A recession is a broad, sustained fall in economic activity — production, jobs, spending, and income all shrink together rather than one sector alone. What causes a recession is rarely a single event. Economists generally describe a small set of forces that recur across downturns: credit cycles that build up debt, sudden shocks that hit spending or supply, and confidence collapses that turn caution into contraction.
The honest qualification comes first: these forces explain why recessions happen in general, but nobody can say in advance which combination will fire, or when. Forecasters with full access to the data have repeatedly missed turning points. This piece explains the mechanisms, not the timing.
The word why itself is worth a pause. As Vocabulary.com defines it, why names the cause or intention underlying a situation — and in economics the cause is usually layered. A recession typically has a trigger, a vulnerability underneath it, and an amplifier that spreads the damage. Confusing the three is how most popular explanations go wrong.
What is a credit cycle, and how does it turn into a downturn?
A credit cycle is the long swing between easy borrowing and tight borrowing. When credit is cheap and confidence is high, households take on mortgages and car loans, and businesses borrow to expand. Spending rises, hiring rises, and asset prices — houses, shares — rise with them. The rise itself feeds the borrowing, because rising collateral makes lenders more willing to lend.
The turn comes when debt outgrows the income available to service it. Borrowers cut spending to keep up payments. Lenders, seeing losses, tighten standards, which cuts off even sound borrowers. Falling asset prices reduce the collateral behind existing loans, so the tightening deepens. This is the mechanism economists call deleveraging: the whole economy pays down debt at once, and because one person's spending is another person's income, the paying-down shrinks everyone's earnings.
Two features make credit-driven recessions distinctive. They tend to last longer, because households and firms need years, not quarters, to repair balance sheets. And they tend to be harder to reverse, because lowering interest rates helps less when the problem is too much debt rather than too little willingness to borrow. Readers following this should also see The economics of local newspaper closures, traced from classifieds to the last printing press.
How do outside shocks start a recession?
Not every downturn begins with debt. Some begin with a shock — a sudden change that hits the economy from outside the normal cycle. An energy price spike raises costs across the economy and squeezes household budgets. A pandemic closes workplaces and shifts what people buy. A disruption to trade or a key input forces factories to slow. A war destroys supply chains and pushes prices up. This connects to our earlier piece, Cloudflare's crawler deadline puts a price on AI's use of news.
Shocks matter through a simple channel: they force choices. A household facing higher fuel and food bills spends less on everything else. A firm facing higher input costs raises prices, delays investment, or cuts staff. None of these decisions is dramatic on its own. Multiplied across millions of households and firms, they subtract enough demand to push output down.
Shocks also interact with debt. An economy with heavy borrowing is more fragile in the face of a shock, because borrowers have less slack to absorb a squeeze. So the two causes often appear together: the shock is the trigger, and the credit build-up is the vulnerability it exposes.
What role does confidence play?
Confidence is the amplifier, and often the least predictable part. Spending depends not only on what people earn but on what they expect to earn. If workers fear layoffs, they save rather than spend. If firms fear weak sales, they postpone hiring and investment. Each act of caution is rational. Together, they produce exactly the weakness everyone feared — a dynamic economists call a self-fulfilling prophecy.
This is why recessions can arrive with a suddenness that the underlying data did not fully show. Expectations can shift in weeks. A bank failure, a market fall, or a run of bad news can move consumers and businesses from spending to hoarding before any official statistic confirms a contraction. The measurement problem is real: output figures arrive with a lag, so a recession is often declared months after it began.
How do policymakers respond — and what are the limits?
Governments and central banks have standard tools. A central bank can lower interest rates, making borrowing cheaper and supporting spending and investment. A government can raise spending or cut taxes, putting demand directly back into the economy. Lender-of-last-resort support can stop a banking problem from freezing credit for everyone else.
Each tool has limits. Lower rates do less when rates are already low, or when the problem is debt overhang rather than the price of credit. Fiscal support adds to public debt, which constrains some governments more than others. And policy acts with delays: it takes time to recognize the downturn, decide the response, and feel the effects. A policy sized for the recession that just ended can arrive during the next one.
There is also a genuine tension. Stimulus that arrives too late or too large can feed inflation, which brings its own squeeze on households. Stabilizing an economy is a calibration problem, not a switch.
Why can't economists predict the next recession?
Because the causes are multiple, the timing depends on shocks that are by nature unforecastable, and the confidence channel can flip without warning. Economic models are good at describing vulnerabilities — where debt is high, where asset prices look stretched, where supply is fragile. They are much weaker at saying when a vulnerability will meet a trigger.
The practical reading is not fatalism. Understanding the mechanisms tells you what to watch: credit growth relative to income, the gap between asset prices and the cash flows behind them, and measures of household and business sentiment. Those indicators mark fragility, not a date. Anyone selling certainty about the timing is going beyond what the tools can deliver.
What this means
What this means for readers is a modest but useful conclusion. Recessions happen because spending, borrowing, and expectations are linked in loops that can run in reverse: credit turns, shocks bite, and caution compounds. The evidence for any specific downturn is always a mix, and the share of each ingredient varies. What remains unknown, in every cycle, is the trigger and the date — which is why the honest answer to what causes a recession describes mechanisms, not predictions.
