Business Cycles Explained: Why Economies Expand, Peak, Contract, and Recover

Understanding business cycles — simply put

A business cycle is the recurring rise and fall of economic activity around its long-run trend. Output, jobs, incomes and spending grow for a stretch, reach a high point, fall back, hit a low point and then start growing again. Economists name the four phases expansion, peak, contraction and trough. A contraction that is deep, broad and lasting is a recession.

The cycle is not a clock. Expansions in the United States have lasted anywhere from about two years to more than ten, and recessions from two months to a year and a half. What repeats is the pattern, not the timing: spending and confidence build on each other on the way up, and pull each other down on the way down.

How does it actually help?

Take the U.S. economy between 2007 and 2009. The expansion that began in November 2001 peaked in December 2007. Over the next 18 months output fell, firms cut investment and the unemployment rate doubled from 5.0% to a high of 10.0% in October 2009. The trough came in June 2009, and the next expansion ran 128 months, the longest on record, until February 2020.

The tool that ties those dates together is the output gap: the percentage difference between actual real GDP (Y) and potential GDP (Y*), the output the economy could produce at full employment without pushing inflation up. Written out, output gap = 100 × (Y − Y*) / Y*. In an expansion the gap closes and can turn positive; in a contraction it turns sharply negative. In 2009 U.S. output sat several percent below the Congressional Budget Office's estimate of potential, which is what a deep recession looks like in one number.

Why does this matter?

The phase of the cycle changes the meaning of almost every economic headline. A strong jobs report late in an expansion can unsettle markets because it signals overheating and higher interest rates; the same report early in a recovery is plainly good news. Central banks set interest rates according to where they think the economy is in the cycle, governments time spending and tax changes around it, and firms use it to decide when to hire and invest.

What makes a cycle turn?

No single force drives every cycle. Demand shocks, such as a collapse in household confidence or a credit squeeze, push spending below what the economy can produce. Supply shocks, such as an energy price spike or a pandemic, cut what the economy can produce in the first place. Financial conditions amplify both: easy credit lengthens expansions and tight credit deepens downturns. Whether a turning point is reliable to call depends on watching several indicators at once rather than any one of them.

The takeaway

Business cycles are the economy's swings around its trend, measured by the output gap and dated by turning points. The rest of this piece works through the real U.S. cycle dates, the indicators used to spot turns early, and the formulas behind them.

A worked example: four U.S. cycles, 1990–2020

In the United States, recessions are dated by the Business Cycle Dating Committee of the National Bureau of Economic Research (NBER), using a broad set of monthly indicators rather than GDP alone. The table below lists the last four peaks and troughs, with the length of each contraction, the length of the expansion that led into it, and the highest unemployment rate reached afterwards.

Three things stand out. Expansions last far longer than contractions: across these four cycles the average expansion ran about 103 months and the average contraction 9. Recessions differ enormously in depth: the 2020 contraction lasted two months but pushed unemployment to 14.8%, while the 2001 recession was mild. And unemployment keeps rising after the official trough. In 1991 and 2001 the jobless rate peaked more than a year after output had started growing again, which is why recoveries often feel like recessions to the people living through them.

Reading a turning point as it happens: the Sahm rule

The NBER announces turning points months after they happen; the December 2007 peak was not declared until December 2008. Forecasters therefore use real-time rules. The best known is the Sahm rule, proposed by economist Claudia Sahm: a recession is likely under way when the three-month average unemployment rate rises 0.5 percentage points or more above its lowest three-month average of the previous 12 months. The table shows the calculation with illustrative monthly figures.

A rise of 0.8 points in the monthly rate looks modest, but because unemployment rarely climbs that fast outside a recession, the rule has historically signaled U.S. recessions within a few months of their start.

The conditions for reading the cycle correctly

• Potential output has to be estimated well. The output gap compares actual GDP with a number nobody observes directly; if potential is overstated, every expansion looks weaker than it is and policy stays too loose.
• Data must be read with revisions in mind. First GDP estimates are revised several times, so a quarter that looked like growth can later turn negative. Turning points called on first-release data are provisional.
• Several indicators have to agree. Coincident indicators (output, employment, real income, sales) confirm the current phase; leading indicators (the yield curve, new orders, building permits, confidence surveys) hint at the next one. Any single series gives false signals.

The formula

Readers who only need the intuition can skip this section. Three relationships do most of the work in business cycle analysis.

Okun's law links the cycle in output to the cycle in jobs. For the United States, a common estimate is that each percentage point of real GDP growth below its trend rate adds about half a point to the unemployment rate over a year.

The Sahm rule turns the unemployment rate into a real-time recession signal.

Estimating potential output itself is a statistical problem. Agencies use production-function methods, building potential from trend labor, capital and productivity, while researchers often use filters that separate a time series into trend and cycle. Both approaches tend to revise the size of past output gaps as new data arrive.

Competing explanations

Keynesian and New Keynesian models treat cycles mainly as swings in aggregate demand that prices and wages adjust to only slowly, which leaves room for monetary and fiscal policy to stabilize output. Real business cycle models, developed by Finn Kydland and Edward Prescott in the 1980s, treat fluctuations as efficient responses to real shocks such as changes in productivity. Monetarist accounts, associated with Milton Friedman and Anna Schwartz, stress swings in money and credit. Financial-cycle research, revived after 2008, adds that booms in credit and house prices make the downturns that follow longer and deeper.

When business cycle analysis isn't the right tool

• Structural change is not a cycle. A permanent shift such as an industry's decline or an aging workforce changes potential output itself; treating it as a temporary gap leads to the wrong policy response. Long-run growth questions belong to growth theory and GDP accounting.
• Turning points are hard to forecast. Leading indicators improve the odds but carry false alarms; the U.S. yield curve, for example, inverted in 2022 and stayed inverted through 2023 while the economy kept growing. Probability models built on several indicators beat any one signal.
• Cycles differ across countries. Rules calibrated on U.S. data, such as the Sahm rule, need re-estimating before use elsewhere; the euro area's turning points are dated by a separate committee at the Centre for Economic Policy Research (CEPR).

The takeaway

Business cycles are measured by the output gap, dated by turning points and tracked in real time with rules like Sahm's. The link from cycles to policy runs through monetary policy and fiscal policy, and to prices through inflation.

Further reading

• Understanding GDP — internal
• Monetary Policy Explained — internal
• Fiscal Policy and Multipliers Explained — internal
• Time Series Analysis Explained — internal
• NBER: U.S. business cycle expansions and contractions — external reference
• Congressional Budget Office: potential GDP (FRED series GDPPOT) — external reference

Data note: cycle dates are from the NBER and unemployment peaks from the U.S. Bureau of Labor Statistics. The two charts pull real GDP, potential GDP and the unemployment rate live from FRED. The Sahm-rule table uses illustrative figures; the Okun example uses a round illustrative coefficient.