When every money figure you’ve ever read about the economy is written in a unit whose supply moves, what changes is perception. The unit is the ruler that stretches while you measure.
Every number you’ve ever read about the economy that’s a sum of money was written in a unit that was changing while you read it. Your first payslip, your parents’ house price, the GDP charts in your textbooks, and the “record highs” in the news were all written in pounds or dollars, and across the post-war decades, the supply of pounds and dollars was higher at each decade’s end than at its start.1 That’s the fact underneath all the others, so far down that almost nobody ever looks at it.
The instrument you see with
The money you spend does more than pay for things. It’s the unit that prices and debts are written in. It’s also the instrument you see the economy with. When the supply of that unit moves, the distortion doesn’t show up as an error you can spot. It disappears into the measurements themselves. You don’t perceive a distorted world. What you perceive hangs together. It reads as the ordinary way of things, not as a fault in the instrument. And it’s wrong about what changed. And because almost every check you’re likely to run is written in the same unit, the world keeps confirming itself.
That’s the root of why nobody sees the distortion. It isn’t stupidity, and it isn’t a conspiracy. There’s never been a fixed point outside the unit to stand on. You can’t test a ruler with the ruler.
What happens to perception
Three things pile up.
First, the sign flips. When the money is being expanded, a thing that hasn’t changed, or has got cheaper to make, can look as if it’s gone up. Your house “doubles” over ten years and you feel richer. But the house didn’t change. It has the same rooms, the same roof. What changed is that each pound measures less, so the same house takes more of them. Meanwhile you still have to live somewhere, and the house you’d move into costs more as well. The tax and upkeep climbed too. The gain you perceive is mostly the ruler shrinking. Same with your pay. A 3% rise reads as progress. If your food, rent, and energy rose 8%, you took a pay cut, and your perception recorded it as a raise. Inflation and wage deflation are one phenomenon seen from opposite sides, and the stretching unit keeps the two sides apart, so the gain arrives as news and the loss arrives as a squeeze you can’t trace. That’s why the illusion is so stable. The half you notice is the half that flatters you, and the half that hurts never leads back to the ruler. People don’t defend the truth of their instruments. They defend the instruments that tell them they’re winning.
Second, the baseline vanishes. Even people who accept that inflation exists miss this part. The natural state of a free market, under a money nobody can add to, is falling prices, because when tools improve, the same goods take fewer inputs to make, and competition passes that saving on. So the honest baseline is minus something rather than zero. With software and AI, it’s minus a lot.2 Measured against that baseline, flat prices aren’t stability. Flat prices mean the entire price fall that year’s better tools should have delivered was absorbed by money creation. The gain from the better tools is real. Part of it reaches you as higher pay, late and unevenly, and the part that doesn’t goes to whoever got the new money first, through house and share prices.3 The weekly shop that should have got cheaper this year cost the same instead, and a bill that hasn’t moved gives you nothing to notice. Perception needs contrast, and the theft removes its own contrast. Nobody mourns the price fall that never happened. The crime scene is a world that never existed, so there’s nothing to point at.
Third, the correction tools inherit the error. This is what closes the loop. You might say we know about inflation, and that’s why we have “real terms” figures. But the adjustment is done with an index built in the same unit, and that unit is managed to hit a target the same index is used to score. The 2% target makes this worse. Once a measurement becomes the goal, the system is managed to hit the number, the way a school that decides the average grade must be a B scales every test until it is. And the deeper distortion never enters the index at all, because the official price index, the CPI, measures against zero, not against the falling-price baseline that technology keeps delivering.
The price index leaves out the price fall that didn’t arrive, and GDP [the official tally of money spent] leaves out the spending that left when a priced thing became free, so the same gain goes unrecorded from both sides. Neither measure breaks its own rule by leaving the gain out. The gain is outside both rules. That makes both a poor guide to what technology delivered. When your phone swallowed the camera, the map, the sat nav, and the postage, your life got cheaper and GDP recorded almost nothing, because GDP counts what was paid for, plus a short list of items priced as if they’d been paid for, and it doesn’t count what a thing is worth to the person using it above what they paid.4 GDP inherits the same error as well. Turning it into “real” growth takes another index, built in the same unit against the same zero, so money borrowed into existence and spent reads as growth.5 So the two official instruments this section has walked through are blind in both places where the distortion lands.
What this does to people
Money is information. It’s the signal you use to decide what your time is worth, what to build, whether you’re getting ahead. Distort the signal and you distort the choices.
