The Business Cycle Explained: Phases, Output Gaps, and How Recessions Are Dated

The business cycle is the repeated pattern of rises and falls in a country's total output around its long-run growth path. It has four phases: expansion, peak, contraction, and trough. Output is measured by real GDP, so the cycle is about real production changing, not prices changing. The phases repeat, but they are not regular: expansions and contractions vary in length and depth every time, which is why no economist can tell you when the next turning point arrives.

This guide works through each phase and what happens to output, employment, and prices inside it; separates the cycle from long-run growth using the output gap; explains how a recession is actually dated (which is not the rule most people quote); sorts the indicators into leading, coincident, and lagging; lists what causes cycles; and connects the whole thing to the AD-AS model and the Phillips curve. It is written to serve AP Macroeconomics, IB Economics, A-level, and college introductory courses at once, and the differences in vocabulary between them are flagged where they matter.

The four phases

### Expansion

An expansion is a period when real GDP is rising. Firms produce more, so they hire more, so household income rises, so spending rises, which pulls production up again. Unemployment falls as this loop runs, and the part of unemployment that falls is cyclical unemployment, the joblessness caused by weak demand rather than by workers switching jobs or by mismatched skills. Frictional and structural unemployment stay behind even in a strong expansion, which is why the target is full employment rather than zero unemployment. The full breakdown is in the guide to the types of unemployment.

Prices in an early expansion are usually calm, because the economy still has idle workers and unused factories, so extra demand can be met with extra output. Late in an expansion the picture changes. Spare capacity runs down, employers compete for a shrinking pool of available workers, wages rise, input costs rise, and inflation picks up.

### Peak

The peak is the turning point at which output stops rising and begins to fall. It is the highest level of real GDP in that cycle, not the fastest rate of growth, and those two are easy to confuse: growth normally slows for a while before the peak arrives.

At the peak, unemployment is at or near its lowest point of the cycle and inflation is usually at or near its highest, because total spending is pressing against what the economy can physically produce. Nobody announces a peak while it is happening. It is identified afterwards, once enough data has been collected and revised.

### Contraction

A contraction is a period of falling real GDP. Firms facing weaker sales cut production, delay investment, run down inventories, and lay off workers. Cyclical unemployment rises. Household income falls, so spending falls, which weakens sales again, and the loop that drove the expansion now runs in reverse.

Inflation typically slows during a contraction, since firms with unsold output and workers with weak bargaining power both push in the same direction. Falling inflation is disinflation, which is different from falling prices. A sustained contraction that is deep, broad, and long enough gets called a recession. A far deeper and longer version gets called a depression, and where the line sits is covered in recession versus depression.

### Trough

The trough is the turning point at which output stops falling and begins to rise. It is the lowest level of real GDP in that cycle. Unemployment is high at the trough and, importantly, often keeps rising for months afterwards, because firms wait until a recovery looks real before hiring. Inflation is usually at its weakest. From the trough, a new expansion starts and the cycle repeats.

You can drag through the phases and watch output move against trend in the business cycle sandbox.

Cycle versus long-run growth: the output gap

The single idea that separates a confident answer from a muddled one is this: an economy has both a trend and a fluctuation around that trend, and they have different causes.

Long-run growth is the trend. It is the rise in potential output, the amount an economy can produce when its labor and capital are used at normal, sustainable rates. Potential output rises when the workforce grows, when the capital stock grows, or when technology and skills improve productivity. These are supply-side forces and they work slowly.

The business cycle is the fluctuation. It is actual output moving above and below potential, driven mostly by swings in total spending. These are demand-side forces (with supply shocks as the main exception) and they work fast.

The measure that ties the two together is the output gap:

Output gap = (Actual real GDP - Potential real GDP) / Potential real GDP x 100

The result is a percentage of potential output. If actual real GDP is 9.7 trillion and potential real GDP is 10 trillion, the gap is (9.7 - 10) / 10 x 100 = -3 percent, so the economy is producing 3 percent below its capacity. A negative output gap means actual output is below potential and resources are idle. A positive output gap means actual output is above potential, which is possible in the short run because workers take overtime and factories run extra shifts, but it is not sustainable.

