Skip to main content

How the NBER defines a recession?

When the U.S. economy enters a downturn, policymakers, investors, and news media wait for an official declaration. That declaration typically comes not from the Federal Reserve or the President's office, but from a private nonprofit research organization headquartered in Cambridge, Massachusetts: the National Bureau of Economic Research, or NBER. This article explains what the NBER is, why its recession definition is the gold standard, and how economists use it to understand economic cycles.

Quick definition: The NBER defines a recession as a significant decline in economic activity lasting more than a few months, visible in employment, industrial production, real income, and wholesale-retail sales. Their definition is the official standard used by economists and policymakers.

Key takeaways​

  • The NBER, not the government, officially declares when recessions begin and end.
  • The NBER definition focuses on a broad decline across multiple economic measures, not just GDP.
  • Their committee uses real-time and historical data to date recessions with a lag of sometimes several months.
  • The "two-quarter GDP decline" rule is a popular shorthand but not the NBER's official standard.
  • NBER recession declarations are more nuanced than simple GDP rules and often catch turning points earlier.

What is the NBER?​

The National Bureau of Economic Research is a private, nonprofit research organization founded in 1920. It operates independently of government and employs dozens of economists who study business cycles, labor markets, health, education, and public policy. While the NBER conducts research on many topics, one of its most visible roles is dating the peaks and troughs of the U.S. business cycle — that is, officially declaring when recessions and expansions begin and end.

The NBER's Business Cycle Dating Committee consists of leading academic economists. This committee meets periodically to review economic data and make a formal determination: has the economy entered a recession, or is it still expanding? Their word carries enormous weight. Major news outlets, policy institutions, and financial markets treat NBER recession dates as the authoritative record.

Why does a private research body have this unofficial-but-official role? Partly because it predates the modern federal statistical agencies, partly because its independence lends credibility, and partly because economists built a consensus decades ago that the NBER's definition was more scientifically sound than any single metric like GDP. Government agencies like the Bureau of Economic Analysis (BEA) report GDP, but they don't declare recessions — the NBER does.

The NBER's official recession definition​

The NBER defines a recession as follows:

A recession is a significant decline in economic activity spread across the economy, lasting more than a few months, normally visible in real GDP, real income, employment, industrial production, and wholesale-retail sales.

Notice what this definition requires. A recession is not just a one-quarter or even one-year decline in GDP. It is a significant decline, spread across the economy, lasting more than a few months, and visible in multiple indicators.

Let's break down each element:

Significant decline: The drop must be substantial, not marginal. A 0.1% quarterly decline does not trigger recession status; a 2% or 3% decline does.

Spread across the economy: The downturn cannot be confined to one sector or region. It must affect broad swaths of economic activity. A decline in oil prices might hurt energy companies badly but leave manufacturing and services intact; that's not a recession by NBER standards.

Lasting more than a few months: Recessions are not one-off quarterly blips. They typically persist for at least six months (two quarters). The NBER explicitly acknowledges that very short, sharp declines (V-shaped downturns that reverse within a quarter or two) might not qualify.

Visible in multiple indicators: Rather than relying solely on GDP, the NBER looks at employment, industrial production, real income, and sales. If GDP falls but employment and incomes hold steady, the committee is more skeptical about calling it a recession. If multiple measures fall together, the case is clear.

This multi-indicator approach reflects a deep truth about how economies work. GDP measures the total value of goods and services produced, but it can be distorted by inventory changes, one-time shifts, or data revisions. Employment is harder to game; you can't really fake job losses or gains. Industrial production is a direct measure of factories running at full capacity or idling. When all of these measures fall simultaneously, the NBER committee is confident a real recession has occurred.

How the NBER's definition differs from the "two-quarter rule"​

Many news stories and popular definitions of recession invoke the "two-quarter rule": a recession is two consecutive quarters of negative GDP growth. This rule is simple, mechanical, and easy to apply. As soon as quarterly GDP turns negative twice in a row, recession declared.

The NBER's definition is much more flexible and judgement-based. Consider a hypothetical scenario: the economy shrinks in Q1 by 0.5%, then rebounds in Q2 by 1.2%, but employment continues falling and people are still reluctant to spend. By the two-quarter rule, there was no recession (one down, one up). By the NBER approach, there might be one, because the employment picture is weak and the overall tone of economic activity is negative despite the GDP rebound.

Conversely, consider another scenario: GDP drops 1% in Q1, then stays flat in Q2, but employment is robust and incomes are rising. The two-quarter rule does not apply (only one negative quarter), but the NBER might still investigate carefully before declaring a recession.

