THE Mirage
OF THE FREE
MARKET
Nine ideas that Milton Friedman gave the world, and what fifty years of evidence did to them.
“The great virtue of a free market system is that it does not care what color people are; it does not care what their religion is; it only cares whether they can produce something you want to buy.”Milton Friedman, 1912 to 2006
“An economist is an expert who will know tomorrow why the things he predicted yesterday didn’t happen today.”Laurence J. Peter
This book takes Friedman seriously enough to disagree with him carefully. He was brilliant, and he was often wrong, and those two facts are not in tension. Every chapter states his argument in its strongest form before taking it apart, and every chapter admits what he got right.
TWO LENSES,
ONE ECONOMY.
Economics has a bad habit. It is either explained to you as if you were five years old, in which case you learn nothing you can use, or it is explained as if you already have a doctorate, in which case you learn nothing at all. This book refuses to choose. Every one of the nine chapters is built twice, and every chapter ends with a steelman of the position it just spent several pages criticizing — a critique that can’t survive the other side’s best argument is not a critique, it’s a mood.
The Picture In Your Head
A story with taps and orchards and boats. No equations, no jargon. The analogies aren’t decoration — they’re the actual logical structure of the argument, drawn in a way a person can hold in their mind.
The Machinery Underneath
The same idea in its formal clothes: endogenous money, NAIRU, price elasticity, capital account volatility. Named scholars, real numbers, live disputes.
You’ll see a box like this every so often. It asks you a question instead of telling you an answer. There is no key in the back. Arguing about them at the dinner table is the entire point.
THE CAST.
Fifty years of argument, condensed into the people who made it. You’ll meet them again in context.
THE MAN WHO
REWIRED THE WORLD.
For roughly half a century, the world ran on one man’s intellectual architecture. That is not a figure of speech. If you have ever heard a politician say that government should be run like a business, that inflation is caused by printing money, that a company’s only duty is to its shareholders, or that competition will fix a failing school, you have heard Milton Friedman, whether or not the speaker knew his name.
Friedman did not invent free market economics. What he did was rarer and more consequential: he made it simple. He produced arguments that fit on a postcard and survived contact with a television audience. Each carried an entire policy program inside it, folded up small enough to travel.
He arrived at the perfect moment. The postwar consensus, built on Keynes, held that governments could and should manage demand to keep employment high. By the 1970s that consensus was in visible trouble — inflation and unemployment rose together, which the standard model said should not happen. Friedman had predicted exactly that, publicly, in advance.
Being right when your rivals are conspicuously wrong is the most powerful thing that can happen to an intellectual. It happened to Friedman on the largest possible stage.
What followed was not merely an academic shift. It was an installation. His ideas became the default settings of central banks, finance ministries, corporate boards, and business schools across dozens of countries.
Then the world ran the experiment. Not a laboratory experiment with controls and placebos, but the messier, slower, more expensive kind: fifty years of actual policy, in dozens of countries, generating data nobody could have simulated in advance. The results came back mixed, and in several cases they came back catastrophic. That is the subject of this book — not whether markets are good, which is too vague to answer, but what happened to nine specific, testable, enormously consequential claims when the world tried them.
The Installation, and the Audit.
Who Turns On
the Tap?
“Inflation is always and everywhere a monetary phenomenon.”Milton Friedman, 1963 · The sentence that reorganized central banking
Set the Pump and Walk Away
Friedman’s monetary theory has the beauty of a machine with one moving part. Prices rise, he argued, when the quantity of money grows faster than the quantity of things to buy. Too much money chasing too few goods. The central bank controls the quantity of money. Therefore the central bank controls inflation, and every inflation in history is ultimately its fault.
From this followed the policy that made him famous among central bankers: the k-percent rule. Stop trying to fine-tune the economy. Simply expand the money supply at a fixed, boring, announced rate every year, forever, and let the private economy arrange itself around that certainty.
The formal backbone is the equation of exchange, old, elegant, and true by construction:
Money supply times how fast money circulates equals the price level times the volume of transactions. It’s an identity — it cannot be false. Turning it into a policy tool requires one extra assumption: that velocity is stable. That assumption is where the trouble starts.