Two notes on wording, because the exams differ. A-level papers usually ask about "negative" and "positive" output gaps and often draw them on a production possibility diagram or against a trend line. AP Macroeconomics and IB usually ask about the same idea under the gap names below, drawn on AD-AS. They are the same concept.

### Recessionary and inflationary gaps

A recessionary gap, also called a deflationary gap in IB materials, exists when short-run equilibrium output is below potential output. Unemployment is above its natural rate, cyclical unemployment is positive, and there is downward pressure on inflation. On an AD-AS diagram it is the horizontal distance between the short-run equilibrium output and the vertical long-run aggregate supply curve, with equilibrium sitting to the left of LRAS.

An inflationary gap exists when short-run equilibrium output is above potential output. Unemployment is below its natural rate, labor and other inputs are stretched, and inflation is being pushed up. On the diagram, equilibrium sits to the right of LRAS.

A useful companion result is Okun's law, the empirical rule of thumb that output below potential and unemployment above its natural rate move together, with each percentage point of extra unemployment associated with output falling roughly two percent below potential. The exact coefficient varies by country and period, so treat it as an approximation rather than a law of nature.

How a recession is actually dated

Almost every student first learns that "a recession is two consecutive quarters of falling real GDP." That is a rule of thumb, and it is worth knowing precisely what kind of rule it is.

The two-quarter rule is a quick, mechanical test based on one variable. It has real advantages: it is easy to compute, easy to compare across countries, and available as soon as the quarterly national accounts are published. Several statistical agencies and news organizations use it as a shorthand for a "technical recession." Its weaknesses are the flip side of the same coin. It ignores employment and income entirely, it can miss a downturn that is severe but concentrated in a single quarter, and it can flag two mild negative quarters that nobody experienced as a downturn. It also flips whenever GDP figures are revised, which they routinely are.

The way recession dating actually works is committee-based and much broader. In the United States, the Business Cycle Dating Committee of the National Bureau of Economic Research (NBER) identifies the months in which peaks and troughs occurred. The NBER is a private nonprofit research organization, not a government agency, but its chronology is the standard reference for US business cycle dates. The committee does not use the two-quarter rule and does not require any fixed threshold. It looks for a significant decline in economic activity that is spread across the economy and lasts more than a few months, weighing three criteria together: depth (how severe the decline is), diffusion (how widely it is spread across sectors), and duration (how long it lasts). An unusually extreme reading on one criterion can partly offset a weaker reading on another.

The committee also looks at several monthly measures rather than quarterly GDP alone, including employment, real personal income excluding transfer payments, real consumer spending, industrial production, and wholesale and retail sales. Because it works with monthly data, it can pin a turning point to a specific month, and its published chronology lists both monthly and quarterly peak and trough dates. Because it waits for data to settle, its announcements come well after the fact, often by many months, and it does not attempt real-time calls.

Other economies use similar committee approaches: the euro area has a dating committee run through the Centre for Economic Policy Research, and several national statistical bodies maintain their own chronologies. The general lesson holds everywhere. A recession is a judgment about the depth, breadth, and length of a broad decline in activity, and the two-quarter rule is a fast proxy for that judgment rather than the definition of it. Say both in an exam answer: quote the rule of thumb, then state that the recognized dating bodies use a broader, committee-based, multi-indicator method.

Leading, coincident, and lagging indicators

Because turning points are only confirmed long after they happen, forecasters sort economic data by its timing relative to the cycle.

Leading indicators turn before the overall economy turns, so they are used to anticipate the next phase. Common examples: new building permits, new orders for capital goods and consumer durables, average weekly hours worked in manufacturing, initial claims for unemployment insurance (which rise before a downturn), consumer and business confidence surveys, stock prices, and the slope of the yield curve. Leading indicators give warning and give false alarms, which is why forecasters watch a composite of them rather than any single series.

