The puzzle classical economics could not stomach
Picture the Great Depression of the 1930s: factories standing idle, machines that worked perfectly gathering dust, and millions of able workers desperate for jobs that did not exist. Nothing was *physically* wrong. The land, the labour, and the capital were all still there. Yet the loop you met in the circular-flow guide had simply slowed to a crawl. The economists of the day, schooled in the classical tradition, found this almost impossible to explain. Their models said that markets clear: if too few goods sell, prices and wages fall until everything sells again, and any worker willing to accept a lower wage finds work. By that logic, mass unemployment lasting a decade should have been impossible. The world refused to read the textbook.
Into this gap stepped John Maynard Keynes, whose 1936 book launched what we now call Keynesian economics. His central heresy was disarmingly simple: an economy can get stuck at less than full employment and *stay there*, not because anything broke, but because total spending — aggregate demand — has fallen short of what the economy is able to produce. Demand, not capacity, was the binding constraint. The factories were idle not because they could not run, but because nobody was buying enough to make running them worthwhile. You already glimpsed the seed of this idea: the circular-flow guide warned that saving more might not smoothly call forth more investment, but could instead just shrink income. Keynes turned that warning into a whole theory of the slump.
Your spending is somebody else's paycheck
If demand drives the show, then anything that pushes spending up gets *amplified* — and this is the engine of the whole theory: the multiplier. The chain runs straight out of the circular flow. Suppose the government hires builders for a new bridge and pays them $1,000. That is not the end of the story; it is the first link. The builders take their $1,000 home and spend most of it — say they spend 80 cents of every dollar and save the rest. So $800 lands as income in the tills of grocers, landlords, and cafés. Those people, in turn, spend 80% of *their* new income — another $640 — which becomes someone else's income, who spends $512 of it, and on and on down the line.
The fraction people re-spend out of each extra dollar has a name: the marginal propensity to consume, or MPC. In our example it is 0.8. Add up the whole shrinking chain — 1,000 + 800 + 640 + 512 + ... — and it does not run away to infinity; it converges to a tidy total. The multiplier is simply 1 ÷ (1 − MPC). With an MPC of 0.8, that is 1 ÷ 0.2 = 5, so the original $1,000 of spending ultimately generates $5,000 of total income across the economy. The single seed dollar bloomed because it kept being passed along, each hand keeping a little and passing the rest on.
Round Spending added Running total 1 1,000 1,000 (MPC = 0.8) 2 800 1,800 3 640 2,440 4 512 2,952 5 410 3,362 ... ... ... total 5,000 = 1,000 x [1 / (1 - 0.8)] Multiplier = 1 / (1 - MPC) = 1 / 0.2 = 5
Run the chain backwards: the paradox of thrift
Now flip the multiplier into reverse, and you arrive at Keynes's most unsettling idea: the paradox of thrift. For any single household, saving more is plainly wise — it builds a cushion against hard times. But what is prudent for one can be ruinous for all at once. Imagine a recession scares everyone into cutting spending to save more. Every dollar a household stops spending is a dollar of income some shopkeeper, supplier, or worker no longer receives. That shopkeeper, now poorer, cuts *their* spending too. The multiplier runs backwards, and aggregate income spirals downward.
Here is the twist that gives the paradox its name. As incomes fall, people cannot actually save much more — you can only save out of income you have. So the grand collective effort to save *more* can leave everyone with a smaller economy and the same total saving, or even *less*. The intention to save more, acted on all at once, defeats itself. This is a fallacy of composition: assuming that what is true for one part must be true for the whole. It is the same flaw as standing up at a concert to see better — sensible for you alone, useless once everyone does it.
Animal spirits and the engine of self-fulfilling slumps
Why would demand collapse in the first place? Keynes pointed to the deeply human, hard-to-predict mood of investors and consumers, which he named animal spirits — the gut confidence (or dread) that drives people to build a factory or postpone a purchase. Investment especially hangs on expectations of an uncertain future, and those expectations can swing on rumour and herd feeling. When confidence cracks, firms freeze hiring and shelve projects, households delay big buys, and demand falls — which makes the slump real, which justifies the fear that started it. Pessimism becomes self-fulfilling. This is what makes a recession more than a tidy market correction: it can be a coordination failure, with everyone waiting for everyone else to spend first.
If a slump is a coordination failure, the cure suggests itself. When private spending freezes, *someone* has to spend to break the spiral — and the one player who can choose to spend even when terrified is the government. By stepping in with its own spending, the state injects fresh demand, the multiplier carries it through the chain, and idle resources are coaxed back to work. This is the intellectual root of stabilization policy: the idea that managing aggregate demand can smooth the business cycle rather than waiting, possibly for years, for prices and wages to grind the economy back to full employment on their own. Fiscal stimulus, the design of automatic stabilizers, the very reflex to "do something" in a recession — all trace back to this chapter of Keynes.
How big is the multiplier, really? The honest fight
Now the honesty this field demands. The clean multiplier of 5 in our worked example is a *ceiling*, not a forecast — it ignores every real-world leak. In practice, people also save more cautiously, pay taxes, and spend on imports made elsewhere, all of which drain dollars out of the domestic chain and shrink the multiplier. Worse, government borrowing to fund stimulus may push up interest rates and discourage private investment — an effect called crowding out that can claw back part of the boost. And if the economy is already near full employment, extra demand mostly just raises prices rather than output. The size of the multiplier is therefore one of the most *contested* numbers in all of macroeconomics.
Careful empirical studies put the real-world spending multiplier somewhere between roughly 0.5 and 2.0, depending on circumstances — and that range is itself fiercely argued. A multiplier below 1 means a dollar of government spending raises total output by *less* than a dollar; above 1 means it more than pays for itself in output. The evidence suggests the multiplier is largest exactly when Keynes's story applies most cleanly — a deep slump with idle resources and interest rates pinned near zero, where crowding out is weak — and smallest in a booming economy. Critics from the monetarist and other traditions go further, doubting that discretionary stimulus can be timed well enough to help at all, given the lags in recognising a recession, passing a budget, and spending the money.
So hold two things at once, as a good economist must. Keynes's core insight — that demand can fall short, that spending circulates and amplifies, that universal thrift can shrink the pie — is genuinely illuminating and reshaped how the world manages recessions. Yet the *quantitative* claims built on top of it remain live, evidence-dependent disputes, not settled facts. The next guide hands you the picture that lets economists argue about all this on a single diagram: the aggregate-demand-and-supply model, where Keynes's demand-driven slumps and his critics' supply-side concerns finally share one stage.