What Is the Sahm Rule? It Triggered in 2024, No Recession

The Sahm Rule is a recession warning built from the US Unemployment Rate. It triggers when the three-month average of the rate rises 0.50 percentage points or more above its recent low. We call that gap the reading. In July 2024 the reading was 0.53. No recession has been dated since.

The National Bureau of Economic Research (NBER) dates US recessions. Its page, opened on 29 October 2026, lists February 2020 as the most recent peak. We tested the rule against NBER's dates for the nine US recessions since 1960. The first trigger is the first month the reading rose to 0.50 or more from a lower level. In seven cases it came one to three months after the recession began. In two cases it came before.

The number behind the rule is the US Unemployment Rate: the number of unemployed people as a share of the labour force. Our guide to the Unemployment Rate covers the measure.

The Federal Reserve Bank of St. Louis publishes the readings on FRED (Federal Reserve Economic Data). Our guide to FRED covers the database.

The rule signals a recession's start when the three-month average of the Unemployment Rate rises 0.50 percentage points or more above the lowest three-month average of the previous 12 months.

Claudia Sahm, then an economist at the Federal Reserve Board, proposed the rule in 2019.

FRED holds two versions. The real-time series uses the rates available in each month. The revised series uses today's data, after the Bureau of Labor Statistics (BLS) revises the rate each January.

This article uses the real-time series unless it says otherwise. Our explainer on why economic data gets revised covers the reason.

The Unemployment Rate was 3.9% in May 2024, 4.1% in June and 4.2% in July. The mean of those three is 4.07%. The lowest three-month average in the previous 12 months was 3.57%, for the three months ending July 2023. The gap is 0.50.

Those figures use today's revised data. The reading published at the time was 0.53, so revision matters near the trigger.

Chart 1 shows the two lines behind the gap, from January 2023 to June 2025. The blue line is the three-month average. The grey line is the lowest three-month average of the previous 12 months.

The pale band marks July to September 2024, when the gap stood at 0.50 or more.

The real-time reading was 0.53, 0.57 and 0.50 in July to September 2024, then 0.43 in October. Since September 2024 no reading has reached 0.50.

NBER's Business Cycle Dating Committee picks the months of each peak and trough in US economic activity. A peak is the high point before a recession. A trough is the low point that ends it. A recession starts in the month after a peak and ends in the trough month.

NBER's dating page, opened on 29 October 2026, says the committee waits for sufficient data to avoid major revisions.

The page also names the two measures NBER has weighted most in recent decades. Payroll employment is the number of jobs on company payrolls outside farms. Real personal income less transfers is household income after inflation, leaving out government payments such as benefits. The Sahm Rule uses neither.

Our guide to what a recession is covers the definition.

We tested the real-time series against the nine US recessions that began after December 1959, the first month FRED holds. For each, we took the first trigger. That is the first month the reading rose to 0.50 or more from a lower level, searching from six months before the recession's first month to its last. A one-month reading counts.

Two first triggers lasted one month: October 1969 and March 1974, both exactly 0.50. The next readings at 0.50 or more came in February 1970 and July 1974, seven months after the 1973 recession began. The series starts at 0.77 in December 1959, so the 1960 trigger may reflect where the data begins.

Chart 2 shows the reading since December 1959. It sat below 0.50 in 624 of the 801 months. It climbed past 1.0 in each of the nine recessions. June 2020 reached 9.5, which is off the scale.

In seven of the nine, the first trigger came one to three months after the recession began. In 1970 it came three months before. In 1960 it came five months before.

Chart 3 gives the months for each recession. Orange bars mark a trigger before the recession began.

The table gives each first trigger and the highest reading by the recession's last month. That reading was at least 1.20 in every one, against 0.57 in 2024.

Real-time Sahm Rule readings, in percentage points. The first reading is the first month the reading rose to 0.50 or more from a lower level. Both columns are searched from six months before the recession's first month to its last month. Source: Federal Reserve Bank of St. Louis (FRED); NBER recession dates

The rule reads the Unemployment Rate alone, while NBER weighs jobs and income. From July 2023 to July 2024 the rate rose from 3.5% to 4.2%. Over the same year, payroll employment rose by 1.7 million. Real personal income less transfers rose 2.9%.

Real GDP, the economy's output after inflation, grew in each of the 13 quarters from the second quarter of 2023 to the second quarter of 2026.

Nine recessions is a small sample. We ran six versions of the method.

Four started the search window 0, 3, 6 or 12 months before a recession's first month. Two moved the trigger level to 0.45 and 0.55. In all six, 2024 was a trigger with no recession around it.

Nothing else was outside a recession in all six. A single reading of exactly 0.50 in November 1976 fell outside in five of them but vanished at the 0.55 level.

Our guide to what the Fed does covers the central bank's role.

The Sahm Rule is a good marker that a downturn is under way. It is a weak early warning, because it triggered after the recession began in seven of nine cases. In 2024 it triggered while payroll employment and real personal income less transfers kept rising.

On 29 October 2026 we opened seven FRED series.

Real-time Sahm Rule (SAHMREALTIME) and revised (SAHMCURRENT). NBER's recession indicator (USREC). From the Bureau of Labor Statistics: the Unemployment Rate (UNRATE) and payroll employment (PAYEMS). From the Bureau of Economic Analysis: real personal income less transfers (W875RX1) and real GDP (GDPC1).

Data end in September 2026, except income (August) and GDP (second quarter). The real-time series holds 801 readings and has no value for October 2025.

For more on recession gauges, see our guide to recession indicators. In the nine US recessions since 1960, the Sahm Rule first triggered between five months before and three months after each recession's first month. In 2024 it triggered with no recession dated since.

The Academy is free to read, here on the web. The Long Trend, our free weekly letter, comes by email.

This explainer is part of the YX Insights Academy. Each one takes a single idea and checks it against real data.

The same approach runs through everything else we publish:

Good places to start on the website:

The formula is the three-month average of the US Unemployment Rate minus its lowest three-month average of the previous 12 months. A gap of 0.50 percentage points or more is a trigger. For July 2024, the average was 4.07% against a low of 3.57%, a gap of 0.50 on revised data.

Yes. On real-time data, two triggers had no recession around them: a one-month reading of exactly 0.50 in November 1976 and the readings from July to September 2024. NBER lists no recession since 2020, but it dates recessions months late, so the record can change.

No. The real-time reading was 0.00 for September 2026, the latest month in the data we opened on 29 October 2026. The last month at or above 0.50 was September 2024. The reading changes each month, when the Bureau of Labor Statistics publishes the Unemployment Rate.

No. The rule usually triggers after a recession has begun. In seven of the nine US recessions since 1960, it first triggered one to three months after the recession's first month. In 1960 and 1970 it triggered before, by five and three months. It reads a rising Unemployment Rate, so it follows a downturn more often than it leads one.

DISCLAIMER: This article is intended for educational purposes only. Any information or analysis in this note does not constitute an offer to sell or a solicitation of an offer to buy any securities. Nothing in this note is intended to be investment advice, nor should it be relied upon to make investment decisions. Any opinions, analyses, or probabilities expressed in this note are those of the author as of the note's date of publication and are subject to change without notice.