The Great Depression shattered the classical faith in automatic recovery. Keynes set out from pure logic to demonstrate that an economy can become chronically stuck in demand deficiency — unable to climb out on its own. The tool for filling the gap — the state as buyer of last resort — was validated at full scale during the Second World War. But two questions remain for those who follow: what price must eventually be paid for distorted price signals, and on whom does the incremental demand ultimately fall?
Three Diagrams That Run Through It
Let's begin with three diagrams. They will run through this essay and every one that follows. The horizontal axis is time; the vertical axis is an economy's total output.
Diagram one: a straight line, with a fixed slope, rising steadily upward. It represents potential output — labor growing, technology advancing, capital accumulating, division of labor deepening. This is the only diagram classical economics (Smith, Ricardo, Say) ever looked at. They did not worry about who would buy the goods, because supply creates its own demand — the boundary of productive capacity is the economy's actual path.
Diagram two: a wavy line oscillating around a horizontal baseline. Peaks: factories running overtime, wages rising. Troughs: workers lining up for relief. This is the business cycle, the investment cycle, financial panic. The long-run trend has been stripped away; only expansion and contraction alternate.
Diagram three: the sum of the first two. A rising straight line, tugged back and forth by the wave wrapped around it. This is the real world — an economy moving on both dimensions at once. Most of the time the actual GDP wobbles only slightly around the trend line. Occasionally the deviation becomes enormous. From 1929 to 1933, U.S. actual GDP fell more than 30% below the potential-output line. The vertical distance between the two lines is the output gap.
These three diagrams are the core of the Keynesian revolution. Classical economics had only the first diagram; it assumed fluctuations were temporary and self-correcting, requiring no independent theory. Keynes proved that the second diagram has its own logic — it can deviate from the first severely and for a long time. No one borrows even at zero interest; no one hires even at lower wages. For the first time in two hundred years, fluctuation was elevated from noise to a phenomenon requiring its own theory.
1929: Black Tuesday, Then a Chain Reaction
October 29, 1929, the New York Stock Exchange. In a single day the Dow fell 12%; over the next three years it dropped from 381 to 41 — nearly 90% of its value wiped out. Industrial output fell 30% from 1929 to 1933. Unemployment went from 3% to 25%. Five thousand banks failed; nine million savings accounts were wiped out.
More vivid than the numbers are the scenes: farmers pouring milk into rivers and slaughtering and burying their pigs; in the same month, children in cities scavenging through garbage for food. Too much produced, too little consumed — the gap between supply and demand so wide that in the same country, in the same month, food was destroyed and people starved simultaneously.
The true cause was buried in the prosperous twenties. In 1920s America, the assembly line had driven the price of a Model T below $300. Radios, refrigerators — things only millionaires could afford a generation earlier — a Philadelphia auto worker could now bring home on installment credit. Factories ran at full throttle. But from 1923 to 1929, manufacturing output per hour rose 32% while workers' real wages rose only 8%. The crack was papered over by credit. Buying cars, radios, washing machines on installment — the worker had no cash in hand, but the banks were willing to lend. Marx's diagnosis had long since said it: the gap between productive capacity and consuming power is temporarily patched by credit — and then the patch bursts before the gap does.
The transmission chain continued: stocks fall → banks call in loans → borrowers forced to sell everything → falling prices make collateral worth less → banks call in more loans → depositors panic → five thousand banks fail → credit freezes → firms lay off workers → the unemployed cannot consume → stores get no orders → stores lay off more workers. The Federal Reserve had been created in 1913, in theory to prevent exactly this. But from 1929 to 1933, it did virtually nothing — and actually tightened, allowing the money supply to contract by a third. The rationale sounded responsible: "Defend the gold standard, let bad banks fail, clear out the rot — then the economy can heal." But by 1933, one in four Americans had no job.
Five Pillars, Dismantled One by One
Hoover's problem was not malice. He genuinely believed markets would heal themselves. Behind him stood the entire standard toolkit of contemporary economics — Say's Law, classical interest-rate theory, classical labor-market theory. Answers that seemed airtight in the classroom were being torn to shreds by the Depression. At the same time, a Cambridge man was setting out from pure logic and dismantling them, one pillar at a time.
Keynes's weapon was essentially one: the fallacy of composition — what holds for one person does not necessarily hold for everyone.