The City Water System
Picture a city’s water supply. One big pump station, pipes running to every house. Friedman’s picture of the economy looks exactly like this: the central bank is the pump station, money is the water. Pump harder, and pressure in everyone’s taps goes up — that’s inflation. So set the pump to a slow, steady speed and never touch it again.
Now look at how the system actually works. The pump doesn’t push water into your kitchen — you open your tap, and water is drawn toward you. In the real economy, the tap is a loan. When a family takes a mortgage or a business borrows to buy machines, money is created at that moment, by the private bank making the loan. The central bank doesn’t decide how much. It decides what the water costs, then supplies whatever quantity people have already decided to draw.
That reversal changes everything. If the central bank is the price tag, not the pump, you can’t fix inflation by setting a dial. You have to ask a harder question: why are so many people opening their taps at once?
Endogenous Money
The technical name for the reversal in the small lens is endogenous money. Money is created inside the economy by private credit, not delivered into it from outside by the state. In Horizontalists and Verticalists (1988), Basil Moore argued the standard picture had the causation backwards: banks don’t lend out deposits they’ve collected, they create deposits by lending, and reserves are supplied afterward by a central bank that has no real choice in the matter if it wants to hold its interest rate target.
A verticalist draws the money supply as a vertical line: a quantity the central bank picks, independent of price. A horizontalist draws it as a horizontal line at the policy interest rate: the bank picks the price, and quantity is whatever the economy takes at that price. These imply opposite theories of what inflation is.
Even granting Friedman his direction of causation, the policy needs velocity to behave — and velocity wanders violently. During crises, households hoard cash; financial innovation continually invents new near-monies. Paul Krugman and others note the awkward consequence: the aggregate you can measure is not the aggregate that matters, and the relationship you need to be stable is precisely the one innovation destabilizes.
In 2014 the Bank of England published a paper stating plainly that commercial banks create money by lending, and that the textbook money multiplier describes the process backwards — the central bank of a major economy formally endorsing the position Friedman rejected.
What Friedman Got Right
Friedman and Anna Schwartz’s Monetary History made a case that stands today: the Federal Reserve turned a bad recession into the Great Depression by allowing the money stock to collapse. Central banks in 2008 and 2020 acted the way they did partly because that lesson was learned.
He was also right that sustained, serious inflation has never occurred without monetary accommodation — there has never been a hyperinflation without a printing press. When American broad money grew by roughly a quarter in a single pandemic year and serious inflation followed, monetarists dismissed for decades were entitled to at least a raised eyebrow.
The warning survived; the mechanism did not. Money still matters, and no central banker ignores it. But the claims that the bank controls the quantity of money, that velocity is stable, and that a fixed rule outperforms judgment have all failed. The Fed tested its own rule and dropped it after three years.
If banks make new money every time they lend, who decides how much money a country has? The government? The banks? Or everyone who walks in and asks for a loan?
The Gardener
They Called
a Fence
“Governments never learn. Only people learn.”Milton Friedman · The premise behind a generation of privatization
The State Is a Referee, Not a Player
The second pillar is a theory of institutions. Government, in this view, is structurally incapable of creating value. It has no price signals to guide it, no competitive pressure to discipline it, no profit motive to focus it, and no mechanism for shutting down its own failures. What it does have is coercive power and other people’s money, which is a dangerous combination.
So the state should be kept to a short list: enforce contracts, protect property, keep the peace, provide the handful of genuine public goods markets cannot supply. Everything else belongs to the private sector, where invention and wealth creation actually happen.
The Orchard and the Fence
Think of the economy as a huge orchard. Friedman’s picture puts government at the edge of it, as the fence. A good fence keeps out thieves and deer. A fence that tries to garden is a disaster — it’s made of heavy machinery and will crush the soil. So: build the fence, then leave the trees alone.
Now walk into the orchard and ask a simple question. Who planted these trees? Not this year — twenty, thirty, forty years ago, when they were seeds with no fruit and no guarantee any of them would grow. Somebody paid for that. Somebody watered dirt for a decade with nothing to show for it.
