The Two Economists: Mises, Hayek and Keynes — What the Austrian School Argued, What Keynes Argued Instead, and Why the Debate Never Ended

• The question: what is an economy, and what should be done about it
• Mises: money is never neutral, and the calculation problem
• Hayek: knowledge is dispersed, and the price system is an information technology
• The boom and the bust, and the politics of the bust
• Keynes at full strength: an equilibrium with involuntary unemployment
• The fight itself: Hayek’s review, Keynes’s reply, and Sraffa on the natural rate
• Why Keynes’s side won
• What actually won was a synthesis
• The close

Two rival accounts of what an economy is were argued out in their own terms in the nineteen thirties, and the one that won the argument was not the General Theory but a synthesis of both.

DeepDives · 2026 · An AI–human collaboration that weighs every source, mainstream or alternative, under strict rules of truth and logic.

§The question: what is an economy, and what should be done about it

The question has been argued for a hundred years and is still unresolved: what is an economy, and what, if anything, should be done about it. Two answers were given in the twentieth century, and the argument between them never ended. It only moved — out of the universities where it began and into the operating systems of every central bank and finance ministry on earth. The first answer came from Vienna, from Ludwig von Mises and his younger colleague Friedrich Hayek, and it said that an economy is a discovery process that no single mind can comprehend, so the attempt to steer it must fail in ways that are not obvious until they are expensive. The second came from Cambridge, from John Maynard Keynes, and it said that an economy can settle into a state of persistent unemployment from which it will not rescue itself, so someone must act.

This examination, the first of two, sets out both theories properly and in their own terms, because they are too often reduced to slogans: the Austrians as people who wanted nothing done, the Keynesians as people who wanted to spend forever. Neither is a fair description, and the argument is much better than the caricature. It then follows the fight itself, which happened earlier and more personally than most accounts allow, and asks why one side became the language of government for the rest of the century while the other spent decades in the wilderness.

Both systems deserve to be stated at full strength, because both were built by people looking at something real and refusing to look away. Mises and Hayek were watching the twentieth century’s great experiment in central planning and concluding, before much of the evidence was in, that it could not work for reasons that had nothing to do with the character of the planners. Keynes was watching the industrial world fall into a slump that lasted years rather than quarters, in the most advanced economies on earth, and concluding that the automatic adjustments the textbooks promised were not happening and might never happen. Neither man was wrong about what he was looking at. That is what makes the argument worth following rather than settling.

§Mises: money is never neutral, and the calculation problem

Start with Mises, because his first important book gives the whole school its shape. He was born in eighteen eighty-one in what was then the Austro-Hungarian empire, taught at the University of Vienna, and published in nineteen twelve a book called The Theory of Money and Credit. The conventional view at the time treated money as a neutral wrapping around real economic activity — something that affected prices but not the underlying structure of production. Mises argued the opposite, and the argument is mechanical rather than moral. Money is the medium in which all other prices are expressed, so a change in the quantity of money does not raise all prices together and evenly. It enters the economy at particular points and particular times, in particular hands, and it travels outward from there. Some prices move first, others later, and the relationships people use to make decisions — between the cost of borrowing and the return on a project, between one industry and another — are distorted on the way. That idea, that money is never neutral and that the point of injection matters as much as the quantity, is the foundation of everything the Austrian school later said about booms and busts, and it anticipates a critique of central banking that became famous much later under a different name.

The second idea is the one that made him famous, and it is worth getting exactly right, because it is commonly misstated as a claim about computers. In nineteen twenty Mises published an article called Economic Calculation in the Socialist Commonwealth, and its argument was not that central planners would lack goodwill or that they would be corrupted. It was that in a system without private ownership of the means of production there are no market prices for capital goods — for the factories, the machines, the raw materials, the half-finished goods — and without such prices there is no way to know whether a particular use of resources is worth more or less than another. A price is not merely a number; it is a ratio that tells you what you must give up. If you cannot form prices for capital goods, you can still decide that a steel mill should be built, and you can still measure the tons of steel, but you cannot know whether the steel and labour and time consumed by that mill would have produced more value somewhere else. You are, in his phrase, groping in the dark.

