Catherine Mann and the Knowledge the Bank Cannot Read
On 2 July 2026, Catherine L. Mann — an external member of the Bank of England's Monetary Policy Committee since September 2021, and before that the chief economist at Citi — delivered a fireside chat at the Natixis CIB Private Debt Forum in which she made a quietly radical argument. The headline numbers on UK inflation are not enough. The disaggregated, firm-level price data on the Bank's Decision Maker Panel carry information that aggregate statistics cannot capture, and her 2026 policy decisions have turned on that information more than on the published CPI. The lecture is sober, technical, and well within the mainstream. It is, without anyone naming him, Friedrich Hayek's "The Use of Knowledge in Society" applied to monetary policy in real time.
What Mann Actually Said
Mann opened by recounting her own evolution in the first half of 2026. At the start of the year, moderation in price and wage growth had moved her closer to voting for a reduction in Bank Rate. Then came the conflict in the Middle East — what the Bank's Monetary Policy Reports since April 2026 call "the Iran war" — and the trade-off between inflation and activity re-emerged. Her view shifted toward "a longer hold, and potentially a need to lean against that risk." That shift was driven less by the CPI print than by the firm-level data: state-dependent pricing (the practice of raising prices in response to observed cost or demand conditions rather than at fixed intervals) spreading, one-year-ahead price expectations shifting rightward, and the wedge between whole-economy and private-sector pay growth persisting.
She then walked the audience through her evidence base. She cited the Bank's working paper on state- and time-dependent pricing (Bunn et al., 2026a), which finds that a higher share of UK firms are now adjusting prices in response to incoming shocks rather than on calendar. State-dependent firms raised prices faster when inflation rose but also raised them by less when inflation fell, because their time-dependent peers were still catching up. Mann cited this as evidence of an "upside inflation bias" — a tendency of the price level to respond more aggressively to positive shocks than to negative ones — that aggregate CPI will smooth away. She cited the Decision Maker Panel's distribution plots, showing firms' one-year-ahead CPI expectations have moved rightward since the war without yet matching the dispersion of 2022. And she cited the wage wedge: private-sector regular pay has fallen within the Bank's target-consistent range, while whole-economy pay has not, because around 80 percent of the private-sector workforce sits in collectively-bargained or cost-minimising segments where wage-setting is stickier than the headline data suggest.
Why This Is a Hayekian Argument
Friedrich Hayek (1899–1992), a Nobel-prize-winning economist who spent his career arguing that useful economic knowledge is dispersed across millions of individuals and cannot be collected by any central authority, made exactly this point in September 1945. His essay "The Use of Knowledge in Society" argued that the statistical aggregates a central planner sees are not the economy. They are summaries of actions taken by individual actors who each know their own costs, customers, and timing. The plan cannot access that knowledge directly. It can only infer it from the prices those actors post. Mann's argument is the same argument, applied to a single central-bank instrument: the Decision Maker Panel. The survey can ask a firm what its price-setting behaviour looks like. The firm can answer. But the firm also holds private knowledge — about its margin position, its customer mix, its exposure to imported energy — that no survey instrument can extract. The price the firm actually posts carries more information than the survey response. Mann has reorganised her decision-making around the price, not the survey.
Ludwig von Mises (1881–1973), the Austrian economist who argued that without genuine market prices no one can rationally calculate whether resources are being allocated well, sharpened the cut. The calculation problem is not only about steel or wheat. It applies to any decision in which the decision-maker lacks a market-generated price signal and substitutes a survey, a model, or a committee judgement. Mann's argument is that the MPC, on the published CPI alone, would have under-weighted the upside risks to inflation in 2026 — read the headline moderation at face value, and moved toward a cut that the firm-level distribution does not support. The disaggregated data is her fix. It is an admission that the aggregate number, alone, is not the price the calculation problem requires.