Follow the money back and you keep arriving at the same gardener: the public. Taxpayers funded the strange, slow, unprofitable experiments that later turned into industries. Private companies arrive at harvest time — genuinely valuable work. But calling the gardener a fence is not a small mistake. It’s the difference between a bill and a gift.
The Entrepreneurial State
Mariana Mazzucato’s The Entrepreneurial State (2013) went looking for receipts. Take the object most often held up as proof of private genius — the smartphone — and trace each core technology back to whoever paid for the research when it was still a bad bet. The internet came out of a defense research agency. Satellite positioning was built by the military. Multi-touch screens trace back to publicly funded laboratories. The voice assistant emerged from a defense-funded AI program.
American business funds around three quarters of all R&D — a decisive victory for the private sector, until you break spending down by type. Business dominates development, turning a known thing into a sellable thing. Federal money dominates basic research, finding out whether the thing exists at all. The further you get from a sellable product, the more the public is paying.
DARPA / ARPANET
DARPA AI Program
US Dept of Defense
Military R&D lineage
Publicly funded labs
Defense procurement
Public energy research
What Friedman Got Right
Government failure is not a fantasy. Public agencies really do capture, really do calcify, and really do keep funding programs long after the evidence is in. Friedman’s attacks on occupational licensing as a cartel, on subsidies as a magnet for lobbying, and on regulators as eventual servants of the industries they regulate have aged extremely well.
There’s also a real methodological objection: the entrepreneurial-state literature tends to count the winners. For every DARPA there’s a portfolio of expensive public failures nobody writes books about. The honest conclusion is narrower: the state is a superb bearer of uncertainty and a poor operator of businesses.
The warnings about bureaucracy and capture survive intact. The central claim doesn’t. The state is not the fence around the orchard — it’s the entity that planted most of what’s growing in it.
If your city pays to build a road, and a delivery company uses it to make a billion dollars, who made the money? Does your answer change if the road took thirty years to build?
The Mall
With Armed
Guards
“Economic freedom is an essential requisite for political freedom.”Capitalism and Freedom, 1962 · The most consequential sentence he ever wrote
Free Markets, Then Free People
This is the moral heart of the whole project, and it deserves to be stated precisely, because it’s the claim most often misquoted in both directions. Friedman’s formal argument is that economic freedom is a necessary condition for political freedom, not a sufficient one. A dissident needs somewhere to work the state doesn’t control, a printer who’ll take their money, a landlord who doesn’t answer to a ministry.
But Friedman went further in practice, repeatedly: he predicted that opening an economy would set political liberalization in motion, that market societies tend to become free societies. Necessity is a modest claim. Prediction is a testable one.
The Mall With Armed Guards
Imagine an enormous shopping mall. Inside, you can buy anything, sell anything, open a shop, hire whoever you like. It is genuinely free in that sense. Around the mall stand guards with rifles. If you criticize the people who own the mall, you disappear.
Friedman’s bet was that the shopping would eventually dissolve the guards. Customers used to choosing between forty kinds of shoes start wanting to choose their leaders. Freedom leaks.
Sometimes it does. But look again at who hired the guards. Very often, it was the shopkeepers — guards keep workers from organizing, keep wages down, keep the doors open during a strike. Buying and selling did not have to be at war with the men holding rifles. On several important occasions, it was their best customer.
The Authoritarian Test Cases
Chile is the sharpest test because the treatment was administered in its purest form. After the 1973 coup, Chicago-trained economists — the Chicago Boys — restructured the economy under a military government that was simultaneously torturing and killing its opponents. Chile eventually became a stable democracy, and Chileans in 1990 were far better off than in 1973. Also: the model crashed catastrophically in 1982, output fell by roughly a seventh in a single year, and the government that had privatized the banks was obliged to seize them to prevent total collapse.
The larger problem is that the world kept producing counterexamples of increasing size. China has generated four decades of extraordinary market-driven growth while building the most technologically sophisticated apparatus of political control in human history. Singapore combines one of the freest economies on earth with tight limits on speech and opposition. Once you stop expecting commerce to produce liberty automatically, an uncomfortable possibility appears: capital sometimes prefers authoritarian arrangements.