Notice how much weaker, and more interesting, that claim is than the slogan version. He was not saying that planners are stupid; he was saying that they are deprived of the instrument that makes economic calculation possible, and that the deprivation is not a matter of effort. The collapse of the centrally planned economies after 1989 is not a proof of his method, but it is a very large piece of evidence for his conclusion — a verdict that belongs to the second half of this story, and this examination will come back to it.

§Hayek: knowledge is dispersed, and the price system is an information technology

The third idea is Hayek’s, and it is the deepest of the three. He was born in eighteen ninety-nine, also in Vienna, and he came into Mises’s orbit in the early nineteen twenties when he attended a private seminar Mises ran, which also drew figures like Fritz Machlup and Oskar Morgenstern. Where Mises’s objection was about prices, Hayek’s was about knowledge. In a famous essay in nineteen forty-five called The Use of Knowledge in Society, he asked what problem we are actually trying to solve when we talk about running an economy rationally. The standard answer assumes that all the relevant information exists somewhere, in a form that could be handed to a single authority. Hayek’s answer is that the knowledge required does not exist in that form at all.

The knowledge that matters — the knowledge of the particular circumstances of time and place — is dispersed among millions of people. It is often tacit and unarticulated, it is constantly changing, and much of it is local, and would be worthless if it had to be transmitted to a centre before it could be used. The problem, then, is not computation. It is that the information is never gathered anywhere in the first place.

And the price system, in his account, is the mechanism that solves it without anyone intending to. When a price changes, it transmits a compressed signal about scarcity to everyone who needs it, while demanding almost nothing in return: the buyer does not have to know why the price moved, only that it did and by how much. Hayek described the price system as one of those formations that man has learned to use after he stumbled upon it without understanding it, and the description has aged extremely well in an age of distributed networks. Read today, the essay is less an argument about markets than an argument about information, and it is the part of the Austrian inheritance that the other side of the argument has largely stopped contesting.

§The boom and the bust, and the politics of the bust

From those foundations the Austrians built their theory of the business cycle, and it is the part with the most direct bearing on the world we live in. The argument runs as follows. There is, in a market, a rate of interest that reflects how much people genuinely prefer present goods to future goods — how patient a society actually is. Call it the natural rate. When banks create credit beyond the supply of real savings, the rate at which money can be borrowed falls below that natural rate. But the interest rate is not just the price of money; it is the price that coordinates production over time. A lower rate tells entrepreneurs that long, roundabout, capital-intensive projects are now worth undertaking: build the factory, plant the orchard, start the decade-long drug trial. So the structure of production lengthens. More is invested in the early and middle stages of production, further from the consumer. The economy looks like it is booming, and in the statistics it is.

The trouble is that the boom is a misallocation dressed as prosperity. The projects were chosen on the basis of a borrowing cost that does not reflect what people actually save, so there is not enough real resource available to finish them all. The reckoning arrives when the credit stops expanding, the rate returns toward its natural level, and the projects that only pencilled out at the artificial rate are revealed as unprofitable. That is the bust, and for the Austrians it is not a disease. It is the liquidation of the malinvestment — the economy discovering, painfully, which of the things it built were never worth building.

Mises and Hayek drew a conclusion from this that put them on the wrong side of the politics of their own lifetimes. If the bust is the necessary correction of the boom, then intervening to prevent the correction prevents the recovery and prolongs the disease. The bust should be allowed to work through: prices should be allowed to fall, wages to fall, unprofitable firms to fail, and the capital released to move to whatever uses are actually viable, after which recovery follows. Hayek, writing in the early nineteen thirties with unemployment rising across the industrial world, argued that the depression was the consequence of the credit boom of the nineteen twenties and that the attempt to stop the crash by propping up the distorted structure would simply delay the adjustment. It was, he said, the only way to a lasting recovery.

