When the Planners Cannot Tell Goods From Services
On 17 June 2026 the Office for National Statistics published its monthly Consumer Price Inflation bulletin for the May 2026 reference period, and the headline was a study in stasis. The Consumer Prices Index including owner occupiers' housing costs — the ONS's most comprehensive measure of inflation — rose 3.0% on the year, unchanged from April. The narrower CPI rose 2.8%, also unchanged. Core CPI ticked up from 2.5% to 2.6%, the kind of drift that fills two paragraphs in a central-bank minutes document. Read past the headline and the picture is the opposite. CPIH goods inflation slowed from 2.4% to 2.0%. CPIH services inflation accelerated from 3.4% to 3.6%. CPI services inflation jumped from 3.2% to 3.7%. The aggregate is "unchanged." The dispersion beneath it has split nearly two percentage points wider in a single month. The ONS publishes the aggregate. It cannot publish the marginal information the aggregate suppresses.
What the Headline Hides
The May 2026 CPIH is a clean test of what a published aggregate can and cannot express. The 3.0% headline is the central tendency of a distribution that has just become bimodal. Goods inflation is decelerating as the supply-chain normalisation that followed the 2021–2023 surge finally works through into shop-floor prices; the post-pandemic restocking cycle, the shipping-cost unwind, and sterling's appreciation have conspired to push goods prices down relative to services. Services inflation is accelerating because services are the segment of the economy where productivity is hardest to lift, where labour is the dominant input, and where the revealed preferences of households are the most informative signal about what the household actually wants. Transport made the largest upward contribution; food and non-alcoholic beverages the largest, partially offsetting, downward. The ONS bulletin is honest. It does not claim that any two-decimal rate is the price the household pays at the next outlet.
The Austrian Diagnosis
Ludwig von Mises's calculation problem, formulated in his 1920 essay "Economic Calculation in the Socialist Commonwealth," is the primary cut. The ONS is not allocating physical capital; it is allocating the published price index that every indexed contract in the country is written against. The 1955 basket weights cannot carry the marginal information of 2026 prices. The basket that mattered in 1955 is not the basket the household is buying now, and the ONS knows this, which is why it reweights periodically. But the reweighting is, itself, a central decision: the ONS's statisticians decide what goes in and what goes out, and the household has no vote on the marginal change. Mises's argument is structural. No central authority, however well-staffed, can compute the right published price of the next purchase at the next outlet. The published price is a single number. The next purchase is a single event. The gap between them is what the household has to manage.
Friedrich Hayek's 1945 "Use of Knowledge in Society" sharpens the cut. The knowledge relevant to the household's real wage is not the data series on the ONS dashboard; it is the dispersed, tacit, locally-held knowledge of the household itself — the next shop on the next Tuesday, the next utility bill, the next renewal on the car insurance. No committee in Newport can aggregate that knowledge in time to publish a two-decimal number that is right at the margin. Eugen von Böhm-Bawerk's theory of time preference, published in 1884, lands the third cut. The 3.0% CPIH is the rate at which the ONS is reporting the public's discount of the future pound. But that rate is not the same in the goods basket as in the services basket. CPIH services rose 3.6% in the year to May; CPIH goods rose 2.0%. The revealed time preference of the household in services is roughly 1.6 percentage points higher than its revealed time preference in goods. The household is signalling, through its actual purchases, that the future pound buys less of what it actually wants than the ONS's single number is willing to admit. The 3.0% headline holds the administered rate down. The 3.6% services print is the revealed rate. The gap between them is the gap between the time preference the ONS is willing to publish and the time preference the household is actually revealing.
Why This Matters for Sound Money
Part 1 of Rails to Freedom — The Foundations of Economic Dysfunction — establishes the diagnostic: price signals are the economy's knowledge system, and any institution that publishes a single aggregate where the underlying reality is a distribution cannot carry the marginal information the household needs. The CPIH is, structurally, the institutional form of the calculation problem at the aggregation step. Mises diagnosed the calculation problem in 1920 because no central authority can compute the right price for capital goods without market prices at the margin. The ONS bulletin in 2026 is the same problem in a different domain: no central authority can compute the right published price of the next purchase without observing every marginal purchase. The published series is two decimals. The marginal signal is dispersed across sixty-seven million households. Part 4 of Rails to Freedom — Broader Implications — identifies the on-chain monetary primitive that does not need a published aggregate to defend its real return. Ethereum's proof-of-stake consensus pays the security budget in ETH; the ETH price is set by the market; the integrity of the ledger does not depend on a committee's published view of what the next inflation print will be. The on-chain real yield is, structurally, the answer to the question the ONS is trying to answer.
What Markets Are Already Doing
The Ethereum tie-in is Lido stETH, a liquid-staking token on Ethereum mainnet. stETH is the structural answer to the ONS's unanswered question. The protocol accepts ETH from any holder and stakes it with over 900 independent node operators, issuing a rebasing token that increases in ETH-denominated balance as validators earn. The yield on stETH is not an administered rate. It is the moving equilibrium of the consensus-layer staking reward, the execution-layer priority fee, the MEV distribution, and the validator-set performance — all priced every block by the chain, with no committee, no forecast, and no two-month lag. The ONS publishes a real-return estimate by subtracting the published CPIH from the published nominal savings rate. The subtraction is two decimals, the inputs are two decimals, and the answer is the published real return. stETH is the on-chain equivalent of the same subtraction, but the inputs are not published — they are continuously-priced market signals, re-priced at every block, against the same ETH-denominated unit the ONS's two decimals are trying to defend.
The secondary tie-in is the MakerDAO Sky Savings Rate on Ethereum mainnet. The DSR is the rate at which holders of the Sky stablecoin (formerly Dai) earn passive yield by depositing into a permissionless vault. No committee sets the rate. The rate is set by the protocol's governance-minimised parameter set, voted on by MKR/SKY holders, and adjusted by the protocol's internal stabiliser mechanism in response to the gap between the secondary-market price of Sky and the dollar peg. The DSR is the on-chain answer to the published savings rate the household gets from a high-street bank — but the rate is not administered by the bank, it is not held down by a base-rate committee, and it is not printed in a monthly bulletin. The structural difference is the point. The ONS will publish the June 2026 figures on 22 July 2026. The chain has already re-priced the return thousands of times. Mises would have recognised the difference immediately: the published aggregate is a forecast by another name; the chain price is the equilibrium.
Looking Ahead
The ONS will publish the June 2026 CPIH on 22 July, and the same dispersion will be visible underneath an "unchanged" or "down one tenth" headline. The pattern is structural. The goods basket is decelerating because supply chains have normalised, the pound is stronger, and the post-pandemic restocking cycle has finally worked through. The services basket is accelerating because the revealed time preference of the household is rising — households are paying more for the next unit of the services they actually want, and the central tendency of the published series is too coarse to carry that signal. The MPC will hold the base rate and call the goods-services wedge "transitory" for the fourth consecutive meeting. The chain will keep pricing the real return at every block, in a unit the household can hold in a wallet it controls, against a settlement the ONS cannot observe. The two-decimal accuracy is not going to improve. The marginal information was never going to fit. The on-chain answer is not a vote against the ONS. It is money that does not need a published aggregate to defend the household's real return. The May 2026 CPIH is the case study. The chain is the answer.