A saver looks at cash losing to everything and concludes, correctly within the frame, that saving is for losers. Holding cash at a loss pushes everyone toward risk just to stand still. Many can’t save enough to follow, so the direct moves into let property and funds stay a minority. A worker feels the squeeze but can’t see the cause, because the mechanism hides inside the unit itself, so it never shows up as a line on any bill. So the anger lands on whatever is visible, the landlord, the boss, the immigrant, the other party. The resentment between labour and capital that dominates politics is largely misdirected, a fight over who stole the gains, conducted by people who can’t see that the gains were absorbed in the money itself, before either side could bargain over them. Inflation manufactures the feeling of scarcity in the middle of an era when technology keeps making almost everything cheaper to produce, and people fight over fake scarcity just as hard as real scarcity.
Perception depends on the unit you think in. It seals itself against any check you run in the same unit. Someone who thinks entirely in pounds sees prices relentlessly rising. Someone holding bitcoin but still thinking in pounds sees a volatile asset “going up”. Someone who thinks in bitcoin sees the same houses, laptops, and flights getting cheaper cycle after cycle. Each of the three observers sees a world that hangs together and is confirmed by its own data. Even the words carry the unit inside them. An everyday phrase that reports a price move names only the thing, so it states the unit’s move as the thing’s move, from a house that “doubles” to an asset “going up”. Argument can’t settle which observer is right, because the arguments are conducted in numbers, and the numbers belong to a frame.
Why the slowness matters
How fast the money loses value is a large part of what decides whether people can name the loss, meaning the day, the amount and who did it. It isn’t what decides whether rising prices get noticed at all. In July 2024 Ethiopia’s central bank cut the birr’s official value by nearly a quarter in a day, from 57 to 75 against the dollar, and let it float from there.6 Shop prices jumped within the week, and by the start of August the authorities were closing shops for raising them, because the stretch happened all at once and everyone could see it, the state included. The same confiscation, running at 2% measured plus the hidden gap to the falling-price baseline, which is a point or two a year on the published figures, compounds for thirty years into a larger theft, and there’s no day to point at.7 Nothing ever happens. Your savings are never seized. There’s no line item, no event, and human perception is built on events. We also think in straight lines, so a small yearly stretch that compounds for a generation is doubly invisible. It’s too slow to alarm you, and it compounds in a way our minds don’t expect. By the time you’re an adult, the stretching ruler is the water you learned to swim in. Every memory you have of “what things cost” was laid down in it.
Where the fixed ruler comes in
This is why the measurement question, not the investment question, is where bitcoin enters the thesis. A measurement needs an independent variable, a reference that isn’t changed by the thing being measured. For the first time the unit itself is one whose supply no single party can change.8 If you hold it directly, you can check that for yourself. It keeps that property only while the network stays decentralised and secure. If you take a claim on a custodian instead, you’re back to a quantity someone else can change. Measured against that ruler, the picture flips. A family home that took roughly 300 bitcoin in 2017 takes a fraction of that today, and it has kept falling across the cycles since, while its pound price rose. The whole of that fall is the new unit hardening, not houses getting cheaper to build. In pounds the house got dearer, so cheaper houses can’t be any part of the explanation. The house didn’t change. Both numbers did, in opposite directions, which makes it a fact about the two rulers and not about houses. That hardening is the loud part while the world is still moving into the new unit. Underneath it runs the slower fall technology has been delivering all along, the one the old unit hid. Nearly everything is getting cheaper against the fixed ruler. Bitcoin “going up” and everything “getting more expensive” turn out to be the same optical illusion, seen through the moving unit. The point isn’t the price. A second unit whose supply no committee can change lets you see the first one move without trusting an index built in the first unit. There are other ways to see it, and there were earlier ones. The second unit is the first reference that isn’t changed by what it measures.
You can’t argue a mind out of a frame using numbers stated inside the frame. Every price and every total arrives already written in a unit, and the mind will file it back into the old unit. What works is starting from the instrument. Take a second ruler and re-measure one thing you already know intimately, your own house or your own payslip, and let the flip happen in your hands. Perception changes by re-measurement, not by persuasion.
The arguments against this, and the unit they’re made in
“Official real-terms figures already correct for this.” The price index does read the stretch of the pound against the goods it prices. What it can’t read is the fall those goods would have shown under a unit nobody can expand, because its adjustments are made in the same stretched pound. It corrects against the wrong baseline, using baskets that give too little weight to the things people must hold, like housing.9 The correction is sincere and it’s circular.
“Markets gather everyone’s information into prices, so a distortion this large would be traded away.” Markets gather that information in the unit. And the anchor of every valuation, the so-called risk-free rate on government debt, is pushed below where free buyers and sellers would set it, so every price worked out from that rate bends with it. You can’t trade the unit away from inside the unit.
“Bitcoin is far too volatile to be a ruler.” Today’s volatility is a small, young asset repricing the whole world, and it’s measured in the stretching unit. The property a measure needs is a supply no committee can change. A calm chart was never the test. And on the record, over multi-year windows, the direction of prices in the fixed unit has been consistent, and it points down.
