Inflation, Communication, and Noise | Mises Institute

In 1948, Claude Shannon published “A Mathematical Theory of Communication” in the Bell System Technical Journal, a paper that established information theory as a formal discipline. Shannon’s central contribution was to show that information can be measured, that communication channels have finite capacity, and that—when noise is introduced into a channel—the receiver’s ability to reconstruct the original message degrades in precise, calculable ways. The paper was concerned with telephone wires and radio signals, but its logic applies to any system in which one agent transmits information to another—including the system through which millions of economic actors coordinate their plans through prices.

A market price is an information packet. It encodes the scarcity of a resource, the cost of producing it, the state of demand from consumers, and the expectations of everyone who has traded it recently; all compressed into a single number that any buyer or seller can act on without knowing any of the underlying details. Friedrich Hayek’s 1945 essay “The Use of Knowledge in Society” described how this compression works: no central planner can gather and process the dispersed local knowledge that prices summarize, which is why decentralized markets outperform central planning at allocating resources. What Hayek described was, in the language Shannon would formalize three years later, a distributed information network operating near its theoretical efficiency limit.

When a central bank expands the money supply, it does not add real resources to the economy. It adds a signal that mimics the appearance of additional resources. From the perspective of an entrepreneur receiving a loan at an artificially-low interest rate, the price signal reads: “real savings are available, real resources can be committed to long-term projects.” This is the same signal that would appear if consumers had genuinely increased their savings and made those funds available for investment. The signal is false—the resources it implies do not exist—but the entrepreneur, acting rationally on the information available, responds as if they do.

This is not a metaphor for what inflation does; it is the literal mechanism. Monetary expansion injects noise into the price signal at the most fundamental layer of the economy—the interest rate, which coordinates the allocation of resources between present consumption and future investment. When that signal is corrupted, the errors propagate through every decision that depends on it. Businesses invest in capital projects that require a supply of complementary resources that will not materialize. Workers are hired into industries that will later contract. Long chains of production are initiated that cannot be completed at the prices prevailing when the boom began.

Ludwig von Mises described this mechanism in detail in his 1912 work The Theory of Money and Credit, and Hayek elaborated on it through the 1930s. The Austrian business cycle theory, as it became known, holds that credit expansion by central banks systematically falsifies the interest rate signal, producing a boom characterized by widespread malinvestment—investment in projects whose profitability depends on conditions that monetary expansion has created artificially and cannot sustain permanently. When the expansion slows or reverses, the false signal corrects, and the investments built on it become unprofitable simultaneously. The cluster of failures that results is what we call a recession.

The distributional consequences of this process were identified even earlier, by the 18th-century economist Richard Cantillon. Writing in his Essai sur la Nature du Commerce en Général around 1730, Cantillon observed that when new money enters an economy, it does not arrive uniformly across all prices and all actors. It enters at specific points—in the modern context, through the banking system and financial markets—and ripples outward from there. Those who receive the new money first can spend it before prices have adjusted to reflect the increase in supply. Those who receive it last—workers whose wages are renegotiated slowly, savers whose deposits are denominated in nominal terms—face higher prices before they see any increase in their nominal income. The monetary expansion transfers purchasing power from the latter group to the former, without any legislative act and without any of the visibility that a direct tax would carry.

This is why the claim that central banks can manage inflation as a policy instrument—keeping it at a “target” of two percent while generating full employment and stable growth—misunderstands what inflation is. Inflation is not a dial that can be turned up or down to achieve desired macroeconomic outcomes, it is the predictable consequence of introducing noise into an information system, and the damage it causes is not proportional to the rate but to the disruption of the coordination that accurate prices make possible. An economy running on two percent inflation is not an economy with a minor, tolerable distortion. It is an economy in which the interest rate signal is chronically falsified at a low level, producing chronic malinvestment at a scale that recessions periodically reveal and correct.

The Federal Reserve, the European Central Bank, and their counterparts do not cause inflation through incompetence or malice. They cause it because the institutional mandate they operate under—stabilize prices, maximize employment, support growth—requires them to intervene in the price system, and intervention in the price system means corrupting the information it carries. There is no version of central banking that avoids this problem, because the problem is structural. A central bank that sets interest rates is, by definition, overriding the rate that would emerge from the voluntary decisions of borrowers and lenders, and the rate that emerges from those decisions is the only rate that accurately reflects the time preferences and resource availability of the economy at that moment.

Shannon’s noisy-channel coding theorem showed that any communication system can transmit information reliably only up to the channel’s capacity, and that noise reduces that capacity. The theorem does not care about the intentions of whoever introduced the noise. It applies to telephone cables and to monetary systems with equal indifference. The price system has a finite capacity to coordinate economic activity, and monetary expansion reduces that capacity by degrading the quality of the signal it carries. The consequences—malinvestment, redistribution from savers to asset holders, boom-bust cycles—are not policy failures that better management could avoid. They are the physics of a noisy channel.