What Friedman Got Right
The necessity half of his claim has never been refuted. There’s still no example of a politically free society running a fully state-controlled economy. Chile also complicates the case against him — it did become a democracy, and living standards did rise substantially. What neither story supports is the claim that markets do the work on their own schedule.
Necessity survives; the prediction fails. Fifty years of evidence show market economies coexisting comfortably with sophisticated authoritarian rule, sometimes because repression made the market work more smoothly.
If a country lets you start any business you want but not say what you think, is it free? Which of those two freedoms would you give up first, and what would you lose?
A Number
That Ate Ten
Million Jobs
“There is no permanent trade-off between inflation and unemployment. Buy jobs with inflation and you will end with both problems at once.”Friedman’s 1968 argument · A prediction that came true within a decade
Some Unemployment Is Natural
Before 1968, policy ran on the Phillips curve: an apparently reliable trade-off in which lower unemployment came with higher inflation. Governments treated it as a menu — choose your point on the curve, accept the price, get the jobs.
Friedman destroyed the menu with the single most successful prediction in modern macroeconomics. Workers, he argued, care about real wages, not the number on the check. Inflate to boost employment and people are fooled once — then they adjust expectations, demand higher wages, and unemployment returns to where it started, now with permanently higher inflation. Keep doing it and you get stagflation, which the old model said was impossible.
He named the resting point the natural rate of unemployment — later, NAIRU: the Non-Accelerating Inflation Rate of Unemployment. In the 1970s, inflation and unemployment rose together exactly as he said they would.
The Doctor and the Fever
A patient comes to a doctor with joint pain. The doctor says: a certain amount of this pain is natural, it’s simply how bodies are. If I give you medicine to take it away, you’ll develop a fever that could kill you. So we won’t treat it.
The patient believes him — he’s the doctor, he has a chart. She lives with the pain for thirty years. Then a different doctor gives her the medicine. The pain goes away. The fever never arrives.
The first doctor wasn’t a liar — he’d seen the fever happen, dramatically, in the 1970s. The problem is what he did with that memory: he turned one observation into a law of nature, put the word natural in front of it, and used it to justify not treating a patient for a generation.
An Unobservable Number With a Body Count
NAIRU cannot be observed. It’s a parameter estimated from a model, revised repeatedly, always after unemployment fell below it without the predicted inflation arriving. Through the 1990s, respectable US estimates sat above six percent. Unemployment fell to four in 2000; inflation stayed quiet; the estimate was lowered. In 2019, unemployment reached the lowest level in half a century; inflation stayed below target; the estimate was lowered again.
The deeper problem is that the empirical relationship itself weakened. Both Janet Yellen and Jerome Powell, as Fed chairs, publicly acknowledged the Phillips curve had flattened. Post-Keynesian economists argue most serious inflations are driven by supply shocks and firm pricing power, not too many people having jobs — under Friedman’s framework, unemployment became the designated shock absorber for the whole system.
What Friedman Got Right
More than almost anywhere else in this book. The 1968 prediction was correct, made in advance, against the consensus of the entire profession. Expectations belong at the center of any inflation model, and every modern central bank watches them obsessively because of him. His successors also got partial vindication in 2021–22, when tight labor markets coincided with the fastest price rises in forty years.
The theory largely survives; the number does not. Expectations matter and there’s no permanent menu — the real insight. But NAIRU as an operational target has been revised downward so many times its main function looks less like measurement and more like permission.
Would you accept slightly higher prices for meaningfully lower unemployment? Notice whether your answer changes depending on whether you own assets or earn wages.
The Tree
Cut Down
for Acorns
“The social responsibility of business is to increase its profits.”New York Times Magazine, 1970 · The essay that became a constitution
Executives Are Employees of the Owners
In 1970 Friedman published a magazine essay that has probably shaped daily corporate behavior more than any academic paper of the century. A corporate executive is an employee of the shareholders. The shareholders own the company. If an executive spends company money on social causes or wages above what the market demands, that executive is spending someone else’s money on their own preferences.
