Does Bitcoin Satisfy the Regression Theorem?

Mises’s regression theorem holds that a good can only become money if it already had exchange value from some non-monetary use. Bitcoin had no industrial or commodity use of the kind gold has, yet it grew from nothing into a widely held money — which is why it is the theorem’s most-argued modern test case. This page states the challenge and the three main responses — the skeptic’s literal reading, the defender’s “some first non-monetary demand” reading, and a property-theory sidestep — and takes no side. The dispute is real and unresolved; what both camps can accept is that the theorem is an entry claim, and Bitcoin’s hardness stands on its own whichever way it goes.

The challenge

The theorem’s payload is a hard corollary, stated by Ludwig von Mises in Human Action:

…no good can be employed for the function of a medium of exchange which at the very beginning of its use for this purpose did not have exchange value on account of other employments.

— Ludwig von Mises, Human Action

Gold satisfies this easily: it was prized as ornament and industrial metal long before it circulated as coin, so its monetary value has a non-monetary starting point to regress to. Bitcoin appears not to. It is a purely digital token with no ornamental or industrial employment; on 3 January 2009 it existed only as a running protocol. If a thing with no prior non-monetary use nonetheless became a medium of exchange, then either the theorem is false, or Bitcoin is not money in Mises’s sense — or the premise, that Bitcoin had no prior non-monetary value, is wrong. Each of these is a way the case has been read.

The skeptic reading: a literal counterexample

The straightforward Austrian objection takes the corollary at face value. Money, on the theorem, must trace back to a good valued for consumption or use. Bitcoin has no such use to trace back to; therefore, on the strict criterion, it either cannot be sound money or exposes a limit of the theorem. Read literally, the corollary tells against Bitcoin: its early history supplies no non-monetary prior use of the kind Mises required. On this reading the burden is on the defender to name the non-monetary value that Bitcoin had before anyone accepted it in exchange — and to show it was value for something other than the expectation of future exchange, which would beg the question.

The defender reading: some first demand, not an industrial one

The defenders’ move is to read the corollary as requiring merely some first non-monetary demand — a reason to hold the good other than to trade it onward — not specifically an industrial or ornamental one. And they argue Bitcoin had it. In its first months Bitcoin was not yet a medium of exchange; it was a curiosity held by a handful of cryptographers for its novelty, its ideological payload (a censorship-resistant, non-state settlement network), and the technical interest of running it. The Bitcoin Standard records that in those months early coins “began to acquire the status of collectibles, albeit in digital form” before the first market price was ever set. On this reading that collectible-and-ideological demand is the non-monetary first layer, and the regress terminates there rather than in a metal’s industrial use.

Nick Szabo supplies the theoretical version of that first layer. His Shelling Out account of prehistoric collectibles — valued as durable, costly, hard-to-forge tokens long before coinage — is exactly the category the defender assigns Bitcoin: a digital collectible whose early worth was non-monetary, from which monetary demand later grew.

Saifedean Ammous carries the Mengerian–Misesian apparatus into the Bitcoin case. In Principles of Economics he restates the theorem as an account of how monetary demand accretes onto a market good — the theorem explained “how a normal market good can develop into a monetary good when it acquires monetary demand, thus raising its value and increasing its salability” — and he treats Bitcoin’s rise as a market phenomenon of the very kind Menger and Mises describe:

Another powerful refutation of the statist theories of money comes from the emergence of bitcoin, which in the last 14 years has grown from nothing to become one of the world’s 20 largest currencies, all without a single legal authority promoting or decreeing its use.

— Saifedean Ammous, Principles of Economics

For Ammous the pre-monetary layer is exactly that early collectible and network value, together with Bitcoin’s strict supply schedule. The unforgeable costliness of proof-of-work then explains why the token could hold saleability across time once monetary demand arrived, but it is the early non-monetary demand, not the mining cost, that the defender offers as the regress terminus.

The sidestep: ownable property, not prior value

A third line questions whether the regression theorem is even the right lens. Konrad Graf, in Are Bitcoins Ownable?, analyzes Bitcoin as ownable digital property through rivalrous, UTXO-level control and a digital-commodity framing, without making the regress the primary anchor of its monetary status. If Bitcoin’s economic standing can be established on property-theoretic and empirical grounds, then whether it fits the historical entry-condition of the regression theorem becomes a narrower question about the theorem’s scope rather than a verdict on Bitcoin.

Why it may not decide much

Both camps can agree on a deflationary point that limits the stakes. The regression theorem, as Mises insisted, is an entry or existence theorem — it “merely explains how a new kind of media of exchange can come into use and remain in use” — not an account of the daily determination of money’s value. It is a claim about how the chain of purchasing power must have begun, not a law that governs a monetary good’s present standing. Bitcoin’s monetary properties that people actually trade on — its strict supply schedule, its saleability across space and scale, and the unforgeable costliness securing it — are established by its design and its market, and hold whether or not the historical regress is judged to terminate cleanly. So a skeptic can doubt that Bitcoin fits the strict corollary while still granting it is hard money on other grounds, and a defender can win the regress point without that settling Bitcoin’s day-to-day value. The contested question is real, but narrower than it first looks.

Where it stands

The dispute turns on a genuine ambiguity in the theorem itself: does “exchange value on account of other employments” mean value from industrial or consumption use specifically, or value from any demand that is not the anticipation of future exchange? Read strictly, Bitcoin looks like a counterexample; read loosely, its early collectible-and-ideological demand satisfies the requirement and the theorem survives intact. The strict reading and the loose reading each have a claim on Mises’s words, the primary texts do not obviously decide between them, and this page records the dispute rather than resolving it — confidence is medium. A narrower point follows from Mises’s own text whichever way the case is scored: the regression theorem is an entry/existence claim, and the hardness criterion it scaffolds can be evaluated separately from it.

Sources

Human Action (Full Text) - Mises’s corollary that no good can serve as a medium of exchange without prior non-monetary exchange value, and the statement that the theorem explains only the entry of a new money, not its daily determination
The Bitcoin Standard (Full Text Aggregate) - Ammous’s history of Bitcoin’s early months, in which coins “began to acquire the status of collectibles, albeit in digital form” before the first market price, and the salability-across-time framing of hardness
Principles of Economics (Full Text) - Ammous’s restatement of the regression theorem as monetary demand accreting onto a market good, and Bitcoin’s stateless emergence as a refutation of statist money theories
Shelling Out (Full Text) - Szabo’s collectibles account — durable, costly, hard-to-forge tokens valued before coinage — the defender’s non-monetary first layer for Bitcoin
Are Bitcoins Ownable? (Full Text) - Graf’s action-based property analysis treating Bitcoin as ownable, rivalrous digital property — the sidestep of the regression theorem

Galtland Index

Explorer

Does Bitcoin Satisfy the Regression Theorem?

Does Bitcoin Satisfy the Regression Theorem?

The challenge

The skeptic reading: a literal counterexample

The defender reading: some first demand, not an industrial one

The sidestep: ownable property, not prior value

Why it may not decide much

Where it stands

See Also

Sources

Graph View

Table of Contents

Backlinks