Whether that was right is the empirical question of the century, and this examination will come to it. But note what the position is not. It is not the claim that nothing bad is happening; it is the claim that the bad thing is the working-off of an earlier bad decision, and that you cannot both cure the hangover and keep drinking. And note what follows from it politically: the programme is austerity during mass unemployment, in the teeth of an electorate that is watching. That is not an incidental feature of the theory. It is, as we will see, one of the reasons the other theory won.

§Keynes at full strength: an equilibrium with involuntary unemployment

Keynes’s central proposition should be stated in the strongest possible form, because it is genuinely radical. He was born in eighteen eighty-three in Cambridge, the son of a distinguished economist, educated at Eton and King’s, a member of the Bloomsbury group, a man who made a fortune in the markets and lost it and made it again, who negotiated for Britain at the peace conference in nineteen nineteen and wrote a bestseller denouncing the terms, and who was a philosopher of probability before he was a policy economist. In nineteen thirty-six he published The General Theory of Employment, Interest and Money, and the word general was an argument in itself: it announced that the old theory was not wrong about the world, but a special case of a more general theory.

The old theory — the classical view — held that an economy finds its way to full employment by itself, because falling wages and prices would restore demand, and because supply creates its own demand. Keynes denied the premise at the level of theory, not merely of policy. His claim was that an economy can rest, indefinitely, at an equilibrium in which willing workers cannot find work. Not a temporary deviation. An equilibrium. If that is true, then unemployment is not a failure of the price mechanism that the price mechanism will fix; it is a failure that requires something outside the mechanism to fix it.

The machinery behind that claim has four moving parts, and each has become part of how the modern world talks about itself. The first is aggregate demand: what an economy produces and sells depends on how much is being spent in total, and spending intentions can fall short of what would be needed to employ everyone. The second is uncertainty in a form that cannot be reduced to risk. In the real world, the future is not a known distribution of outcomes you can average; it is genuinely unknown, and the decisions that matter most — investing in a factory, hiring for a decade — are made by people who know they do not know. Keynes called the resulting impulse animal spirits, and he meant that investment is driven partly by confidence that cannot be justified by calculation alone, which means it can evaporate for reasons that have nothing to do with the available facts.

The third is liquidity preference: in a panic, people want money, not assets, and no interest rate will make them lend if their expectation is that prices will fall. That is why cutting interest rates can fail to revive an economy — the mechanism by which monetary policy is supposed to work goes through the very investor psychology that has collapsed. And the fourth is the multiplier, the idea that a pound of spending does not stop when it is spent: it becomes someone’s income, part of which is spent again, and again, so that a change in spending produces a larger change in output. The corollary is the paradox of thrift — that what is prudent for one household in a slump, to save more, can be ruinous for all of them together, because one person’s saving is another person’s income.

The policy conclusions follow from the mechanism rather than from sentiment. If the problem is a shortfall in total spending, and if the private sector cannot be induced to make up the shortfall, then the government can. It can borrow and spend — and in a slump, with idle resources and unemployed labour, the argument is that it is not competing for resources that anyone else is using. It can run a deficit deliberately, as a tool rather than a scandal, and repay in the upswing. In the last chapter of the General Theory, Keynes went further and said something his followers have spent decades either emphasising or soft-pedalling: that a somewhat comprehensive socialisation of investment would be necessary to secure full employment, and that the rentier — the person living on the return to capital — would in time be euthanised by the fall in the rate of interest. He was not a socialist. He was a liberal who thought capitalism was the best available system, provided it was managed. But he had concluded that the management could not be left to the automatic workings of the market, because the market’s automatic workings had just failed in front of the whole world, for a decade, in the most advanced economies on earth.

§The fight itself: Hayek’s review, Keynes’s reply, and Sraffa on the natural rate

The fight happened earlier and more personally than most people realise. Keynes published A Treatise on Money in nineteen thirty, and in nineteen thirty-one Hayek reviewed it at length in two articles in the journal Economica, arguing that Keynes had neglected the structure of production — that he was treating changes in money expenditure as an instant effect on output, when in reality the effect travelled through the capital structure with an order and a timing that mattered. Keynes replied in the same journal, sharply, complaining that Hayek’s argument was obscure and that his own position had been misread.

