The Regression Theorem¶

This chapter goes deeper into a single, specific debate introduced in Ludwig von Mises and Monetary Theory: does Bitcoin's emergence as a traded, valuable good violate the regression theorem Mises proposed in 1912? This question has generated a genuine, sustained literature within Austrian-influenced economics since 2011, and it is worth treating carefully because it is one of the few places where Austrian economists have directly and publicly argued about Bitcoin's theoretical legitimacy on the school's own terms, rather than simply adopting or dismissing it.

Restating the theorem precisely¶

Mises's regression theorem holds that money's current exchange value must be explained, ultimately, by tracing backward to a point where the good was valued for a non-monetary use, some direct utility independent of its function as a medium of exchange. Without that anchor, Mises argued, the chain of "I value it because others will accept it, who value it because others will accept it..." has no logical starting point and cannot explain why the good has exchange value at all, rather than being circular. See Ludwig von Mises and Monetary Theory for the theorem in its original context.

The objection¶

Economist Robert Murphy raised the core challenge in a 2011 blog post, later developed further in Austrian economics circles: Bitcoin, at the moment it first began trading with a positive price around 2010, had no established non-monetary use of the kind Mises's theorem seems to require, no industrial application, no ornamental use, nothing comparable to gold's pre-monetary utility. If the regression theorem is correct as literally stated, Murphy argued, this should have made it impossible for Bitcoin to acquire exchange value as a medium of exchange in the first place. Yet it demonstrably did. Murphy described this as exposing a "loophole" in Mises's argument: perhaps a good could acquire value purely from the anticipation that it might become a medium of exchange, without ever needing an independent non-monetary use first.

Konrad Graf's response: redefining "use"¶

Economist Konrad Graf, in a paper published in the Quarterly Journal of Austrian Economics ("Bitcoin, the Regression Theorem, and the Emergence of a New Medium of Exchange"), argued the apparent contradiction dissolves once "non-monetary use" is understood the way Austrian subjective value theory actually defines value in general, not as a physical or industrial utility, but as anything a specific individual subjectively values, for any reason particular to them. Graf's argument: early Bitcoin holders valued it directly, before it had significant exchange value, for reasons including its novelty as a cryptographic achievement, the intellectual and technical challenge of participating in a new network, ideological interest in a currency outside state control, and simple curiosity or collector's interest, direct-use motivations that, per subjective value theory, do not need to be shared or recognized by anyone beyond the individual valuing the good, and do not need to resemble gold's ornamental or industrial utility to count as genuine, non-monetary, direct-use value.

Graf argues that "possible direct uses" for early Bitcoin included qualities like mystique, geek appeal, curiosity, and the challenge of participation, subjective sources of direct value that satisfy the regression theorem's requirement without needing to resemble a physical commodity's use. Summarized from Konrad Graf, Bitcoin, the Regression Theorem, and the Emergence of a New Medium of Exchange, Quarterly Journal of Austrian Economics, 2013

Graf's paper also makes a second, distinct point: Bitcoin did not emerge into a pre-monetary barter economy the way Mises's original theorem envisioned gold emerging. It emerged into an economy that already had a fully functioning fiat money system, with established prices for goods and services in that existing money. This meant Bitcoin could be priced and exchanged in terms of an already-existing monetary unit (dollars, for instance) from its very first trades, which Graf argues is a meaningfully different (and easier) starting condition than the one Mises's original theorem was built to explain.

Where the debate stands¶

This is not a settled question even within Austrian economics. Some Austrian-influenced economists have accepted Graf's resolution as sufficient; others maintain that stretching "non-monetary use" to include purely psychological or speculative motivations empties the regression theorem of most of its explanatory content, since almost any speculative asset could then claim to satisfy it retroactively. Mainstream (non-Austrian) monetary economists mostly do not treat the regression theorem as a binding constraint at all, since it is a specifically Austrian theoretical framework rather than a widely shared requirement across the economics profession, meaning this entire debate is, in a sense, internal to one school of thought rather than a universal test Bitcoin must pass to be considered economically legitimate.

This book presents the debate without declaring a winner, because doing so would require taking a side in an active, unresolved dispute among people working within the same theoretical tradition Mises founded, exactly the kind of disputed interpretation this book's historical accuracy commitments require labeling as disputed rather than settled.

Common misconceptions¶

The regression theorem is not empirically about whether Bitcoin has a price today. Bitcoin obviously trades with a price; the theorem is a theoretical claim about how that fact should be explained within Austrian value theory, not a prediction that Bitcoin couldn't have a price at all.

This debate is specific to Austrian economics' internal theoretical framework and should not be confused with a general economic consensus question about whether Bitcoin is "real" money. Mainstream economists asking whether Bitcoin functions as money generally use the functional framework in Functions of Money, not the regression theorem specifically.