So when every money figure you’ve ever read about the economy is written in a unit whose supply moves, perception succeeds at seeing a false world, and the illusion recruits your own gains, your own language, and your own correction tools to keep it false. The ruler that stretches while you measure doesn’t just mismeasure the plank. It convinces the carpenter he’s getting taller.
Sources 9
What this essay rests on, as checked on 20 October 2026.
- 1 Board of Governors of the Federal Reserve System, H.6 Money Stock Measures, M2, seasonally adjusted (series M2SL), monthly to August 2026: M2, billions of dollars, seasonally adjusted, at each decade start: 1959-01 286.6; 1970 589.6; 1980 1,482.7; 1990 3,166.3; 2000 4,666.1; 2010 8,473.1; 2020 15,425.0; 2026-08 23,342.8. The series starts January 1959 and covers dollars only, not pounds.
- 2 Stanford University Institute for Human-Centered Artificial Intelligence, The 2025 AI Index Report: "Driven by increasingly capable small models, the inference cost for a system performing at the level of GPT-3.5 dropped over 280-fold between November 2022 and October 2024." Cost fell from $20 to $0.07 per million tokens.
- 3 Stansbury and Summers, National Bureau of Economic Research Working Paper 24165, “Productivity and Pay: Is the link broken?”: "over 1973-2016, one percentage point higher productivity growth has been associated with 0.7 to 1 percentage points higher median and average compensation growth"; "other factors orthogonal to productivity have been acting to suppress typical compensation". Pessoa and Van Reenen, Centre for Economic Performance Discussion Paper 1246, “Decoupling of Wage Growth and Productivity Growth? Myth and Reality”, October 2013: "In the UK net decoupling is basically zero: both wages and productivity rose by the same amount since 1972"; UK 1972 to 2010, "real productivity grew by 114 per cent" and "median real wages by 72 per cent".
- 4 Australian Bureau of Statistics, Australian System of National Accounts: Concepts, Sources and Methods, 2020-21, Chapter 23, “Unpaid work”, quoting System of National Accounts 2008 paragraph 6.27: “the production of all goods and services that are supplied to units other than their producers” and “the own-account production of all goods that are retained by their producers for their own final consumption” Office for National Statistics, Changes to National Accounts: imputed rental, 23 March 2016: “It can be thought of as the amount that non-renters pay themselves for the housing services that they produce.” “Imputed rental represents around 10% of GDP as measured by expenditure.” Brynjolfsson, Collis, Diewert, Fox and Eggers, NBER working paper 25695, “GDP-B: Accounting for the Value of New and Free Goods in the Digital Economy”, March 2019: “The welfare contributions of the digital economy, characterized by the proliferation of new and free goods, are not well-measured in our current national accounts”; the paper “introduce[s] a new metric, GDP-B which quantifies their benefits, rather than costs”.
- 5 ONS, Chain-linking methods used within the UK National Accounts: “If a suitable deflator (or equivalently referred to as a price index) exists” … “The resultant time-series of production figures has the effects of price changes removed.”
- 6 International Monetary Fund, Press release, “IMF Executive Board Approves Four-Year US$3.4 billion Extended Credit Facility Arrangement for Ethiopia”, 29 July 2024: "The Executive Board approved an SDR 2.556 billion (about US$3.4 billion) ECF arrangement for Ethiopia"; "Move to a market-determined exchange rate to help address external imbalances and relieve FX shortages"; approved July 29, 2024.
- 7 Bank of England, Inflation and the 2% target: "The Government sets us a 2% inflation target"; also "The Government has set us a target of keeping inflation at 2%". The page prints 2% as the target set for the Bank, not a measured rate.
- 8 Nakamoto, “Bitcoin: A Peer-to-Peer Electronic Cash System”, 2008: "there is no central authority to issue them"; "Once a predetermined number of coins have entered circulation, the incentive can transition entirely to transaction fees and be completely inflation free." Bitcoin Core (the published consensus rules), Bitcoin Core source file src/consensus/amount.h, master branch: Code: "inline constexpr CAmount MAX_MONEY{21'000'000 * COIN};" with the comment "No amount larger than this (in satoshi) is valid. Note that this constant is *not* the total money supply ... but rather a sanity check."
- 9 Paul Johnson, UK Statistics Authority, UK Consumer Price Statistics: A Review, published 8 January 2015: "Paul Johnson's report finds that, other than not accounting for owner occupiers' housing costs, the Consumer Prices Index (CPI) is a well constructed measure of inflation." The review prints the omission of owner occupiers' housing costs and otherwise calls the CPI well constructed; it does not say basket weights are too low.