Then Keynes did something more effective than replying. He asked Piero Sraffa, an Italian economist at Cambridge with a formidable talent for finding the flaw in an argument, to review Hayek’s own book. In nineteen thirty-two Sraffa published a review of Prices and Production that attacked the very foundation of the Austrian theory: the idea that there is a single, determinate natural rate of interest at all. In a world of many commodities produced over different periods of time, Sraffa argued, there is no such unique rate to be measured — and if the thing that is supposed to be distorted cannot be identified, the theory of distortion becomes impossible to test or even to state precisely. Hayek answered, but the answer did not persuade the profession, and the damage was done. Economists are still arguing about whether Sraffa’s logic was decisive. It is not, however, the reason Hayek lost.

§Why Keynes’s side won

The deeper reason is about what was happening outside the window of the London School of Economics while the two men were publishing in journals. In nineteen thirty-one, when the debate began, unemployment in Britain was already severe, and by the middle of the decade it had reached about one worker in five in Britain and about one in four in the United States, and the depression had been running for years, not quarters. Hayek’s position required the reader to accept that this was the necessary working-out of malinvestment from a boom that had ended in nineteen twenty-nine, and that the remedy was to let it finish. Keynes’s position required the reader to accept that the economy was stuck, that waiting was not a strategy, and that government could shorten the misery. In that context the second argument did not merely win on logic; it won on consequence.

It had another advantage that is easy to underestimate: it came with a diagram and a machine. Within a year of the General Theory, John Hicks had reduced its core to a two-line model — the relationships between saving, investment, money and interest — which could be taught to students, put into equations, and given to governments as a lever. Hayek’s theory of the capital structure resisted that treatment. It was a story about heterogeneous goods, time and dispersed knowledge, and it could not be compressed into a tool that a treasury official could use on a Monday morning.

There is one more reason, and it is the most consequential for everything that follows, so it deserves to be said plainly. Mises’s method ruled out the very kind of argument that was being won. He held that economics is a branch of what he called praxeology — the a priori science of human action — deduced from the axiom that people act purposely toward chosen ends. From that axiom, and its subsidiary facts about the world, you can derive economic law by reasoning, in his view, without needing to test it statistically; indeed he regarded the attempt to test economic propositions with historical data as a category error, because the data are themselves shaped by the theories the people in them hold. Whatever one thinks of that position philosophically, it had a strategic consequence. A framework that declines to be tested cannot be falsified, but it also cannot win the kind of argument that was happening in the nineteen thirties, which was about whether specific policies worked. Keynesianism, by contrast, made claims that were precise, testable and dramatic: the multiplier has a size, the Phillips curve has a slope, the stimulus does or does not pay for itself. That made it vulnerable, when the tests came back badly. But it also made it the only one of the two that could enter a government, be implemented, be measured, and be improved.

§What actually won was a synthesis

And so the second half of the century belonged, in policy terms, to a version of Keynes. What is important to be precise about is that it was not Keynes’s own theory that won, and this is where most accounts become lazy. What won was a synthesis — the neoclassical synthesis, built by Hicks and by Paul Samuelson’s enormously influential textbook of nineteen forty-eight — which claimed that the classical conclusions about markets were right in the long run and in conditions of full employment, while Keynes’s conclusions about demand management were right in the short run, when prices and wages were sticky. That gave the post-war world a comfortable doctrine: use markets for allocation, use fiscal and monetary management for the cycle.

Then, from the late nineteen seventies, came a further transformation: New Keynesian economics, which rebuilt the theory on the assumption that firms and workers have reasons for not adjusting prices and wages instantly, and which accepted the rational-expectations insights of the opposing camp. This is what central banks actually use today. It is recognisably descended from Keynes, but it is not the General Theory, and it concedes a great deal to the people who attacked it.