The Time-Dependent versus State-Dependent Cut
Mann's most interesting empirical move is the distinction between time-dependent firms (those who reprice on a fixed calendar regardless of incoming shocks) and state-dependent firms (those who reprice when costs or demand move). Before 2022, the bulk of UK firms were time-dependent. Since 2022, the share of state-dependent firms has risen sharply. State-dependent firms amplify positive shocks and dampen negative ones. Translated into a macroeconomic variable, this means the price level has an asymmetric response to shocks — it goes up faster than it comes down — the mechanism Mann uses to justify "a longer hold." She does not use the language of Austrian business-cycle theory. She does not need to. The mechanism is the same one Böhm-Bawerk (1851–1914), the Austrian economist who founded capital theory and the analysis of time preference, described in his critique of artificial credit expansion: an asymmetric adjustment process, set in motion by an external shock, that creates a structural bias in the price level until the distortion unwinds.
The Information That Cannot Be Surveyed
Read Mann's speech for what it is — a working central banker making the best case she can with the data the Bank publishes — and there is still an unresolved gap. The Decision Maker Panel is a survey of roughly 2,800 UK firms. It produces distributional evidence on price expectations and price-setting behaviour. It does not, and cannot, read the actual prices each firm posts at the moment a shock arrives. The information Mann most wants — whether the marginal firm is repricing now, by how much, on what expectation — is held by that firm alone, revealed only through the price it posts. The Bank can observe the resulting transaction. It cannot reconstruct the private knowledge that produced it. This is the knowledge problem (Hayek's claim that useful economic information is dispersed, tacit, and locally held, and cannot be transmitted to any central authority by statistical reporting alone) in its purest operational form.
What a Market Does That a Survey Cannot
The tool the Bank of England does not use is the one the Austrian framework predicts will work: a continuous, priced, decentralised aggregation of beliefs. Polymarket — the world's largest active prediction market — runs on Polygon, a public Ethereum Layer 2 that settles every contract back to Ethereum mainnet for security. It hosts markets whose outcomes are the same ones the MPC is trying to forecast: whether UK CPI will print above or below specific thresholds in coming months, whether the next MPC vote will be a hold or a hike. The price of a YES contract on any of these questions is not a survey median. It is the point at which the marginal informed buyer will take the other side — and that price updates with every trade, against a verifiable public resolution source. The Decision Maker Panel asks firms what they expect. Polymarket asks anyone with money and a view to put both on the table. The latter aggregates more information per unit time, without anyone deciding which response to trim.
Part 3 of Rails to Freedom makes the same structural argument for Ethereum generally. Decentralisation is not a technical feature; it is the property that lets dispersed knowledge act on a shared state without being collected by a central counterparty. Chapter 7 returns to Hayek's spontaneous order — the insight that complex beneficial coordination can emerge from millions of independent decisions without anyone designing it — to argue that what is true of market prices is true of programmable settlement: a public base layer lets private knowledge become public information, by action, at the margin. Mann's speech is, almost despite itself, a case for why the price is the data — and why a bank that surveys instead of observing must argue, in technical footnotes, for the relevance of its own survey.
The Lesson for the MPC — and for Sound Money
Mann is not asking for abolition of the Bank of England. She is asking that the Bank treat its firm-level data with the seriousness the published aggregates do not deserve. The Austrian diagnosis goes one step further. If the price is the data, and the data is dispersed across millions of price-setters, then any institution whose job is to read that data should be built on the principle that the price is the most honest signal available. Decentralised prediction markets on Ethereum settlement are exactly that principle, operationalised. They will not replace the MPC. But they will increasingly price the questions the MPC is asked to answer, in public, on chain, with the kind of asymmetric-information weighting that Mann's firm-level evidence has only begun to approximate.
On 6 August 2026 the Bank will publish its next Inflation Report. On the same morning, a Polymarket contract on whether UK CPI will print above or below the Bank's central projection will have updated dozens of times. The latter is the disaggregated, real-time, continuously priced equivalent of the Mann speech — and, structurally, what Hayek described in 1945: an institution that lets dispersed knowledge act on prices without anyone collecting it.