices change]. A system built on those debts can take a slow fall that comes with better tools. In that fall it’s the cost of making things that drops, and pay drifts down slowly behind it, so the money to meet a fixed payment is still there. What it can’t take is a sudden fall in the prices credit pushed up, houses first, because then the losses spread through the banks. In that fall it’s the lending that has stopped, so incomes and the house fall together and at once, and the fixed payment is hit from both sides in the same year. Houses come first because people buy them with borrowed money. The debt takes a bigger bite of every income, yours, your employer’s, the government’s. And where the loan was written against a house, the house is worth less too, so the bank stands to lose more on that loan. Push that far enough and defaults cascade through the chain of lenders, because each lender’s asset is someone else’s promise to pay. So every time a sudden fall in those prices threatens, the system is rescued with new money, and the rescues keep getting bigger. The danger is the debt design, not the cheaper goods, which are exactly what better tools should deliver.

In between the rescues, policy targets rising prices of about 2% a year, forever.³ Central banks publish why they aim there.⁴ They want room to cut interest rates in a downturn, they know wages resist falling, and they say the official measure of prices overstates the rise.⁵ Those are the reasons they give, and the debts aren’t among them. The debts show up in the rescues, not in the everyday 2%. The starting line for that 2% isn’t zero. Under a money nobody can add to, the natural direction of prices is down, because people keep getting better at making things and competition passes the saving on.

So when a central bank targets 2% inflation, it doesn’t need two points of push. It first has to cancel whatever better tools would have taken off prices that year, then add two on top. How big that fall would be is about how fast output per hour of work grows, which is what the productivity figures measure. On the published figures that’s a point or two a year.⁶ No statistic records that fall, because it never happened. Say that fall is 1%. Then the system has to engineer three points of upward pressure. If technology accelerates and the natural fall becomes 5%, it has to engineer seven. It’s like cycling up a hill that keeps getting steeper. You can’t stop and hold your place. Just staying where you are takes a harder push.

The push itself is made of ordinary policy moves. Interest rates are cut so households and firms borrow more. Governments run deficits [spend more than they raise in tax and borrow the difference]. Central banks buy bonds with newly created money [quantitative easing, QE] to keep borrowing cheap. Every one of those channels ends in the same place. Each one either creates new claims on future income or makes creating them cheaper, and both leave more debt behind.

The gears that make it compound

Gear one. The gap widens on its own. Software already pushed the cost of copying anything digital toward zero. AI is now pushing the cost of knowledge work the same way, and robotics is starting on physical work. That downward force on the cost of making things compounds, because better tools build better tools. So even before you count any debt, what it takes to cancel the fall keeps growing. The official output-per-hour figures don’t show that pull growing. They miss part of it, because what becomes free drops out of the count, though on the measures tried so far that part is small.⁷ It shows in what the tools can do.

Gear two. The pile grows, and the floor rises. Each rescue’s borrowing joins the stock of debt. That stock never meaningfully shrinks, and it has to be kept serviceable. Worse, the share and house prices the rescue propped up are kept up through credit, because loans are secured against them. What that protects is whether the lenders survive, and the state acts wherever a lender is about to break. Where lenders borrow against things that are valued every day, it acts within days. Where a fall has to eat into the banks’ own capital, it acts later and larger, after the fall has already been taken as a loss. Below the point where a lender breaks, prices are allowed to fall, and UK house prices did fall after 2022 with no purchase programme to hold them up.⁸ But let house prices that cheap credit lifted fall far enough to reach the banks, and the level the lenders can survive is the level the next rescue defends. Each rescue lifts those prices further from where buyers and sellers alone would set them. New loans are then written against the lifted prices, so the level the lenders can survive rises with each rescue. There’s also a mechanical trap. After a price spike, prices would naturally fall back the next year, and unless the new money keeps coming, the price index [a basket that tracks consumer prices over time] reads that as deflation. You have to keep pushing against the level your own last push created.

Gear three. The medicine weakens. Between 2000 and early 2018, the world added roughly 185 trillion dollars of new debt to get about 46 trillion dollars of growth. Call it four borrowed pounds for one pound of growth. The ratio is the same whichever currency you count in. That ratio worsens because more of each new pound goes to paying the interest on old debts and renewing them rather than building anything. Cheap credit also keeps firms alive that competition should have replaced [zombie firms, companies that survive on cheap borrowing rather than real profits]. Those firms sit on capital and talent that better firms would have used, which drags on the very growth the borrowing was meant to buy.

