The science of human action

The following series of articles presents the Austrian School of economics, 1000 words at a time. Nine economists. Twenty-seven articles. One coherent tradition that the establishment has been trying to ignore for 150 years. They were right.

The unpaid seminar that built a movement: Mises at NYU. (Article 7b)

New York City. Fall, 1950.

The course is listed in the NYU Graduate School of Business Administration catalog as “Interventionism and the Profit System.” The professor’s name means nothing to most students who read it. A few sign up out of scheduling convenience. What they find when they arrive is a 69-year-old Austrian in a formal suit, standing at the front of a plain seminar room, about to explain why almost everything they have been taught about economics is wrong.

Ludwig von Mises teaches at NYU for twenty-four years. He gives his first lecture there at sixty-four. His last at eighty-eight. For the entirety of that period, New York University pays him nothing. His salary is covered by the William Volker Fund (a private foundation) because the university is happy to have him teach, provided it costs them not a penny. Among the institutions of higher learning that host him across his career, both the University of Vienna and NYU distinguish themselves, as one student puts it, by tolerating his teaching “provided it did not cost them a penny.”

His wife Margit attends nearly every session. Around the seminar table on any given evening: Henry Hazlitt, who has already written the most successful popular economics book of the century. Murray Rothbard, twenty-something, taking notes with a ferocity that suggests he intends to build a cathedral from the foundations being laid in this room. Israel Kirzner, who will spend the next forty years extending one corner of what Mises teaches here into a complete theory of entrepreneurship. Percy Greaves, George Reisman, Ralph Raico, Leonard Liggio, a generation of economists and scholars who carry the tradition forward long after the seminar ends. Rothbard later called it “the most exciting intellectual adventure of my life.”

Two miles north, Columbia University’s economics department runs econometric models. Across the Atlantic, the profession celebrates British economist John Maynard Keynes and builds mathematical frameworks for managing aggregate demand. The mainstream is moving, confidently and unanimously, in a direction Mises is certain is wrong.

He keeps teaching.

One axiom, an entire science

The central claim of Human Action, the one that separates it from every other work in economics, is that the entire structure of economic theory can be derived from a single, self-evident truth.

Humans act purposefully.

Mises calls this the action axiom, and he is precise about what it means. Action is not mere behaviour. A stone falling is not acting. A dog scratching an itch is not acting in the relevant sense. Human action is the conscious deployment of means to achieve ends; the deliberate choice of one course over another, driven by a judgement that the chosen course will move the actor from a less satisfactory state to a more satisfactory one.

This axiom is not an empirical claim subject to testing. You cannot falsify it with data, because any attempt to falsify it requires you to act purposefully in order to do so. It is, in the philosophical sense, a priori; prior to experience, known with certainty from the nature of consciousness itself. It is the only foundation solid enough to build a science on.

From this single starting point, Mises builds outward by pure logic. Action implies scarcity: if there were no scarcity of means, no choice would be necessary. Scarcity implies cost: every choice forecloses alternatives. Choosing among alternatives implies ranking; the actor values some ends above others. Ranking implies that value is subjective and ordinal, not objective and measurable. From here the entire structure unfolds: exchange, the division of labour, prices as signals, capital as deferred consumption, interest as the price of time, money as the most saleable commodity. This is all derived, step by logical step, from the single premise that humans act purposefully.

Mises calls this method praxeology. The science of human action. Economics, in his framework, is the branch of praxeology that studies action conducted under conditions of monetary exchange. Its theorems are not generalizations from observed data, but rather logical deductions, as certain as geometry, that hold wherever human beings act, which is everywhere, always.

The three stages Mises describes in every act of human action: first, unease. The individual feels dissatisfied, a discomfort that generates the impulse toward change. Second, perception. Identifying what action might relieve the unease or improve the situation. Third, substitution. Choosing the action, trading a less preferred state for a more preferred one. Every transaction you have ever made, every choice between options, every plan formulated and executed, fits this structure. The man who wakes up hungry and decides to make breakfast. The entrepreneur who sees a gap in the market and builds a business. The government that taxes one group to subsidize another. All of it praxeology. All of it human action.

