The Sahm Rule
Detect recessions in real time through a sustained rise in unemployment
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 5
- Confidence
- 99%
The Sahm Rule converts a widely followed labor statistic into a transparent recession signal. Calculate the three-month moving average of the unemployment rate and compare it with the lowest three-month average recorded during the previous twelve months. A rise of at least half a percentage point signals that a recession has begun. Sahm developed the threshold by testing historical and real-time data, seeking a rule that activated during the opening months of every recession without the false positive produced by the Federal Reserve’s more sensitive rule of thumb. Its simplicity is central to the mechanism: lawmakers can understand it, the public can verify it, and fiscal programs can use it as an automatic trigger.
Origin
Claudia Sahm developed the indicator while designing automatic recession payments for the Recession Ready volume. She tested real-time unemployment data in a large spreadsheet to find a reliable trigger suitable for congressional fiscal policy.
Core principles
- 01Use labor-market harm as the recession signal
- 02Favor a simple public trigger over an opaque model
- 03Test indicators against data available at the time
- 04Minimize false positives when public money is triggered
- 05Act within the first months of a recession
How to run it
- 1
Build the real-time dataset
Collect unemployment releases as they were available at each historical point, rather than relying solely on subsequently revised figures.
Pro tip Preserve each release vintage so the test reflects what policymakers actually knew.
Watch out Revised data can make a historical signal appear cleaner than it was in real time.
- 2
Smooth monthly volatility
Calculate the three-month moving average of the national unemployment rate.
Pro tip Use the same calculation consistently across the full historical series.
Watch out A single monthly increase can be noise and should not independently trigger action.
- 3
Establish the recent low
Identify the lowest three-month average unemployment rate during the preceding twelve months.
Pro tip Use a rolling twelve-month window so the baseline adapts over time.
Watch out Do not compare the current rate with an arbitrary long-term average.
- 4
Measure labor-market deterioration
Subtract the recent low from the current three-month average. Treat an increase of at least half a percentage point as the recession signal.
Pro tip Publish the inputs alongside the result so others can verify the trigger.
Watch out A more sensitive threshold can activate outside recessions.
- 5
Connect the signal to action
Use the trigger to activate preauthorized fiscal relief, then monitor unemployment to determine when support should phase down.
Pro tip Define both activation and deactivation rules before a crisis begins.
Watch out Leaving implementation to fresh political negotiations can delay relief after the signal fires.
In the wild
A government preauthorizes direct household payments tied to the Sahm Rule. When the three-month unemployment average rises half a percentage point above its prior twelve-month low, the payment system activates without waiting for an official recession declaration.
→ Households receive support during the recession’s opening months, when rapid relief has the greatest stabilizing value.
An analyst observes several weak labor reports but calculates that the smoothed unemployment increase remains below the threshold. The analyst reports rising risk without declaring that the recession trigger has fired.
→ The rule separates concerning labor-market noise from the stronger deterioration historically associated with recessions.
Common mistakes
Using revised historical data
Testing only the latest revised series ignores what observers could actually see when decisions had to be made.
Lowering the threshold without testing
A more sensitive rule may provide extra warning but can generate false positives and prematurely activate expensive programs.
Treating the rule as a full forecast
The rule identifies recessionary labor-market deterioration; it does not explain every cause or predict every dimension of the economy.
Is it for you?
Best for
It is best for policymakers, analysts, and institutions that need a transparent real-time recession trigger.
Not ideal for
It is not ideal as a complete economic forecast or as a substitute for broader analysis of unusual labor-market conditions.
From the transcript
“a small increase in the unemployment rate is a bad sign.”
“And so I found a rule that always triggered within the first few months of a recession.”
“You have to have a simple rule if you want to put fiscal policy on autopilot.”
From the episode
At the Money: Forecasting Recessions with Claudia Sahm
Claudia Sahm