Worse, argued Friedman, they’re doing it without a mandate — appointing themselves a legislator, a tax collector, and a judge of the public good, accountable to nobody. Better to maximize profits inside the rules of the game and let democracy set the rules.
The Oak Tree
An old oak tree stands in a field, giving shade, holding soil together, dropping acorns every autumn. A person arrives with a deed and asks: what is this tree for? Answer: acorns. So the caretakers stop watering the roots — watering costs money and roots produce no acorns. They stop treating the soil. They cut the lower branches to speed the harvest. Every decision raises this year’s acorn count, and every one is defensible on its own.
For three or four seasons, it’s the most productive tree anyone has seen. Then the roots rot, the soil washes away in the first hard rain, and the tree comes down. The acorn count for the following year is zero, and the field is now a place where things do not grow.
The Legal Claim Is False
Shareholders do not own the corporation. In The Shareholder Value Myth, corporate law scholar Lynn Stout demonstrated that a corporation is an independent legal person — it owns itself and its assets. Shareholders own shares, which are contracts with a defined, limited set of rights. American courts generally apply the business judgment rule, which gives directors broad latitude. Shareholder primacy is not a legal requirement — it’s a management ideology that spread because it was simple and aligned with stock-based executive pay.
Once share price became the scoreboard, buying back your own stock — mechanically raising earnings per share — became a legitimate use of corporate cash. It requires no new product, no new hire, no new factory. The corporation began as a chartered body created for a public purpose: a bridge, a port, a university. The idea that it’s a private wealth-extraction device answerable to a single constituency is roughly fifty years old.
What Friedman Got Right
His strongest argument is democratic, not economic, and it’s rarely answered well. If a CEO decides which social causes deserve corporate money, who elected them? Stakeholder capitalism can become a license for unaccountable managers to pursue whatever the current fashion rewards. His reply to vague corporate virtue was: pass a law, and enforce it on everyone equally.
The accountability argument survives; the legal premise was never true. A doctrine sold as legal fact was a management fashion all along, coinciding with stagnant wages, record buybacks, and collapsed corporate time horizons.
If a company earns a hundred dollars, who should get it? Shareholders, workers, the town that built the roads, or the customers? Try splitting it and defend your split.
The Coupon
That Made
It Worse
“Give parents the money and let them choose. Competition will force bad schools to improve or close.”The voucher argument in one sentence · Proposed by Friedman in 1955, scaled up after 2010
Fund Students, Not Schools
Friedman proposed vouchers before he was famous, and defended them for fifty years. Public education is a government monopoly. Families are assigned to schools by home address, which means the school has a captive customer and no reason to improve. So: take the money the state spends per pupil, give it to the family as a voucher, and let them spend it anywhere. Good schools attract families and expand. Bad schools lose families and close.
The Taxi Coupons
A city has a bus system, old and slow. Someone proposes: stop funding the buses, give every resident a taxi coupon instead. Watch the buses improve through competition.
Here’s what happens when you try it. Taxi companies notice every passenger holds a coupon of fixed value, so they raise fares to absorb it. Wealthy people, already taking taxis daily, now ride for free — the coupon just paid their existing bill. People in far-out neighborhoods discover taxis won’t drive to them; no coupon forces a driver to take a fare he doesn’t want.
Meanwhile new taxi companies appear overnight, because coupons exist. Some are fine. Some have no insurance and no brakes. And the bus, still carrying everyone the taxis refused, runs on a fraction of its old budget.
The Most Alarming Result in Education Research
After 2010, Louisiana, Indiana and Ohio scaled voucher programs up, and oversubscribed programs allocated places by lottery — a randomized controlled trial policy handed researchers for free. Compare students who won a voucher to those who applied and lost: the difference is causal. Susan Dynarski and colleagues ran the comparison, and the results were so negative several researchers said publicly they’d expected the opposite.
