Inflation has many causes. Only one is permanent. The Austrian error is treating the permanent cause as the only cause.
Links: Economics, Business Cycles, Value and Profit, Yarvin x McCormack — Fake Science of Economics, Gauge Theory Applied to Economics — Weinstein x Murphy, Evo-Cap: Optimal Money Supply
“Inflation is always and everywhere a monetary phenomenon.” — Milton Friedman
Many Austrians adopt this line or go further: some define inflation as monetary expansion, making price increases a secondary effect they call “price inflation.” This definitional move makes their claim tautologically true — if inflation means money supply growth, then of course only money supply growth causes inflation. But it’s empirically empty. It tells you nothing about why the prices at the grocery store went up this month.
The corrected view: monetary expansion is the only permanent source of inflation. But it is not the only source. Multiple forces cause prices to rise, and they operate on different timescales, through different mechanisms, with different resolutions.
The Austrian/monetarist core claim is correct as far as it goes: if you increase the money supply faster than the economy’s productive output grows, you get sustained price increases. More money chasing the same goods.
This is the only inflationary force that is:
The Cantillon effect (see Evo-Cap: Optimal Money Supply) adds a distributional dimension: new money enters through banks and government, enriching early recipients before prices adjust. So monetary inflation is not just a general price increase — it’s a wealth transfer from those far from the spigot to those near it.
Why it’s permanent: Every other inflationary force below eventually resolves — supply recovers, demand shifts, shocks dissipate. But once the money is printed, it stays in circulation. The price level ratchets up and stays up. This is why the Austrians focus on it, and they’re right to — it’s the most dangerous source precisely because it doesn’t self-correct.
When the supply of a critical good drops — oil embargo, pandemic disruptions, war destroying productive capacity, natural disaster — prices rise. No monetary expansion required.
Examples:
The Austrian objection: Supply shocks cause relative price changes, not general inflation. Oil goes up, but if the money supply is fixed, other prices should fall to compensate — total spending power hasn’t changed, it’s just been reallocated.
Why the objection fails in practice:
Why it’s transient: Supply eventually recovers, substitutes emerge, or demand adjusts. The 1973 oil shock led to fuel-efficient cars, North Sea drilling, and conservation. The 2020 supply chain crisis resolved as shipping normalized. The inflationary impulse fades when the underlying constraint is relieved.
The danger: when government responds to a supply shock with monetary expansion to “cushion the blow,” the transient shock becomes entrenched. The supply-side cause resolves, but the printed money stays. This is exactly what happened in the 1970s — the oil shock was real, but the sustained inflation that followed was the Fed’s monetary response.
When demand for a good or class of goods surges — due to new technology, cultural shifts, government mandates, or speculation — prices in that sector rise.
Examples:
Why it’s transient: Demand shifts trigger supply responses. High rare earth prices incentivize new mining, recycling, and substitute materials. High housing prices incentivize construction. The market’s price signal does exactly what it’s supposed to: attract capital to where it’s most needed. The inflation is the signal, not the disease.
The quantity theory of money: MV = PQ (money supply × velocity = price level × real output). The Austrians and monetarists focus on M. But V — the speed at which money changes hands — matters too.
Austrians tend to treat V as roughly stable. It isn’t — it can swing dramatically in both directions, especially during crises. The deflationary shock of 2008–2009 was largely a velocity collapse: banks stopped lending, consumers stopped spending, and the Fed’s massive M expansion barely offset the V decline. This is why the predicted hyperinflation from QE never materialized in consumer prices — M went up, V went down, and PQ stayed roughly flat (though asset prices, per Yarvin’s Z1 argument, did inflate enormously — see Yarvin x McCormack).
If the economy’s ability to produce goods declines — through regulatory burden, war, brain drain, resource depletion, or institutional sclerosis — more inputs are needed to produce the same output. Prices rise.
This is the quietest form of inflation and the hardest to measure because it’s not a sudden shock but a slow erosion. It shows up as things gradually getting more expensive in real terms despite no obvious crisis.