The NBER's multi-indicator, judgement-based approach often makes the committee the final arbiter in borderline cases. In 2020, for instance, the COVID-19 recession was extraordinarily sharp but also extraordinarily short. GDP fell by about 3% in Q2, then bounced back by 7% in Q3 — the largest swings in modern history. By a pure two-quarter reading, there was one down quarter (Q2) followed by recovery, so no clear recession. But employment collapsed and the economic pain was acute. The NBER dated this as a recession lasting from February to April 2020, just two months, much shorter than the two-quarter rule would have implied.

This flexibility is one reason markets and policymakers respect the NBER's judgment: they understand that a mechanical rule misses the real economic experience.

The NBER's dating process​

The NBER committee does not declare recessions in real time. Instead, it waits for data to accumulate and then looks backward. When policymakers and markets are hoping for clarity about whether a recession is happening now, the NBER typically remains silent until several months after the fact.

Here's the practical timeline:

Month N: The economy enters a recession, but data is incomplete.

Months N through N+6: Economic indicators (employment, GDP, industrial production) are released regularly, but they are often revised. The committee gathers data.

Months N+6 through N+12: The committee reviews all available data and meets to discuss whether a recession occurred, when it started, and (if it is over) when it ended.

Month N+12 or later: The NBER issues an official announcement. Sometimes this announcement is made months after the recession has already ended and the economy has begun recovering.

For example, in 2020, the COVID recession began in February and ended in April (per NBER). But the NBER did not formally announce this until June 8, 2020 — four months after the fact.

This lag is frustrating for real-time decision-makers. Policymakers cannot wait 6–12 months for official confirmation. They must act on preliminary data, forecasts, and their own judgment about whether a recession is underway. That's why the two-quarter rule and other real-time measures exist: they offer a working definition for the here and now, even if they sometimes differ from the NBER's eventual retrospective judgment.

The components of the NBER's dating​

When the NBER committee meets, it examines four to five main economic indicators:

Real Gross Domestic Product (GDP): The total value of goods and services produced, adjusted for inflation. The committee looks at both quarterly growth rates and the level of output relative to trend.

Real Personal Income excluding Transfers (RPI-X): This measures household income from wages, business profits, and capital gains, excluding government transfer payments like unemployment benefits or social security. It reflects whether workers are actually earning more or less.

Employment: The total number of jobs in the economy. The committee tracks the payroll employment level and looks for sharp drops. A loss of millions of jobs is a clear recession signal.

Industrial Production: An index of output at factories, mines, and utilities. This measure is especially sensitive to recessions because firms typically cut production capacity before laying off workers (they reduce hours and shift to part-time first).

Wholesale and Retail Sales: The total sales of goods by wholesalers and retailers. A drop in sales signals that consumers and businesses are pulling back on spending.

The NBER publishes a business cycle chronology — a historical table of all officially dated recessions and expansions since 1854. This table is a public record and is updated periodically. Economists and policymakers reference this table constantly when discussing historical business cycles.

Why the NBER's judgment matters​

Why not just let a computer apply a simple rule? The answer is that real economic life is messier than any single formula. Consider a few cases:

Supply shocks: An oil embargo or a pandemic might cause a sharp, temporary drop in production and employment, but the underlying productive capacity of the economy is intact. Consumers and workers know the shutdown is temporary. The NBER might record a very brief recession (as it did for COVID), but economists recognize that the fundamentals differ from a typical demand-driven recession.

Measurement issues: GDP is revised repeatedly. A first estimate might show a 0.5% decline, but a revised estimate (released months later) might show 0.2% growth instead. The NBER waits for these revisions to stabilize before making a call.

Regional variation: Sometimes one region of the country is in recession while another is booming. The NBER looks for broad-based declines, not isolated sectoral downturns.

The role of policy: If policymakers respond quickly to early warning signs of a downturn with aggressive fiscal or monetary stimulus, they might prevent a full recession or cut short a mild one. The NBER committee considers whether the underlying economic forces called for a recession even if policy intervened.

By exercising judgment, the NBER committee makes a call that is informed by statistical rigor but not enslaved to a mechanical rule. This is valuable precisely because it forces the committee to think deeply about what a recession really is: not a statistical artifact, but a real period of broad-based economic weakness that causes hardship for workers and firms.

Real-world examples​

The 2008 recession: The Great Recession began in December 2007 and ended in June 2009, as documented by the NBER Business Cycle Dating Committee. The NBER dated this recession based on a sustained collapse in employment (9.3 million jobs lost over two years), a sharp decline in industrial production, and negative real income growth. Employment data from the Bureau of Labor Statistics confirmed the severity of job losses. GDP fell, but the employment picture was so severe that there was no doubt about recession status.