Three mechanisms explain most of the failure. First, the money mostly went to families who’d already chosen — roughly seven in ten recipients were already enrolled in private school. Second, the best schools declined to participate, since accepting vouchers means accepting state testing and reporting rules. Third, the market that appeared attracted entities that could open a building fast, not necessarily excellence — Joshua Cowen documents financially distressed private schools using vouchers as a lifeline.
What the Other Side Can Still Say
These results indict specific programs rather than choice itself. Some early city programs found gains in graduation and college enrollment even where test scores didn’t move. Charter schools — also a choice mechanism, but under public accountability rules — have produced genuinely large gains for disadvantaged students in several major cities.
The clearest empirical failure in the book, because it was the cleanest test. Lotteries removed the usual excuses about selection, and the results came back among the most negative in modern education research.
How would you actually know if a school was good before you sent your child there for a year? Write down three ways. Now ask which of them a family with no car and no free time could use.
Who Actually
Cashes the
Check?
“Abolish the welfare bureaucracy. Send poor households cash through the tax system. Then abolish the minimum wage, which only prices the unskilled out of work.”The negative income tax proposal · The package matters more than either half
Cash Instead of Programs
Friedman hated the welfare state for a reason that had nothing to do with cruelty. He thought it insulting and inefficient: dozens of separate programs, each with its own bureaucracy, each deciding what poor people were permitted to want. His alternative was the negative income tax — set a threshold; earn above it, you pay tax; earn below it, the tax system pays you, automatically, in cash, with no caseworker.
It became law, in modified form, as the Earned Income Tax Credit. But Friedman attached a condition: he wanted it to replace the minimum wage, which he regarded as pricing the least-skilled workers out of employment entirely.
The Bakery and the Bread
A worker needs three loaves of bread a day to live. The bakery pays her two. Friedman’s solution: the government gives her the third loaf. No bureaucracy, no shame, and nobody had to tell the bakery what to pay.
Run it forward a year. The bakery owner notices the government always provides the third loaf. So he offers one loaf instead of two — the worker takes the job, because one loaf plus a guaranteed government loaf still beats nothing.
The worker is no better off. The taxpayer is paying double. The bakery owner now pays one loaf and keeps the difference. The subsidy meant for the worker landed in the owner’s pocket — and everybody behaved perfectly rationally on the way there. Unless there’s a law saying the bakery may not pay less than two loaves. Then the floor holds, and the money goes where it was aimed.
Incidence, or Who Really Gets the Money
Labor economists call this the incidence problem. An income-contingent benefit shifts the labor supply curve outward: workers become willing to accept lower wages because the state fills part of the gap. In a competitive labor market with no floor, that shift lowers the market wage — some of the subsidy reaches the worker, some is captured by the employer.
History ran this experiment before economics existed as a discipline. In 1795, English magistrates at Speenhamland supplemented agricultural wages with a scale tied to bread prices. Employers cut wages, knowing the parish would cover the difference. Public money flowed into private profit; the system was condemned and abolished in the 1830s. Karl Polanyi later made it the centerpiece of his account of how markets and societies collide. The lesson isn’t that cash transfers fail — it’s that a wage subsidy and a wage floor are complements, not substitutes.
What Friedman Got Right
Almost everything except the pairing — this may be his most vindicated idea. Give poor people money rather than supervising their consumption has won across the political spectrum. The Earned Income Tax Credit lifts millions out of poverty. His minimum-wage worry was also serious rather than ideological, and the empirical debate is genuinely unfinished. He had the right instrument and the wrong package.
The highest score in this book. Direct cash is efficient, dignified and effective, and Friedman was early and right. The failure is the condition he attached — without a wage floor, a wage subsidy becomes a transfer from taxpayers to low-wage employers, the Speenhamland result, documented two centuries before he proposed it.
If the government tops up low wages, why would an employer ever raise them? And if your answer is competition for workers, what happens where there’s only one big employer in town?
The Draft
Nobody Calls
a Draft
“Conscription is a tax collected in bodies. If the army needs soldiers, let it pay a wage and hire them, like any other employer.”The case Friedman made to the Gates Commission · The draft ended in 1973
An Army Is a Labor Market
Of all the policies in this book, this is the one Friedman was proudest of. Conscription forces a young person to accept a job at below-market pay under threat of prison — a tax levied entirely on the people least able to avoid it, invisible in the budget because the cost is paid in unpriced human time.