Gear four. The pain threshold falls. Because markets have learned the rescue comes every time, they build on top of that expectation. Companies borrow to buy back their own shares. Investors take more risk because when the bets fail, the rescue passes the loss to savers and wage earners, through the new money it takes to fund it. The whole structure is built around permanent support, so ever smaller wobbles now demand a rescue. Time after time, when central banks hinted at shrinking support, markets convulsed and policy came back in. Support can be pulled back part of the way for a time, as it was from 2022 to 2024, when rates rose and bond holdings shrank.⁹ But each place the pull-back strained got new support within days, the Bank of England’s in September 2022 and the Federal Reserve’s in March 2023, and the promise to step in was never withdrawn.¹⁰ You can’t ease off the pedals for good, because withdrawing the promise would cause the exact collapse the rescue was meant to prevent.

And a fifth gear. The rescue steepens the hill. When policy pushes up wages, rents, and input costs while customers still demand lower prices, automation pays for itself faster. Firms replace labour with software and machines sooner than they otherwise would. So the intervention adds to the force it has to cancel, which widens next year’s gap again. The fifth gear is what feeds the first.

The gears multiply each other. The next rescue equals a wider gap, times a taller pile, times weaker medicine, times a lower pain threshold. Every term in that multiplication is fed by the previous rescue. That’s why it compounds rather than settles. The record shows the stock of debt growing, and the rescue that each downturn gets growing with it. The yearly pace of new money hasn’t trended up. Central bank holdings jump in a crisis and come only part of the way back.¹¹ Rescues went from hundreds of billions of dollars of government spending and bond-buying in 2008 to trillions in 2020, and in the first year of the pandemic response the United States’ stock of money, cash and bank deposits together, grew by more than one and a half times the before-tax profits of every US company for 2020.¹² Taxes couldn’t fund that scale. Only newly created money could, which tells you what the system actually runs on.

Where the bill lands

The bill lands in four places. The first is invisible, the price falls you never got. Technology made photography, navigation, communication, and computation nearly free, and it’s been cutting the cost of making food, energy, and goods the whole time. Under a money nobody can add to, those gains would have reached you as a falling cost of living. Instead the expanding money supply absorbed them, so prices stayed flat or rose. Against what prices should have done, your yearly loss is the fall you didn’t receive. The rise you paid on top of it is the second bill. You worked just as hard, and the saving existed. Part of it reached you as higher pay, late and unevenly.¹³ The rest went to whoever got the new money first, through house and share prices.

The second is your wage and your savings. Inflation and wage deflation are the same event seen from two sides. Your pay rises slowly, so it lags the money creation. Your cash savings are guaranteed to lose purchasing power [what your money can buy], which forces you to become an investor and take risk just to stand still. A mechanic who saves in cash falls behind through no fault of his own.

The third is the transfer. New money usually enters through bank loans and central bank bond purchases, landing with whoever sold the house or the bonds, so it reaches the people who already own houses and shares first, before prices adjust for everyone else. The landlord’s house and rents rise while the tenant’s pay lags.¹⁴ That’s the transfer working as designed, and it widens the gap between owners and earners every cycle, feeding much of today’s political anger.

The fourth is the future. Capital flows to whatever policy props up rather than to what serves people. Houses stop being shelter and become savings vehicles, pricing out the young. Fragility builds. And as stress rises, the system defends itself with more centralised control, because a structure that must keep the support going eventually has to manage what people hear and how they behave too. The two doors at the end of this road are a depression if support stops, or escalating political repression if it continues. History’s versions of this story, from Weimar [Germany’s 1920s currency collapse that preceded Nazi rule] onwards, ended in resets, strongmen, or wars.

If the mechanism is right

If this mechanism is right, the collision between the debt design and cheaper tools steepens from here, because no industry is safe from AI’s push on the cost of making things, and that push is compounding. For a state carrying debt at the scale the United States, the United Kingdom, and Japan now carry, the ordinary exits are blocked. The debt can’t be repaid in real terms [after allowing for what money will buy] at this scale, default is the very thing the rescues exist to prevent, growing out of it fails because each unit of debt buys less growth, and raising rates until the bad debt is forced out is closed too, for the reason the last objection below gives. Whichever exit is refused, the interest keeps adding up. With less debt against a year’s output there’s a door that list doesn’t name. A long run of a state taking in more than it spends is ordinary, and Canada walked it, its debt falling from about two thirds of a year’s output in the mid-1990s to under a third by 2009, with its currency rising against the US dollar.¹⁵ What closes that door at today’s debt levels is how big the surplus would have to be, and none of the three is running one.