Coincident indicators turn at roughly the same time as the economy, so they tell you which phase you are in now. Common examples: real GDP, total nonfarm employment, industrial production, real personal income excluding transfers, and manufacturing and trade sales. These are the closest thing to a live reading of the cycle, and they are close to the list official dating committees weigh.

Lagging indicators turn after the economy has already turned, so they confirm a phase that has begun. Common examples: the unemployment rate, the average duration of unemployment, the ratio of consumer credit to income, outstanding commercial and industrial loans, the prime interest rate, and unit labor costs. The unemployment rate is the one exam questions ask about most, and the reason it lags is behavioral: firms cut hours before they cut staff, and they raise hours and hire temporary workers before they commit to permanent hires.

More background on the composite measures is in the leading economic indicators entry.

What causes business cycles

No single cause explains every cycle. Four families of explanation cover most of what courses ask about.

Demand shocks. In the mainstream demand-side account, sudden changes in total spending are the most common driver. A collapse in consumer confidence, a fall in business investment, a drop in export demand from a trading partner, a fall in asset or housing prices that reduces household wealth, or a financial crisis that restricts lending all cut aggregate demand and push output below potential. Positive versions (a construction boom, a surge in exports) push the other way. Because spending changes feed on themselves through the multiplier effect, even a moderate initial shock can move output a long way.

Supply shocks. A sudden change in production costs or productive capacity shifts short-run aggregate supply. An energy price spike, a crop failure, a war, or a disruption to global supply chains raises costs, cuts output, and raises prices at the same time, which is stagflation. Favorable supply shocks (a fall in oil prices, a productivity leap) raise output and reduce inflation together. Supply shocks are the reason a downturn does not always come with falling inflation.

Monetary policy. Central banks change interest rates to steer demand, and those changes are themselves a source of fluctuation. Raising rates to fight inflation deliberately slows spending, and if the tightening goes further than intended, it can turn a slowdown into a contraction. Cutting rates does the reverse. Credit conditions matter as much as the policy rate: when banks tighten lending standards, investment and durable purchases fall regardless of the headline rate.

Expectations. Beliefs can be self-fulfilling in macroeconomics. If firms expect weak sales they cut investment and hiring now, which produces the weak sales they expected. If households expect job losses they save more and spend less, which causes the job losses. This is why confidence surveys are treated as leading indicators and why central bank communication is treated as a policy tool in its own right.

Two mechanisms are worth adding for stronger answers. Inventory cycles amplify small demand changes, because firms cut orders more sharply than sales fall while they run down stock, then reorder sharply once stock is depleted. And real business cycle theory argues that fluctuations are largely the efficient response of an economy to genuine productivity shocks rather than a failure of demand, which is the main intellectual challenge to the demand-side account.

Policy responses: automatic stabilizers and discretionary action

Governments respond to the cycle in two distinct ways, and exams reward keeping them separate.

Automatic stabilizers work with no new legislation and no decision by anyone. A progressive income tax collects proportionally less as incomes fall in a contraction and more as they rise in an expansion. Unemployment benefits and means-tested transfers pay out more automatically when joblessness rises and less when it falls. Corporate taxes fall sharply with profits. The combined effect cushions disposable income on the way down and restrains it on the way up, which damps the cycle at both ends. The mechanism, with a worked multiplier example, is in automatic stabilizers explained and summarized in the automatic stabilizers entry.

Discretionary policy requires an active decision. On the fiscal side, that means changing tax rates or government spending through the legislature. On the monetary side, it means a central bank changing its policy rate, its asset holdings, or its guidance. Expansionary versions of both shift aggregate demand right to close a recessionary gap; contractionary versions shift it left to close an inflationary gap. The comparison of the two toolkits, including which authority controls which, is in fiscal policy versus monetary policy.