So the scoreboard is not the one either camp claims. Mises was right about a great deal: that money is not neutral and that its point of injection matters; that capital goods require prices to be allocated rationally; that a system which suppresses prices for the means of production will misallocate resources on a scale that is invisible from inside. Hayek was right about something even more general: that the knowledge needed to run a complex economy is dispersed beyond the reach of any centre, and that the price system is an information technology of astonishing power and frugality. Keynes was right that an economy can be stuck, that waiting for the automatic correction can be catastrophic, and that in a crisis the state is the only actor with the balance sheet to stop a collapse. Three things are true at once: central planning cannot calculate; markets cannot self-correct in every circumstance; and the people who act in a crisis are the people who define the next forty years.

§The close

The examination set out to give both systems their strongest form rather than their slogans, and here is what it found. The Austrians are not a party of doing nothing. Their case rests on three separate claims, and the claims are not equally strong: that money is never neutral and the point of injection matters as much as the quantity; that without market prices for capital goods there is no instrument of calculation, a point about a missing tool rather than missing virtue; and that knowledge of particular circumstances of time and place is dispersed, tacit and local, so that it is never gathered anywhere and the price system is the only thing that transmits it cheaply. The cycle theory then follows from a rate of interest that coordinates production over time, and the politics follow from the cycle theory: if the bust is the working-off of an earlier error, then preventing it prevents the recovery. That position is coherent. It is also, in a slump, an argument for austerity while an electorate watches, which is why it lost.

Keynes is not a party of spending forever. His central claim is a technical one about equilibrium, and it is the claim, not the policy, that does the work: an economy can rest with willing workers unemployed, so the correction has to come from outside the mechanism. The four parts of the machinery, aggregate demand, uncertainty and animal spirits, liquidity preference, and the multiplier with its paradox of thrift, are the reasons the policy conclusion follows rather than being asserted. The argument was decided by consequence rather than by proof — by unemployment of one worker in five, and by the fact that one side offered a diagram a treasury official could use and the other offered a story about heterogeneous capital and time. And it was decided, in the end, by method: a framework that declines statistical testing cannot be falsified, and it also cannot win an argument about whether particular policies worked.

Because an account that cannot be checked is not an account, here is what would falsify this one. If a system that suppressed prices for capital goods had been shown to allocate them as well as a market does, the calculation argument would be refuted. If it could be shown that knowledge of local circumstance is in fact concentrated enough to be handed to a centre, Hayek’s argument would lose its base. If a slump such as the one that began in nineteen twenty-nine could be shown to have corrected itself quickly and at no lasting cost once wages and prices were allowed to fall, the Austrian case against intervention would gain the evidence it has never had. And if involuntary unemployment at an equilibrium could be shown never to occur — if willing workers at going wages could always find work — Keynes’s central proposition would fall at its root. None of those conditions has been met. The next examination asks what happened when Keynes’s answer was put into the world’s institutions, whether the Austrians were right about what it would cost, and what today’s economy looks like from Vienna.

Sources: this examination names its sources inline — Ludwig von Mises’s The Theory of Money and Credit and his article Economic Calculation in the Socialist Commonwealth; Friedrich Hayek’s The Use of Knowledge in Society, his Prices and Production, and the two-part review of Keynes in Economica together with his reply to Sraffa; John Maynard Keynes’s A Treatise on Money and The General Theory of Employment, Interest and Money, including its closing chapter on the socialisation of investment and the rentier; Piero Sraffa’s review of Prices and Production; John Hicks’s two-line reduction of the General Theory and Paul Samuelson’s textbook of nineteen forty-eight, which between them built the neoclassical synthesis; the New Keynesian literature that rebuilt the theory on price and wage stickiness and rational expectations; and the unemployment figures of the nineteen thirties in Britain and the United States, as reported in the historical statistics of the period. Every source is weighed with the same skepticism regardless of faith, following or business model. The audio version of this article, read by DeepDives, is available on Wavlake.