The first pillar: Say's Law — "Supply creates its own demand"
Say argued: you produce $100 worth of cloth and pay out $100 in wages, profit, and rent; those who receive the money will spend it. True for one person. But for everyone? The people receiving that money are doing two things at once: spending and saving. Someone who saves may be worried about tomorrow — spend less, save more — but the bank will not automatically find an entrepreneur confident enough in the future to borrow those savings. Saving and investing are done by different people with different motives. When everyone saves more and spends less at the same time, aggregate demand shifts left and output falls. Supply does not automatically create its own demand.
The second pillar: Thrift is a virtue
Classical economics said: if everyone spends less and saves more, bank savings increase, the supply of loanable funds shifts right, interest rates fall, borrowing becomes cheaper, and investment rises. Consumption is transformed into investment. Keynes asked a question: who decides to save? Households. Who decides to invest? Entrepreneurs. Two different people, two different motives. Classical economics collapsed them into one sentence — saving automatically equals investment. In the normal interest-rate zone, the chain holds. But when rates are already very low, the liquidity trap sets in — entrepreneurs do not look at interest rates; they look at whether anyone will buy their product tomorrow. If they cannot be sure, they will not borrow even at zero.
The third pillar: What is interest?
Classical economics said interest is the reward for saving — you defer consumption, deposit money in the bank, and the bank pays you interest. Keynes said no. Interest is the reward for giving up liquidity — you exchange cash you could spend at any moment for a bond, and receive interest as compensation. This is not the same thing as saving.
The fourth pillar: Lower wages to restore employment
Classical economics said unemployment means wages are too high; cut them and employers will hire more. Keynes raised two objections. First, wages are sticky — workers do not accept nominal wage cuts; they have contracts, they have a sense of fairness. Second, even if wages were cut — if everyone's wages fall, everyone's consumption falls too, aggregate demand deteriorates further, and employment does not rise but falls.
The fifth pillar: Markets heal themselves
Classical economics assumed the first four pillars held, so the market could self-correct. Keynes showed that each pillar holds only locally. An economy can remain trapped on a flat segment — factories operate when orders come in and sit idle when they do not. It does not automatically slide back to full employment.
Two Trailers
Keynes did not write The General Theory only after the Depression. He had already demonstrated the tools twice before.
Sitting in a hotel room near the conference hall, he wrote The Economic Consequences of the Peace. Not a word wasted on "poor Germany." Every bullet aimed at the same target: the math is wrong — you are demanding an impossible outcome.
Trailer one: The Versailles reparations
At the Paris conference, French premier Clemenceau wanted to crush Germany, British prime minister Lloyd George vacillated in the middle, and U.S. president Wilson pushed his Fourteen Points. Keynes did the arithmetic: Germany's post-war export capacity had been cut by at least a quarter; to generate a trade surplus under the weight of reparations, it needed exports vastly exceeding imports — but the Allies simultaneously demanded reparations and erected tariff walls to keep German goods out. The only path was for Germany to drive down wages and living standards to squeeze out exports — but any German government pursuing that would not survive the next election.
He wrote The Economic Consequences of the Peace in that hotel room. Four years later, the German mark went from 60 to the dollar in 1921 to 4.2 trillion to the dollar in November 1923. His arithmetic was validated at full scale.
Trailer two: Britain's return to the gold standard
In 1925, Britain pegged the pound back to its pre-war rate of $4.86. Chancellor of the Exchequer Winston Churchill saw it as a mark of "responsible economic policy." Keynes wrote a pamphlet titled The Economic Consequences of Mr. Churchill, naming names: you have set the pound too high; British exports will be 10% more expensive everywhere. Exports fall, factories lay off workers, wages drop, consumption drops. The very stability of the gold-standard exchange rate was the mechanism locking demand down.
No one listened. In 1926, Britain was paralyzed by a general strike — coal miners, railway workers, dockworkers; the army was deployed to the Liverpool docks. In September 1931, Britain was forced off the gold standard. Keynes had written the answer in 1925. The validation took six years.
Four Responses
The Depression produced four different answers to the same question.
Hoover — Waiting for self-healing
"Prosperity is just around the corner." He said it in 1930, in 1931, and in 1932 — the year unemployment hit its all-time high — he was still saying it. The problem: liquidating one bank is market discipline; liquidating all banks simultaneously is a credit freeze. Cutting wages at one factory is cost optimization; cutting wages at every factory simultaneously is an aggregate-demand collapse.
Hitler — Invisible deficits
In Germany, with six million unemployed, the Weimar Republic was crushed between austerity and violence. Finance Minister Hjalmar Schacht designed the Mefo bill — a nominally private shell company issued promissory notes to arms manufacturers, banks discounted them, the Reichsbank rediscounted them. The government's books showed a deficit of zero, but currency was pouring into production lines through the promissory-note channel. By 1938, the outstanding balance was 12 billion Reichsmarks — roughly 12% of German GDP. Military Keynesianism, with tanks instead of dams as the final product — validated in the darkest possible register.