The reason discretionary policy is hard is time lags. The recognition lag is the delay before anyone knows the phase has turned, which the dating discussion above shows can be substantial. The implementation lag is the delay in deciding and enacting a response, which is long for fiscal policy (budgets must pass) and short for monetary policy (a committee can act at a scheduled meeting). The impact lag is the delay before the action changes behavior, which is short for fiscal spending and long for interest rates, since investment plans take time to adjust. Add these up and a stimulus aimed at a contraction can arrive during the recovery, adding demand the economy no longer needs. Automatic stabilizers have essentially no recognition or implementation lag, which is their central advantage.

The link to AD-AS

The business cycle and the AD-AS model describe the same events at different levels of detail. The cycle diagram shows output over time; AD-AS shows why output sits where it does at a point in time.

An expansion driven by demand is aggregate demand shifting right along an upward-sloping short-run aggregate supply curve, raising both real GDP and the price level. A contraction driven by demand is AD shifting left, lowering both. A supply-driven contraction is SRAS shifting left, which lowers output while raising the price level. Vertical long-run aggregate supply sits at potential output, so the horizontal distance from short-run equilibrium to LRAS is the output gap in levels, which the formula above converts into a percentage of potential output.

The model also explains self-correction. With a recessionary gap, high unemployment eventually pushes nominal wages and other input prices down, SRAS shifts right, and output returns to potential at a lower price level. With an inflationary gap, tight labor markets push wages up, SRAS shifts left, and output returns to potential at a higher price level. The policy debate is largely about how long that adjustment takes and whether it is worth waiting. You can shift the curves directly in the AD-AS sandbox and step through worked shocks in the graph walkthroughs.

The link to the Phillips curve

The Phillips curve is the cycle viewed through the pairing of inflation and unemployment. Along a short-run Phillips curve, a demand-driven expansion moves the economy up and to the left: unemployment falls below its natural rate and inflation rises. A demand-driven contraction moves it down and to the right: unemployment rises and inflation falls. That is the standard trade-off, and it is simply the AD shifts above re-plotted.

The long-run Phillips curve is vertical at the natural rate of unemployment, which corresponds to potential output and a zero output gap. So a positive output gap, an inflationary gap, and unemployment below the natural rate all describe one situation, and a negative output gap, a recessionary gap, and unemployment above the natural rate all describe another. Supply shocks break the trade-off by shifting the whole short-run curve to the right, or equivalently upward, raising inflation and unemployment together, which is stagflation again. The full treatment is in the Phillips curve and the inflation-unemployment trade-off, and the causes of the inflation side are set out in what causes inflation.

Common mistakes to avoid

Confusing the peak with the fastest growth. The peak is the highest level of output, which occurs when the growth rate has already fallen to zero. Growth normally slows for some time before the turning point.

Saying a recession means falling nominal GDP. The cycle is measured in real GDP. Nominal GDP can fall from deflation alone and can rise during a contraction if inflation is high enough. The real versus nominal distinction is doing the work here.

Treating the two-quarter rule as the official definition. State it as the rule of thumb it is, then note that dating committees use depth, diffusion, and duration across several monthly indicators.

Calling the unemployment rate a leading indicator. It is the standard lagging indicator. Initial jobless claims lead, total employment is coincident, and the unemployment rate lags.

Treating a contraction as a fall in potential output. In a normal downturn, factories and workers still exist and are simply idle. Potential output is the trend, and the contraction is the gap opening beneath it.

Mixing up automatic stabilizers and discretionary policy. If it required a vote, it is discretionary. If it happened because the tax and benefit rules were already written, it is automatic.

Practice and connect

Work the phases and gaps until you can do three things without hesitating: name what happens to output, unemployment, and inflation in each phase; compute an output gap from actual and potential GDP and state whether it is recessionary or inflationary; and translate any of it onto an AD-AS diagram. Then drill the model in the business cycle module, test the graph work in the business cycle sandbox and the Phillips curve sandbox, and connect the labor market side through the unemployment and inflation module. For a structured route through the rest of macroeconomics, start at the learn economics hub.