The Sahm Rule Explained: Real-Time Recession Signal Hidden in Jobs Data
The Sahm Rule triggered before every U.S. recession since 1970 with zero false positives. Here's how it works, why the 2024 trigger was different, and how to monitor it with the right datasets.
Most recession indicators have a credibility problem. The yield curve inverts, pundits panic, and then... nothing happens for 18 months. By the time the recession actually arrives, everyone has moved on to the next narrative. The unemployment rate, meanwhile, only confirms what the economy already knows: it rises *after* the damage is done.
The Sahm Rule is different. Developed by former Federal Reserve economist Claudia Sahm in 2019, this real-time recession indicator was designed to solve a specific policy problem: how do you detect a recession in real time, fast enough to trigger automatic fiscal stimulus? The answer turned out to be elegant, mechanical, and historically flawless -- until 2024 complicated the picture.
In this article, we break down exactly how the Sahm Rule works, why its economic mechanism is so powerful, what the 2024 trigger revealed about its limitations, and how to build a labor market deterioration dashboard that goes beyond a single indicator.
What Is the Sahm Rule? Definition and Formula
The Sahm Rule is a recession indicator that triggers when the three-month moving average of the national unemployment rate rises 0.50 percentage points or more above its low point during the previous 12 months. Created by economist Claudia Sahm while at the Federal Reserve, the rule was originally proposed as an automatic trigger for direct stimulus payments to households at the onset of a recession.
That's it. One formula. One data series. One threshold.
How to Calculate the Sahm Rule Indicator
- Take the U.S. unemployment rate (seasonally adjusted, monthly)
- Compute the 3-month moving average
- Find the lowest 3-month moving average in the prior 12 months
- Subtract the low from the current value
- If the result is >= 0.50 percentage points, the Sahm Rule triggers a recession signal
FRED Series: SAHMREALTIME
FRED publishes the Sahm Rule indicator directly as a pre-calculated series, updated monthly. The real-time version uses the unemployment rate as initially reported -- before any BLS revisions -- which is critical for capturing the signal as policymakers and analysts would actually experience it.
Why Is the Threshold 0.50 Percentage Points?
Claudia Sahm calibrated the threshold by backtesting against every post-war U.S. recession. The 0.50pp threshold was the sweet spot that caught every recession early while producing zero false positives from 1970 through 2019. A lower threshold (0.30pp) would have generated false alarms. A higher one (0.70pp) would have been too late.
Sahm Rule Historical Accuracy: 11 for 11 Since 1970
The Sahm Rule's track record is what separates it from most recession indicators. Here is how it performed across every U.S. recession since 1970:
| Recession | Sahm Rule Trigger Date | NBER Recession Start | Lead Time | Peak Sahm Value |
|---|---|---|---|---|
| 1970 (Dec 1969 - Nov 1970) | Jan 1970 | Dec 1969 | 1 month after start | 0.90pp |
| 1973-75 | Sep 1974 | Nov 1973 | During recession | 1.70pp |
| 1980 | Mar 1980 | Jan 1980 | 2 months after start | 1.30pp |
| 1981-82 | Oct 1981 | Jul 1981 | 3 months after start | 2.20pp |
| 1990-91 | Sep 1990 | Jul 1990 | 2 months after start | 0.70pp |
| 2001 | Jun 2001 | Mar 2001 | 3 months after start | 0.80pp |
| 2007-09 | Feb 2008 | Dec 2007 | 2 months after start | 4.00pp |
| 2020 | Apr 2020 | Feb 2020 | 2 months after start | 11.10pp |
An important nuance: the Sahm Rule is not a *leading* indicator in the traditional sense. It doesn't warn you six months early, the way the yield curve sometimes does. Instead, it confirms that a recession has likely begun -- within one to three months of the start. That speed matters enormously for fiscal policy. The NBER's official recession dating committee typically doesn't announce a recession until 6-12 months after it begins. The Sahm Rule closes that gap.
False positives from 1970 to 2019: zero.
The Economic Mechanism: Why Rising Unemployment Accelerates Into Recession
The Sahm Rule works because of a well-documented feedback loop in labor markets that economists call the unemployment multiplier effect.
Here is the chain of causation:
- Initial shock: Some sector of the economy weakens -- housing, manufacturing, exports, technology
- Layoffs begin: Firms reduce headcount, pushing the unemployment rate up
- Spending contracts: Newly unemployed workers cut consumption immediately -- research shows spending drops 10-15% within the first month of job loss
- Secondary layoffs: Businesses in consumer-facing sectors (retail, restaurants, services) see demand fall, triggering another round of cuts
- Credit tightens: Banks see rising delinquencies and pull back lending, especially to small businesses
- Confidence collapses: Employed workers start saving more ("precautionary saving"), further reducing demand
This is why a 0.50pp rise in unemployment is not just noise. It signals that the feedback loop has activated. Small increases in unemployment tend to become large increases because the mechanism is self-reinforcing. The economy doesn't slowly drift into recession -- it accelerates into one.
FRED Series: UNRATE (Civilian Unemployment Rate, seasonally adjusted)
July 2024 Sahm Rule Trigger: Why It Wasn't a Recession
In July 2024, the Sahm Rule triggered for the first time since the pandemic. The 3-month moving average of unemployment reached 4.13%, which was 0.53pp above its 12-month low of 3.60%. Markets reacted sharply -- the S&P 500 fell over 3% on August 2nd, and recession fears dominated financial media for weeks.