Roosevelt — Hadn't read Keynes, but walked the same path
"The only thing we have to fear is fear itself." The man who said it was paralyzed below the waist and delivered the speech standing with the help of steel braces.
The New Deal was an aggregation of bills — 12,000 pages, more than 400 agencies, no blueprint. But its effect landed on the same logic. The Works Progress Administration (WPA) employed over 8.5 million people, building roads, bridges, parks. The Tennessee Valley Authority (TVA) brought electricity to rural areas across seven states. Before the TVA, the U.S. rural electrification rate was below 10%.
But the New Deal also did foolish things. The National Industrial Recovery Act (NIRA) was effectively the legalization of cartels — minimum prices, output limits, frozen wages. The Supreme Court struck it down unanimously in 1935. The Agricultural Adjustment Act (AAA) paid farmers to slaughter six million pigs and plow under ten million acres of cotton. In the same month, breadlines stretched around city blocks. Keynes, had he read the AAA, would have said: you are treating a demand problem as if it were a supply problem.
The prescription among the four
Hoover waited for self-healing. It never came; the demand hole only deepened. Hitler validated the multiplier through war procurement — at the cost of an irreversible path. Roosevelt groped in the right direction but understood only half of it; he did the right things and also the wrong ones. Keynes's logic went further than the New Deal. He did not touch the ownership structure, did not alter distribution — from a purely logical starting point he derived: the state can stand in front of the output gap and fill it with its own purchasing power. One dollar of filling, amplified through the multiplier, becomes several dollars of incremental demand.
1937: A Reverse Validation
This is the single most important data point in New Deal history, and the most direct real-world test of The General Theory after its publication. By 1936, the economy had climbed back to 95% of its 1929 level; unemployment had fallen from 25% to 14%. Roosevelt worried about the deficit — the 1936 deficit was 5.5% of GDP — and decided to pull back. In 1937, he cut public works spending, tightened the budget, and the Social Security tax began to be deducted from wages. The Federal Reserve raised reserve requirements in the same year.
The result: from mid-1937 to mid-1938, GDP fell about 10%, manufacturing output dropped nearly 40%, and unemployment rebounded from 14% to 19%. After eight years of Depression, fiscal austerity pushed the economy back into the pit.
Keynes wrote Roosevelt a letter in February 1937: "You are pulling back at the worst possible moment. The patient has just managed to sit up, and you are pulling the blanket off him." Roosevelt did not change course immediately. Nearly a year passed before he opened the spending valve again — Congress passed $3.3 billion in public works and relief appropriations, about 3% of GDP. The economy recovered.
1937–1938 proved the reverse logic of the multiplier. Every dollar the government cut pushed total output down through the multiplier's negative side — by several times the amount of the cut.
World War II: The Ultimate Answer
December 7, 1941 — Pearl Harbor. Over the next four years, the U.S. government ran an annual fiscal deficit exceeding 20% of GDP. A number that today would send any country's credit rating to junk.
But here is what happened to the economy: factories ran at full capacity, unemployment fell from 25% to 1%, wages rose, consumption rose. The last unemployed person of the Depression found work in a munitions factory. Market self-healing has no standing in this history. It was the state, with the largest fiscal deficit in human history, forcibly filling the demand gap. Every theoretical prediction of Keynesianism was validated at full scale — in the most extreme and most unsustainable way possible.
Two Questions Left Unanswered
Keynes opened the door to demand management. But after he left, two question marks remained in the doorway.
Hayek's question: When you use public spending to fill the gap, you distort price signals in the process. What price will eventually have to be paid to restore them? The transactions that were never made because public spending covered the gap will never tell you what the price would have been if you hadn't filled it. In private correspondence, Keynes called Hayek's book "one of the most confused books I have ever read." But after Hayek published The Road to Serfdom in 1944, Keynes read it and wrote him a letter: "I agree with you morally and philosophically on almost everything, but your economic analysis is not persuasive enough. Planning itself is just a tool; the real danger is that the planner believes he can know everything." Hayek treasured that letter for most of his life.
Keynes's own question: Of the total demand that the state stimulates, how much of the incremental demand falls on domestic products and how much becomes imports? Keynes solved "who buys the goods" but did not solve "whose goods get bought."
The depth, breadth, and shape of the output gap — each subsequent essay will draw different countries' versions.