His argument to the commission that ended the American draft: set the wage high enough and people will serve willingly. He was right about the mechanics — the all-volunteer force is a formidable professional military. The question this chapter asks is what else was purchased along with it.
The Forest Fire
A fire is coming toward a family’s house. The old way: everyone grabs a bucket — grandmother, cousins, everyone, because it’s everyone’s house. The new way is more efficient: the family pools money and pays the poorest cousin to fight the fire alone. He accepts freely, because he needs the money. The rest stay inside and watch television.
Something has been solved here and something has been broken. Nobody was compelled, which genuinely matters. But now the cousin risks his life, and the others risk nothing and decide everything. Because the fire no longer costs the deciders anything, they’ll be much less careful about how many fires they choose to fight.
The Market Draft
Michael Sandel calls the result a market draft. A choice made freely between good options is one thing; a choice between military service and a set of civilian options with no affordable path to higher education is a different thing wearing the same word. Douglas Kriner and Francis Shen documented the consequence in The Casualty Gap: American war deaths have fallen increasingly on lower-income communities.
Under conscription, a decision to go to war reached every household with a young person in it, generating enormous political friction. Remove it, and war becomes something that happens to other people’s families. The most direct measure of that insulation is what’s happened to the legislature that declares war.
What Friedman Got Right
Conscription is coercion, and no amount of civic poetry changes what it does to the person conscripted. The draft that romantics remember was also deeply unequal in practice — deferments and exemptions meant the wealthy were largely spared. Millions of young people have not been forced into uniform because Friedman won this argument.
The liberty argument holds and shouldn’t be dismissed. What the model missed is that a market solution to a civic problem changes the politics of the thing it solves.
Is there anything a country should never be allowed to buy, even if the seller agrees to the price? Make a list of three things. Then defend the hardest one.
Cutting
the Anchor
“Let currencies float. Trade imbalances will correct themselves smoothly, without crises, committees or capital controls.”Friedman’s case against Bretton Woods · Made from 1953 and won in 1971
The Market Can Price Money Too
After WWII, currencies were pegged to the dollar and the dollar to gold. Governments defended pegs with reserves and capital controls. Friedman argued this was absurd — why should currency price be the one price in the economy set by a committee? Let currencies float, and adjustment becomes automatic: a country importing more than it exports sees its currency drift down, closing the imbalance on its own.
The Boat and the Ocean
A boat sits at anchor. The anchor is heavy and drags, and everyone agrees it’s a nuisance. Friedman’s proposal: cut it loose. The ocean has currents, and a free boat will drift naturally toward where it needs to go. The currents here are trade — they really do push in the direction Friedman described.
What the plan didn’t account for is the weather. Above the currents there’s wind — speculative finance — and it’s not a breeze. It’s a hurricane system that can change direction in an afternoon because of a rumor. A boat with no anchor in calm water drifts gently. A boat with no anchor in a storm is thrown wherever the storm wants.
Money Moves Faster Than Goods
The fatal miscalculation was one of scale. In Friedman’s model, exchange rates are driven by trade flows, which move slowly. In the actual world, they’re driven overwhelmingly by capital flows — interest differentials, risk appetite, the collective mood of fund managers — which move at the speed of a keystroke.
A currency’s price is set by opinion about the future, not the trade balance. Opinion is subject to herding, and herding produces sudden stops: capital arrives for years, then leaves in weeks. The Asian financial crisis of 1997 is the canonical case — Thailand, Indonesia and South Korea had been growth models; within months, capital reversed, currencies collapsed, and unemployment and poverty rose across a region that had been the standing example of successful development.
What Friedman Got Right
Every country devastated in 1997 was defending a peg. A Friedmanite will say the disaster was caused by the fixed rate, not by floating. Bretton Woods didn’t fall because Friedman argued against it — it fell because it was unsustainable, and floating rates have survived fifty years, several oil shocks, a global financial crisis and a pandemic without the systemic breakdown critics predicted.
