Economic indicators are often sorted by when they tend to turn relative to the business cycle. A leading indicator tends to turn before the overall economy does, so it is watched for early warning of a peak or a trough. A coincident indicator moves at about the same time as the economy, so it describes where activity is now. A lagging indicator turns after the economy has already turned, so it confirms a change that has happened. The Conference Board publishes a leading, a coincident and a lagging index for the United States built on exactly this idea, and the OECD publishes composite leading indicators for many economies.
This page is for AP Macroeconomics and intro economics study. It is educational only and is not investment, financial, tax, or legal advice. It makes no forecast about the current economy and no claim about markets.
Why sort indicators by timing at all
The business cycle is a sequence of expansions and contractions separated by turning points. The NBER, which keeps the official chronology of US business cycles, describes expansions as the periods between a trough and a peak and recessions as the periods between a peak and a trough. The hard part is that nobody knows a turning point has arrived on the day it happens. The NBER says its own approach is retrospective: it waits until enough data are available to avoid major revisions, so it tends to identify a peak a number of months after it actually occurred.
That delay is the reason the three groups exist. If you want an early hint, you look at series that historically move first. If you want to know the state of the economy right now, you look at series that move with it. If you want confirmation that a turn really happened, you look at series that move afterwards. The Conference Board says its composite indexes are designed to signal peaks and troughs, and that combining many independent indicators reveals common turning points more clearly and convincingly than any single component.
Leading indicators
The Conference Board describes its Leading Economic Index (LEI) as a predictive tool that anticipates turning points in the business cycle by around seven months. Its US page lists ten components:
- Average weekly hours in manufacturing
- Average weekly initial claims for unemployment insurance
- Manufacturers' new orders for consumer goods and materials
- The ISM Index of New Orders
- Manufacturers' new orders for nondefense capital goods excluding aircraft
- Building permits for new private housing units
- The S&P 500 Index of Stock Prices
- The Leading Credit Index
- The interest rate spread, measured as 10 year Treasury bonds less the federal funds rate
- Average consumer expectations for business conditions
The Conference Board's description of components explains why several of these move early. Employers usually adjust hours before they hire or lay off, so average weekly hours tend to lead. New claims for unemployment insurance are more sensitive to business conditions than total employment, so they also lead; because claims rise when conditions worsen, the series is inverted inside the index. Building permits signal construction activity, which typically leads most other production. The interest rate spread is the gap between long and short rates, often called the yield curve, and the Conference Board notes that when it turns negative its record as an indicator of recessions is particularly strong. You can practise the arithmetic of a spread with the yield curve spread calculator.
The OECD runs a similar idea across countries. It defines its composite leading indicator (CLI) as an index designed to provide early signals of turning points in business cycles, showing how economic activity fluctuates around its long term potential level. Each country's CLI is built from a set of short term economic series, it is measured as an amplitude adjusted index with a long term average of 100, and it is compiled for the G20 countries plus Spain and five zone aggregates. The OECD also stresses that CLIs show short term movements in qualitative rather than quantitative terms. In plain words: a CLI is meant to tell you about the direction and turning of the cycle, not to give you a growth rate.
Coincident indicators
The Conference Board says its Coincident Economic Index (CEI) reflects current economic conditions and is highly correlated with real GDP. It has four components:
- Payroll employment
- Personal income less transfer payments
- Manufacturing and trade sales
- Industrial production
These four line up closely with how US recessions are dated. The NBER says its committee looks at a range of monthly measures of aggregate real economic activity, including real personal income less transfers, nonfarm payroll employment, employment from the household survey, real personal consumption expenditures, real manufacturing and trade sales, and industrial production. It adds that in recent decades it has put the most weight on real personal income less transfers and nonfarm payroll employment. The Conference Board notes the same overlap: its four coincident components are among the data used to determine recessions in the US. That is why coincident series are the ones to read when the question is whether the economy is shrinking or growing right now.
Lagging indicators
The Conference Board also publishes a Lagging Economic Index (LAG). Its description of components page (last revised 6 February 2012) lists seven series:
- Average duration of unemployment
- The ratio of manufacturing and trade inventories to sales
- The change in labour cost per unit of output in manufacturing
- The average prime rate charged by banks
- Commercial and industrial loans outstanding
- The ratio of consumer installment credit outstanding to personal income
- The change in the Consumer Price Index for services
The same page explains the lag in each case. The average duration of unemployment falls only after an expansion gains strength, and its sharpest increases tend to come after a recession has begun, so it is inverted inside the index. The inventories to sales ratio typically peaks in the middle of a recession, because inventories pile up when sales fall short. Business loans tend to peak after an expansion peaks, and their troughs are typically seen more than a year after a recession ends. Consumers tend to hold off borrowing until months after a recession ends. The Conference Board suggests that service sector inflation tends to rise in the first months of a recession and fall in the first months of an expansion because of recognition lags and other rigidities.
A lagging series is not useless because it is late. Its job is confirmation. If a leading index fell and coincident series weakened, a later turn in the lagging index is evidence that the change in the cycle was real rather than a blip.
One sketch of the timing
The diagram below is a teaching sketch, not real data. It shows three curves passing through the same peak and trough of the business cycle, each shifted in time.