What the equations cannot capture

By the time Human Action is published in 1949, economics is well along in its transformation into a mathematical discipline. Paul Samuelson is rewriting the field’s foundations in terms of constrained optimization. The Keynesian framework is being formalized into equations. Econometrics (the statistical analysis of economic data) is becoming the benchmark for serious academic work. The prestige of physics, borrowed.

Mises finds this development not impressive but alarming.

Mathematical economics, he argues, mistakes the map for the territory. An equation can model an equilibrium: a state where all plans are mutually consistent, all markets clear, and all information is known. What it cannot model is the process by which that equilibrium, if it is ever reached at all, comes about. It cannot represent the entrepreneur who discovers a profit opportunity that nobody knew existed. It cannot represent the moment a consumer changes her mind. It cannot represent the flood of information that flows through prices in real time, generated by millions of individual decisions and updated continuously as circumstances change.

The economy is not a system tending toward a computable optimum. It is a process: restless, creative, perpetually unfinished. Driven by human action in conditions of genuine uncertainty. The entrepreneur does not optimize a known function. He acts on a judgment about an uncertain future, using knowledge that is partly tacit, partly local, partly a hunch refined by experience. His action generates new information that changes the situation for everyone else, requiring further action, generating further information, in a process that never settles.

You cannot put this in an equation. The information the market generates didn’t exist before the actions created it. A model built on historical data captures what people did under past conditions. It tells you almost nothing about what they will do under new ones, because new conditions generate new knowledge, new preferences, new opportunities that no past data could have anticipated.

The consequences of this error are not merely academic. Governments armed with mathematical models believe they possess knowledge they do not have, that they can manage, optimize, and fine-tune what is in fact a self-regulating process that operates on information no central authority can access or replicate. Every technocrat who has ever claimed to know the “right” interest rate, the “correct” exchange rate, or the “optimal” level of government spending is making an epistemological claim that Human Action shows is incoherent.

Mises wrote this clearly and at length in 1949. The profession responded by building more elaborate models and ignoring him more comprehensively.

The room where it was kept alive

Here is what makes the NYU seminar remarkable. Mises was not, by the early 1950s, a figure the mainstream engaged with. He was too old, too European, too uncompromising, too hostile to the mathematical turn, and too opposed to everything the profession’s most powerful figures had staked their careers on. He had no institutional support, no research budget, and no PhD students funded by grants. He had a room, a handful of attendees, and twenty-four years of evenings.

What came out of that room: Murray Rothbard, who reconstructed the entire Austrian system from first principles in Man, Economy, and State, which Mises himself called “an epochal contribution.” Israel Kirzner, who built the Austrian theory of entrepreneurship into one of the most important strands of thought in modern economics. A generation of scholars who founded and staffed the institutions (The Foundation for Economic Education, the Cato Institute, the Mises Institute) that kept Austrian economics alive through the decades when the mainstream had no interest in it.

The Volker Fund stopped paying Mises’s salary in 1962. Other foundations stepped in. He kept teaching until 1969. He died four years later, in October 1973, as the stagflation that Keynesian economics said was impossible was proving, month by month, that someone had got something badly wrong.

The economists who ran economic policy in the 1960s and 70s had mathematical models. They had government appointments, Nobel Prizes, and the full institutional weight of the academy behind them. They did not have what Mises had: a correct theory of how economies actually work, taught in a seminar room nobody paid for, to a dozen students who went on to change the intellectual landscape of the late twentieth century.

Human Action is still in print. It has never been out of print. The models that were supposed to make it obsolete have been revised, extended, complicated, and quietly abandoned, one after another, as the phenomena they were supposed to explain kept failing to behave as predicted. The action axiom has not required revision. It was true when Mises stated it in 1949. It is true now.