Who decides when a recession starts, and why it takes so long
The 'two quarters of falling GDP' rule is not the official US definition. Here is how the NBER dates recessions, which indicators it watches and why its calls come months late.
Every time growth slows, a familiar argument breaks out over whether the economy is in a recession. Part of the confusion is that the most quoted definition, two consecutive quarters of falling gross domestic product, is not the one the United States officially uses. The official dates come from a small committee of academic economists, and they usually arrive well after the fact.
The two-quarter rule of thumb
Two straight quarters of negative real GDP growth is a simple test, and many other countries' commentators use it. It is easy to check because GDP is published quarterly. But it has real problems. GDP is revised, sometimes heavily, so a quarter that first looked negative can later turn positive. GDP also measures only output, and a single quarter can be distorted by swings in inventories or trade that say little about whether households are losing jobs and income.
In the first half of 2022, initial estimates showed US real GDP falling for two quarters in a row. At the same time, employers were adding hundreds of thousands of jobs a month and unemployment was near historic lows. The official arbiter did not declare a recession for that period, and later revisions to GDP changed the picture of those quarters.
The NBER Business Cycle Dating Committee
In the US, recession dates are set by the Business Cycle Dating Committee of the National Bureau of Economic Research, a private, nonprofit research organization. The committee's members are economists, mostly from universities. Its chronology of peaks and troughs goes back to the 1850s and is the standard reference used by government agencies, researchers and the press.
The committee defines a recession as a significant decline in economic activity that is spread across the economy and lasts more than a few months. It looks at three criteria, depth, diffusion and duration, and treats them as somewhat interchangeable: an extremely deep but very short downturn can still qualify, as the early 2020 pandemic recession did.
What the committee watches
Instead of relying on one number, the committee looks at a set of monthly indicators. Its published materials highlight:
- Real personal income less transfer payments, which shows earned income adjusted for inflation.
- Nonfarm payroll employment from the Bureau of Labor Statistics establishment survey.
- Employment as measured by the household survey.
- Real personal consumption expenditures.
- Wholesale and retail sales adjusted for price changes.
- Industrial production from the Federal Reserve.
It also considers quarterly measures, including real GDP and real gross domestic income, which estimates the economy's size from the income side. The two should be equal in principle but often differ in practice, and when they disagree the committee looks at both.
Peaks and troughs
The committee does not declare that a recession is happening. It dates turning points. A peak is the month when activity reached its high before falling, and the recession begins the month after. A trough is the month when the decline bottomed out, and the expansion begins after that. The period from peak to trough is the recession.
Why the calls are so late
The committee waits until it is confident, which means waiting for data to be revised and for a turning point to be clear in hindsight. Announcements have typically come many months after the peak. The December 2007 peak that began the Great Recession was announced in December 2008, a full year later. The February 2020 peak was announced in June 2020, unusually quickly because the collapse in activity was so sudden and so large.
This delay frustrates people who want a real-time answer, but it is deliberate. The committee's purpose is an accurate historical record, not a forecast or a warning. Its dates are used to study how economies behave across cycles, and getting them wrong would distort that work.
Real-time signals people use instead
Because the official call comes late, analysts track indicators that have tended to move early.
- The Sahm rule, named for economist Claudia Sahm, signals the start of a recession when the three-month average of the national unemployment rate rises half a percentage point or more above its low over the previous 12 months. The Federal Reserve Bank of St. Louis publishes it on its FRED database. It is a statistical regularity, not a law, and its creator has noted it can give false signals when labor supply grows quickly.
- The yield curve, particularly when short-term Treasury yields rise above long-term yields, has preceded many past recessions, though with long and variable lags.
- Initial jobless claims, released weekly by the Labor Department, can show layoffs rising before the monthly jobs report does.
- Purchasing managers' surveys and business sentiment surveys give early reads on orders and hiring plans.
None of these is perfect, and each has given false alarms. They are useful because they are timely, not because they are definitive.
Why the label matters, and why it does not
A recession declaration does not trigger automatic policy actions in the US. The Federal Reserve and Congress respond to conditions as they see them, not to the NBER's calendar. For households, the practical signs of a downturn, such as rising layoffs, fewer job openings, slower wage growth and tighter credit, matter more than the label.
Still, the word carries weight in politics and in markets, which is why debates over it can be heated. Knowing who makes the call, what they look at and why they wait helps separate a real turning point from a headline.
This article is general information, not investment advice.