How the signals are read, and why they can mislead
A single fall in a leading index is not a recession call. The Conference Board's own reading uses what it calls the 3Ds: duration, depth and diffusion. Duration is how long a decline has lasted, depth is how large it is, and diffusion is how widespread it is across the components. Its rule signals an impending recession only when two things happen at the same time: the six month diffusion index is at or below 50, meaning most components are weakening, and the six month growth rate of the LEI, at an annualized rate, falls below a threshold of minus 4.3 percent.
The need for a two part rule is itself the caveat. Several leading components reflect orders, credit, stock prices and consumer expectations, which is part of why they move early. It also means a dip in some of them does not always become a broad downturn, which is why the rule asks for both depth and breadth. A falling leading index is a warning to look harder at coincident data, not proof. The OECD's point that its CLIs are qualitative rather than quantitative makes the same warning from another angle. The timing advantage is real but approximate, and a lead of around seven months on average does not mean seven months every time.
Coincident data have their own limit: they are published with a delay and revised later, which is part of why the NBER waits before dating a turn. Lagging data are, by design, too late to warn you about anything. Using all three together is the point.
A quick checker
Pick the group each series belongs to in the Conference Board's US indexes.
Series 1. Building permits for new private housing units.
Yes. Building permits are one of the ten LEI components. Permits come before construction, which typically leads other production.
Not quite. Building permits are a leading component, because permits are issued before construction begins.
Series 2. Industrial production.
Yes. Industrial production is one of the four CEI components, and the NBER lists it among the measures it uses to date recessions.
Not quite. Industrial production measures current output, so it is a coincident component.
Series 3. Average duration of unemployment.
Yes. It is in the lagging index. Spells of unemployment shorten only after an expansion gains strength, and lengthen most sharply after a recession has begun.
Not quite. The average duration of unemployment responds late, so the Conference Board places it in the lagging index.
A dated example from a real release
This example is here to show how the three indexes are reported side by side. It is a snapshot of one release, not a forecast. In its release updated 22 October 2026, the Conference Board reported that the US LEI was unchanged in September 2026 at 99.6 (2016 = 100), that the CEI rose 0.3 percent to 115.3, and that the LAG rose 0.5 percent to 121.1. It said there was no recession signal under its 3Ds rule, because the LEI's six month growth rate increased and the diffusion index rose slightly above 50. Notice the structure: one leading reading that was flat, one coincident reading that rose, one lagging reading that rose, and a rule that looks at breadth and duration rather than a single month. Figures are revised, so always check the latest release before quoting them.
How this fits the rest of the course
Indicators are the data side of the business cycle. Coincident data tell you whether the economy is in an expansion or a contraction, which links to how a recession is defined in the US and the UK. A weak coincident picture usually goes with a negative output gap, meaning actual output below potential. For the measurement side of output, start with the GDP topic hub. The AP Macroeconomics study hub maps these ideas to the course units.
Common exam mistakes
- Treating a leading indicator as a forecast. A leading index gives an early signal of a possible turning point. It does not say how deep a downturn will be or promise that one will happen.
- Mixing up coincident and lagging. Payroll employment moves with the economy and is coincident. The average duration of unemployment responds late and is lagging. Both are about jobs, but they sit in different groups.
- Expecting the NBER to call a recession in real time. The NBER says its dating is retrospective and that it waits months after a peak or trough before announcing it.
- Reading one month as a trend. The Conference Board's rule needs both a large enough six month decline and a decline spread across most components.
- Forgetting that some series are inverted. Initial jobless claims and the average duration of unemployment rise when conditions worsen, so the Conference Board reverses their signs inside its indexes.
Related reading
- The business cycle explained
- What is a recession?
- Output gap and the output gap calculator
- What is the yield curve? and the yield curve spread calculator
- Supply shocks versus demand shocks
- GDP topic hub
- AP Macroeconomics study hub and the explainers index
Sources
- The Conference Board, US Leading Indicators (release updated 22 October 2026; retrieved 9 Nov 2026)
- The Conference Board, Business Cycle Indicators: Description of Components (revised 6 February 2012; retrieved 9 Nov 2026)
- NBER, Business Cycle Dating (retrieved 9 Nov 2026)
- OECD, Composite leading indicator (CLI) (retrieved 9 Nov 2026)
Frequently asked questions
What is the difference between leading, coincident and lagging indicators?
A leading indicator tends to turn before the overall economy, a coincident indicator moves at about the same time, and a lagging indicator turns afterwards. Leading series are watched for early warning, coincident series describe current conditions, and lagging series confirm that a turn has happened.
What are examples of leading economic indicators?
The Conference Board's US Leading Economic Index includes average weekly hours in manufacturing, initial claims for unemployment insurance, new orders, building permits, stock prices, a leading credit index, the spread between 10 year Treasury yields and the federal funds rate, and consumer expectations.
Is unemployment a leading or lagging indicator?
It depends on the series. New claims for unemployment insurance are a leading component in the Conference Board's index, payroll employment is coincident, and the average duration of unemployment is lagging.
Can leading indicators predict a recession?
They can give early signals of a possible turning point, but they are not guarantees. The Conference Board signals an impending recession only when a decline in its leading index is both large enough over six months and spread across most components, and the OECD describes its leading indicators as qualitative rather than quantitative.