Examples:
Why it’s structural but not permanent in the monetary sense: Productivity decline can reverse — deregulation, technological breakthroughs, institutional reform. It’s not self-correcting in the way supply shocks are, but it’s not irreversible in the way monetary expansion is (you can’t un-print money, but you can un-write regulations).
A special case worth calling out separately: technology doesn’t just shift demand — it creates and destroys it. Things that were worthless become essential. Things that were essential become obsolete. Both directions produce price changes that have nothing to do with the money supply.
Technology creates demand (inflationary):
Technology destroys demand (deflationary):
Technology does both simultaneously:
This is real inflation (and deflation) in the sense that it genuinely changes what you need to spend. But it’s not monetary — it reflects real changes in the world. The Austrian framework, focused entirely on money supply, has nothing to say about any of this except to deny it’s “really” inflation. Which is precisely the definitional move that makes their theory unhelpful.
The broader point: technology is constantly reshuffling which things are valuable and which aren’t. This produces a continuous churn of sectoral inflation and deflation that is completely invisible to monetary theory. No amount of money supply control will prevent lithium from spiking when everyone wants EVs, or whale oil from collapsing when kerosene arrives.
| Source | Timescale | Self-correcting? | Amplified by monetary policy? |
|---|---|---|---|
| Monetary expansion | Permanent | No — money doesn’t un-print | N/A — it IS monetary policy |
| Supply shock | Months to years | Yes — supply recovers or substitutes emerge | Yes — if government prints to cushion the shock |
| Demand shock / FOMO | Months to years | Yes — supply response + speculation collapse | Yes — cheap credit fuels the FOMO |
| Velocity change | Variable | Partially — confidence returns, but slowly | Yes — rate policy directly targets velocity |
| Productivity decline | Decades | Possible but requires structural reform | Yes — printing masks the decline instead of forcing reform |
| Technology-driven demand | Permanent shift, price stabilizes | Partially — supply scales, but baseline is higher | Modestly — not the primary driver |
The key insight: every non-monetary source of inflation is either transient or addressable through real-economy adjustments. Monetary expansion is the only source that ratchets permanently and never self-corrects. This is why the Austrians focus on it. Their error is treating the most important cause as the only cause.
→ Full treatment: Measuring Inflation — Why Disaggregation Doesn’t Save You.
The short version: “the price level” is a fiction, because CPI requires choosing and weighting a basket and the weighting is subjective (The Weighting Problem; Weinstein × Murphy). Different baskets, different weights, different “inflation.” Consequences: CPI understates asset inflation (post-2008 printing went into equities, real estate and bonds — Yarvin’s Z1 argument), overstates technology deflation via hedonic adjustment (the Boskin manoeuvre), and hides enormous sectoral and household-level divergence — measured household inflation rates have an annual interquartile range of 6–9 percentage points.
When Austrians say “inflation is only monetary,” they’re making a claim about a quantity that has no unique objective definition — ironic for a school that prides itself on methodological rigor.
And the obvious fix fails: personalizing the basket relocates the problem rather than solving it, because one person still buys many goods and their own basket still changes over time. The index-number problem reappears at n = 1. See the linked page for why the Weighting Problem is scale-invariant.
| Claim | Orthodox Austrian/Monetarist | Corrected view |
|---|---|---|
| Only monetary expansion causes inflation | Yes | No — but it’s the only permanent cause |
| Supply shocks are “not really inflation” | Yes (relative price changes only) | Wrong — they’re real inflation with real effects, just transient |
| Technology-driven price increases aren’t inflation | Definitionally excluded | They’re real price increases that affect real purchasing power |
| “The price level” is a meaningful concept | Assumed | No — it requires subjective weighting (Weinstein/Murphy) |
| Control the money supply, control inflation | Yes | Partially — you eliminate the permanent source but not the transient ones |
| Velocity is stable | Approximately | No — it swings dramatically and can offset monetary policy entirely |
The orthodox position says: control M, control inflation.
The corrected position says: controlling M eliminates the only permanent ratchet, which is the most important thing you can do. But you’ll still have transient inflation from supply shocks, demand shifts, velocity swings, and productivity changes. These resolve naturally if you don’t make the critical error of printing money to paper over them — which, per the political ratchet in Business Cycles, is exactly what governments always do.