When the exits are blocked and creative destruction [the market process where better methods and firms replace old ones] is suppressed in markets, it moves up a level, to the money itself. The depression and the repression, the two doors at the end of the road, both end in the same place. The eventual reset happens at the money itself, the stick everything else is measured against. That comes in two forms, and they aren’t the same thing. The slow form is a long loss of what the money buys, and for the dollar since it left gold in 1971, that’s a measured fact.¹⁶ The event form is a break in the money itself, and that’s a forecast. I can’t put a date on that break, because the path to it is chaotic and no one can time the trigger. What I hold is that some form of reset is unavoidable, and that each rescue makes it larger. Technology’s speed also brings it on faster than the old debt cycles, which took decades to unwind.

Because the reset lands on the money, my conclusion is a base money [the foundation money everything else is priced in and settles to] whose supply can’t be expanded. Under a money like that, the gains from better tools can reach people as falling prices, and debts shrink in importance because saving works again. Credit can still be built on top of a fixed base, that credit can still go bad, and the firm that lent it can fail. What changes is that the base can’t be expanded to rescue it, so the loss stays with the people who dealt with the firm that failed instead of spreading to everyone holding the money. Those people needn’t have known they were taking a risk, and in the American bank failures before the federal government insured deposits most of them were ordinary depositors.¹⁷ Whether the credit built on top is kept fully reserved [every claim backed by money held] is a social choice, and the money’s rules don’t settle it.

Five ways this could be wrong

“If the system needed exponentially bigger rescues, it would have collapsed already.”

It’s been failing in slow motion. Two things masked it. The same price falls the new money absorbs, cheaper goods from better tools, kept a lid on shop prices for decades, and reserve-currency demand for dollars [the dollars the rest of the world, including its central banks, holds for trade and savings] spread the cost across the whole world. Both are wearing thin as the rescues grow. You’ve been watching the first of those happen since 2020.

“You can taper once growth returns.”

Each exit so far has been partial, as the pull-back of 2022 to 2024 was, because positions built with borrowed money on the promise of support have to be marked down and sold the moment the promise wavers, and new support arrives within days. A system that can’t withdraw that promise is hostage to the borrowing built on it.

“Growth will outpace the debt.”

The arithmetic runs the other way. Four pounds borrowed for one pound of growth, and the ratio worsening, while the interest on the pile compounds regardless.

“The rescues are worth it, because deflation would be worse.”

There are two reasons prices fall, and they need separating. Prices falling because credit is collapsing is destructive.¹⁸ Prices falling because we got better at making things is cheaper life, and people still buy phones and food when they expect them to be cheaper next year. The rescues defend the debt design, not your living standards. Debt of this design can’t coexist with accelerating technology.

“A Volcker moment could purge it, the way high rates did in the early 1980s.”

A purge on the pattern of the early 1980s can’t do now what it did then. Back then the debt was mainly with households and companies rather than the state, so crushing rates punished borrowers, and the state was well enough placed to apply the cure. Now the state itself is the heavily indebted party. So the cost of higher rates lands first on its own interest bill, on the part of its debt whose interest changes at once. Rates can be raised, and they were. Both the Federal Reserve and the Bank of England took them from near zero to above 5% between late 2021 and mid-2023 while shrinking their bond holdings.¹⁹ What I hold is that the political cost, and the cost to the system itself, of going far enough to force the bad debt out is likely intolerable, so the policy I expect instead is holding returns below the rate prices rise. That’s what happened to the lenders who held ordinary government debt through 2021 and 2022. It isn’t what a new lender has been offered since. On the United States’ inflation-protected ten-year debt, the return above inflation has been above zero since the middle of 2022.²⁰

When the next rescue arrives and it’s larger than the last, you’re watching the design work. Nobody has to be incompetent for the rescues to keep growing. A system built on debts fixed in pounds can’t take a sudden fall in the prices credit pushed up. Every time one threatens, it has exactly one move. Put that system up against technology that keeps cutting the cost of making things, and the moves keep getting bigger.