The Economic Cycle Phases: What Actually Dates a Recession

Almost every explanation of the economic cycle opens with the same definition of a recession: two consecutive quarters of falling real GDP. The body that actually dates recessions in the United States does not use that rule and says so plainly.

That gap matters more than it looks. If the definition in circulation is not the one being applied, then the calendar of expansions and contractions you carry in your head is not the one the official record will show.

What follows sets out what a cycle is, what genuinely decides a turning point, which data the decision rests on, and how long the label takes to arrive.

What an Economic Cycle Is, and What It Is Not

An economic cycle is the repeated movement of aggregate activity in an economy between periods of broad expansion and periods of broad contraction. The word aggregate is doing real work there. That two-way framing also has a blind spot, since the regime the phases cannot represent is the one where prices and activity move in opposite directions at the same time.

A cycle is not a single number rising and falling. It is a condition of the whole economy, which is why a decline confined to one sector does not constitute one.

It is also not periodic. Cycles are recurrent but they are not regular, so no phase has a fixed length and none of them arrives on schedule.

The most common misreading is treating the cycle as a forecast. Turning points are identified from data that already exist, which makes the framework a way of describing the past rather than a way of anticipating the future.

The Phases, and Why the Lists Disagree

Ask three sources how many phases a cycle has and you can easily get three answers. The four-phase version runs expansion, peak, contraction, trough. Longer versions split expansion into early and late stages, or separate recovery from expansion, producing five or six.

None of these lists is authoritative, because the phase vocabulary is descriptive rather than defined. What is defined is narrower and more useful: the turning points themselves.

A peak is the month in which activity reaches its high before a decline begins. A trough is the month in which activity reaches its low before a rise begins. Everything between a trough and the next peak is an expansion; everything between a peak and the next trough is a contraction.

That reduction is worth keeping. Two dated turning points are verifiable. A claim that an economy is in the late stage of an expansion is a judgment, and different sources will place the same month in different buckets.

What Actually Defines a Recession

In the United States, business cycle turning points are determined by the Business Cycle Dating Committee of the National Bureau of Economic Research. Its definition of a recession is a significant decline in economic activity that is spread across the economy and lasts more than a few months.

Read that definition again for what it does not contain. There is no number of quarters in it, no threshold percentage, and no reference to gross domestic product as the deciding series.

The definition instead points at three criteria: depth, diffusion and duration. Depth is how large the decline is, diffusion is how widely it is spread across the economy, and duration is how long it lasts.

The committee treats these as partly interchangeable. Its own explanation states that while each criterion needs to be met individually to some degree, extreme conditions revealed by one criterion may partially offset weaker indications from another.

That single sentence is why the two-quarter rule fails as a description. A decline that is unusually deep and unusually widespread can qualify without lasting long, and a shallow decline that drags on can qualify without ever being severe in a given quarter.

It also means a recession call is a judgment applied to a body of evidence, not the output of a formula. The two-quarter rule is popular precisely because it is mechanical, which is the same reason it does not match the record.

The Monthly Indicators That Decide It

Because a turning point is dated to a month, quarterly data cannot do the work on its own. The committee’s measures of aggregate economic activity are monthly series.

Those measures are real personal income less transfers, non-farm payroll employment, employment as measured by the household survey, real personal consumption expenditures, manufacturing and trade sales adjusted for price changes, and industrial production.

Real GDP is not absent, but it occupies a specific place. The expenditure-side and income-side estimates of real gross domestic product are described as important for determining quarterly peaks and troughs, and as not available monthly.

Two features of that list are worth noticing. It is dominated by measures of income, employment, production and sales rather than by any single headline aggregate, and it contains two separate employment series drawn from two different surveys.

The presence of non-farm payroll employment on it is the practical link to the release calendar. A series watched monthly by traders is also one of the inputs to the eventual dating decision, which is not true of most of what the calendar carries.

The same applies, less directly, to the consumer price index. Price data does not date the cycle, but it is what converts nominal sales and income into the real measures that do.

Why the Official Label Arrives Late

The dating decision is deliberately slow. The committee’s stated practice is to wait to identify a peak until a number of months after it has actually occurred, and it waits similarly before identifying a trough.

The reason is data quality. Monthly economic series are revised, sometimes substantially, and a turning point called on first estimates can be contradicted by the revised series later.

The consequence for anyone watching markets is direct. By the time a recession is officially dated, the months in question are long past and every price that was going to move on the underlying data has already moved.

This is what makes the framework poor as a timing device and reasonable as an interpretive one. It tells you what the economy was doing, with a confidence that only hindsight provides.

What This Means for Reading Data Releases

The cycle framework changes how a single release should be weighted, rather than telling you what to do about it.

A release that belongs to the dating committee’s list of measures carries information about the economy’s overall condition. A release that does not may still move prices, but it is not evidence about the cycle itself.

Diffusion is the criterion most often ignored when a single number disappoints. One weak month in one series is not a decline spread across the economy, and treating it as an early recession signal misapplies the definition.

Revisions deserve the same caution. Because the committee waits for revised data, a first estimate that looks decisive can lose that quality entirely once it is restated, which is why our page on the economic calendar treats the revision line as part of the release rather than a footnote.

Cycle conditions also sit behind the language central banks use, which is where central bank policy stance becomes relevant, and behind the co-movement of different markets covered under intermarket analysis. Neither of those relationships is stable enough to be treated as a rule.

Who This Page Is Not For

This page is not for anyone looking for a phase-based allocation rule. It names no asset that performs well in any given phase, because a relationship that holds across some past cycles is not a property of the phase.

It is not for anyone wanting a view on where any economy currently sits. Identifying a turning point requires data that are not yet complete, and the body that does the identifying will not name one until it is.

It is also not a forecasting method. Nothing on this page indicates when a peak or trough will occur, and the framework described is not designed to do that.

What it is for is reading the vocabulary correctly, and knowing which claims about the cycle can be checked against a published record.

Frequently Asked Questions