The 2001 recession: The dot-com bubble burst in 2000–2001. GDP actually contracted only mildly (never two consecutive negative quarters), but employment fell, industrial production declined sharply, and the tone of economic activity was clearly negative. The NBER dated a recession from March to November 2001 based on this multi-indicator view, even though GDP was not the focal point.

The 2020 recession: The COVID pandemic caused the sharpest, shortest recession in U.S. history. GDP fell 3.1% in Q2, then rebounded 7.1% in Q3. Employment collapsed from 130 million to 127 million jobs in March–April, then recovered. The NBER classified this as a recession lasting February–April 2020, a mere two months, based on the severity of the employment decline and the speed of the downturn.

Common mistakes​

Mistake 1: Assuming that "the NBER has not declared a recession, so there is no recession." The NBER works with a lag. By the time they announce a recession, it may already be over and the economy recovering. Real-time forecasters and policymakers must make judgments before the NBER's blessing.

Mistake 2: Confusing the NBER definition with the two-quarter rule. They are related but distinct. The NBER is more flexible and judgment-based; the two-quarter rule is mechanical. Both have merit, but they sometimes reach different conclusions.

Mistake 3: Treating all NBER recessions as equivalent in severity. The 2020 COVID recession lasted two months; the Great Recession lasted 18 months. Both are recessions by the NBER's definition, but their economic and human impact was vastly different. Severity matters for policy response.

Mistake 4: Ignoring employment data in favor of GDP. The NBER puts heavy weight on employment because job losses are real and visible to households. A recession that shows up in employment is more serious than one that shows up only in GDP noise.

Mistake 5: Assuming the NBER will declare a recession as soon as it happens. There is always a lag. Market participants and policymakers often know a recession is underway before the NBER formally announces it, based on employment data and other real-time indicators.

FAQ​

Q: Who sits on the NBER's Business Cycle Dating Committee? A: The committee is composed of leading academic economists, usually 5–7 members, chosen for their expertise in business cycles. Members rotate and are affiliated with universities and research institutions. The chairman is elected from among the members. Members are not government appointees and serve without compensation.

Q: How often does the committee meet? A: The committee typically meets a few times per year, both in regular sessions and as needed when economic conditions warrant a review. After a clear turning point (like a sharp drop in employment or a change in direction of GDP), the committee may accelerate its timeline.

Q: Can the NBER reverse its decision about when a recession started or ended? A: Yes, but rarely. The NBER has occasionally revised the dating of past recessions when new data became available or methodology was refined. These revisions are infrequent and well-publicized. The official chronology is treated as the final word, but historians occasionally debate whether particular downturns should have been classified differently.

Q: How does the NBER date expansions? A: An expansion is the period between the end of one recession and the beginning of the next. The NBER simply dates the trough (the lowest point of the recession) and the next peak (the highest point before the next recession). Everything in between is an expansion. The longest expansion in U.S. history, by NBER dating, lasted 128 months (June 2009 to February 2020).

Q: What's the relationship between NBER recession dates and fiscal policy decisions? A: Some automatic stabilizers in U.S. tax and transfer law are triggered by "recession" as defined by the NBER. For example, extended unemployment insurance benefits have historically been tied to NBER recession dates. So the NBER's judgment has concrete policy consequences.

Q: Can a country be in a recession by the NBER standard but not by others? A: Yes. Other countries have their own recession-dating bodies or definitions. For instance, the European Commission has its own rules for dating euro-area recessions. The NBER's definition applies specifically to the U.S. economy, though similar standards are used by other nations.

Q: What happens if real GDP data is revised drastically after the NBER dates a recession? A: The NBER looks at all available evidence, not just GDP. If employment, income, and production all confirm a downturn, the NBER is confident in its dating even if GDP is later revised. Conversely, if only GDP moved, the committee is more cautious. Historical revisions to GDP do happen, but they have not typically reversed the NBER's broad recession calls.

Summary​

The National Bureau of Economic Research (NBER) is the official arbiter of U.S. recessions. Unlike the mechanical "two-quarter GDP decline" rule, the NBER defines a recession as a broad, significant decline in economic activity visible across multiple measures—employment, industrial production, real income, and sales—lasting more than a few months. This multi-indicator judgment-based approach is more nuanced than simple rules, and the NBER committee's declarations are treated as the gold standard by economists, policymakers, and markets. However, the NBER works with a data lag, often announcing recessions months after they have begun or ended. Understanding the NBER's approach is essential for anyone trying to make sense of when recessions truly occur and what they mean for the broader economy.

Next​

→ The two-quarter technical recession rule