The regime survived; the promise didn’t. Currencies float and the world hasn’t ended — but the gentle self-correcting mechanism never materialized, because the market that sets exchange rates is a hundred times larger than the trade it was supposed to be balancing.
Should money be allowed to leave a country as fast as it arrived? Who benefits from the speed, and who is standing underneath when it leaves?
HOW MUCH OF EACH
IDEA IS STILL STANDING
Ranked by how much survives contact with the evidence. These percentages are editorial judgments, not measurements — given as numbers so you can disagree with them precisely.
Cash beats programs. He was early and right.
0%
Real liberty bought. A civic bond sold to pay for it.
0%
The theory survived. The number never should have.
0%
The regime survived. The gentle adjustment never arrived.
0%
The warning stands. The control lever was never attached.
0%
A management fashion sold to the world as a legal duty.
0%
Necessary, plausibly. Sufficient, demonstrably not.
0%
The gardener was misfiled as the fence for fifty years.
0%
The cleanest test in the book. The worst result in it.
0%
Every idea scoring above 40 was about individual freedom from coercion.
Every idea below 30 replaced a public institution with a market.
THE OTHER SIDE
OF THE LEDGER.
A book that only prosecutes is not analysis, it’s advocacy with footnotes. So here, in one place, is the case for the defense, stated as strongly as it can be stated.
In 1968 he told the entire economics profession its central empirical relationship would break down, and gave the mechanism. Within five years it broke down, by the mechanism he described. Very few economists have ever made a prediction that consequential.
Inflation is a regressive tax that falls hardest on people holding cash and earning fixed wages. Friedman insisted on this when the fashionable view treated a few points of inflation as a small price for full employment.
Agencies are captured. Programs outlive their purpose. Friedman said this loudly for fifty years, and the parts of it that were controversial then are consensus now.
The negative income tax is one of the most humane policy ideas of the twentieth century. He argued for treating poor people as adults who can be trusted with their own money.
Whatever the civic costs, millions of young people were not forced into uniform against their will because he won that argument — a substantial and durable gain in human freedom.
The necessity half of his freedom argument stands unrefuted. Every experiment in placing all economic power in the state has produced political power that could not be checked.
The strongest critique of Friedman is not that he was foolish. It is that he was right about several important things and then generalized from them with a confidence the evidence never earned.
WHAT REPLACES IT.
The purpose of an audit is not demolition. Friedman’s architecture was tested more thoroughly than almost any body of social thought in history, because it was actually installed, in dozens of countries, for decades. Three things fifty years of evidence taught us.
One. Mechanical Models Break on Human Behavior
The theories that failed hardest were the ones that needed a human constant: stable velocity, a knowable natural rate, parents who can evaluate school quality in advance, currency markets driven by trade fundamentals. In every case the constant turned out to be a variable, and the variable turned out to be a mood.
Two. The State Is Not a Fence, and Not a Saint Either
The evidence supports a specific and testable division of labor. The state is the only actor with the time horizon to fund things that might not work for twenty years, and generally a poor operator of businesses. Design accordingly, and stop arguing about whether government should act.
Three. Some Things Are Corrupted by Being Priced
Defense, education and the duties of citizenship were treated as commodities that happened to be publicly provided. Turning these into markets didn’t just produce worse outcomes — it changed what the thing was.
None of this is an argument for abolishing markets, which would be both impossible and stupid. Markets are the best instrument ever discovered for aggregating dispersed information into prices. That is an extraordinary achievement and no serious critic disputes it.
The claim is narrower and older than the last fifty years of argument. Markets are excellent servants and catastrophic masters. Moving beyond the Chicago School doesn’t mean finding a better model to defer to for the next fifty years — it means recovering the older discipline Friedman’s generation renamed: political economy, in which the rules of the market are understood to be written by societies, revisable by societies.
THE GLOSSARY THAT
DOES NOT ASSUME.
TEN QUESTIONS FOR
THE DINNER TABLE.
There are no answers in the back. Tap each one to open it.