But as of April 2026, the NBER has not declared a recession for 2024 or 2025. GDP growth remained positive. Consumer spending held up. The labor market stabilized.
What Caused the 2024 False Signal?
Several structural factors distorted the signal:
1. Immigration-driven labor supply expansion
The U.S. experienced a significant surge in immigration during 2023-2024. This expanded the labor force rapidly, pushing the unemployment rate up even as the number of employed workers continued to grow. The rise in unemployment was partially a *supply* story, not just a *demand* story -- and the Sahm Rule doesn't distinguish between the two.
2. Post-pandemic labor market normalization
The 3.4% unemployment rate of early 2023 was historically extreme. Some reversion toward 4.0% was expected as pandemic-era labor hoarding unwound and the economy rebalanced.
3. Sectoral rotation, not broad-based weakness
Job losses concentrated in a few sectors (technology, financial services) while healthcare, government, and leisure/hospitality continued adding jobs. The Sahm Rule aggregates to a single national number and cannot capture this compositional nuance.
What the 2024 Sahm Rule Trigger Teaches Analysts
The 2024 episode doesn't invalidate the Sahm Rule -- but it reveals an important limitation. The rule was calibrated during a period of relatively stable labor force growth. When immigration or demographic shifts cause rapid labor supply changes, the unemployment rate can rise for non-recessionary reasons. Any serious labor market monitoring framework needs to pair the Sahm Rule with supply-side indicators.
How to Build a Labor Market Recession Dashboard
The Sahm Rule is a powerful starting point, but relying on a single indicator is always risky. Here are the datasets that professional macro analysts use alongside it to confirm -- or rule out -- a genuine deterioration signal.
Tier 1: Real-Time Recession Speed Indicators
These move fastest and provide the earliest signal of labor market stress.
| Indicator | FRED Series | Frequency | What It Tells You | IQ Score Range |
|---|---|---|---|---|
| Sahm Rule (Real-Time) | SAHMREALTIME | Monthly | Recession threshold breach | 90+ |
| Initial Jobless Claims | ICSA | Weekly | New layoffs in real time | 90+ |
| Continued Claims | CCSA | Weekly | Duration of unemployment spells | 85-90 |
| Insured Unemployment Rate | IURSA | Weekly | Claims as % of covered employment | 80-85 |
Initial claims (ICSA) deserve special attention. They are the single highest-frequency labor market indicator available -- published weekly, with only a one-week lag. When initial claims breach 250,000 and trend higher for four or more consecutive weeks, labor market trouble is almost certainly underway.
During the 2007-2009 recession, initial claims crossed 300,000 in January 2008 -- the same month the Sahm Rule triggered. During the 2001 recession, claims began rising above 350,000 in March 2001. The two indicators tend to confirm each other.
FRED Series: ICSA
Tier 2: Labor Market Breadth and JOLTS Indicators
These tell you whether weakness is broad-based or concentrated in a few sectors.
| Indicator | FRED Series | Frequency | What It Tells You |
|---|---|---|---|
| U-6 Underemployment Rate | U6RATE | Monthly | Broadest measure of labor slack |
| Job Openings (JOLTS) | JTSJOL | Monthly | Employer demand for workers |
| Quits Rate (JOLTS) | JTSQUR | Monthly | Worker confidence in finding new jobs |
| Hires Rate (JOLTS) | JTSHIR | Monthly | New employment activity |
| Average Weekly Hours | AWHMAN | Monthly | Hours worked in manufacturing |
The JOLTS data is where the real analytical edge lives. Job openings peaked at 12.2 million in March 2022 and have since declined significantly. But the *quits rate* is arguably more telling. When workers stop voluntarily quitting, it means they no longer believe they can find something better. The quits rate fell from 3.0% in April 2022 to 2.1% by late 2024 -- a level consistent with pre-pandemic norms. That decline signaled cooling, not crisis.
Average weekly hours (AWHMAN) in manufacturing is another underappreciated leading indicator. Employers cut hours before they cut headcount. A sustained decline below 40.0 hours per week historically precedes significant layoff waves by 3-6 months.
FRED Series: AWHMAN
Tier 3: Confirming Macro Indicators for Recession Risk
These provide the broader economic context that helps you interpret the labor data.
| Indicator | FRED Series | What It Confirms |
|---|---|---|
| Real Personal Income (ex transfers) | W875RX1 | Whether consumer spending power is actually falling |
| Industrial Production | INDPRO | Whether the goods economy is contracting |
| Real Retail Sales | RSXFS (deflated) | Whether consumer demand is declining |
| Temporary Employment | TEMPHELPS | Firms cut temps before permanent staff |
Temporary help services (TEMPHELPS) is a particularly clever leading indicator. When firms start cutting temporary and contract workers, it signals that they expect demand to weaken but haven't yet committed to permanent layoffs. Temp employment peaked in March 2022 at approximately 3.2 million and has been declining steadily -- shedding roughly 500,000 jobs through mid-2025. In past cycles, sustained temp declines of this magnitude preceded broader layoffs by 6-12 months.
FRED Series: TEMPHELPS
Sahm Rule vs. FRED: The Workflow Problem
Tracking the Sahm Rule alone is easy on FRED -- search "SAHMREALTIME," download the series, done. But building a comprehensive labor market deterioration dashboard across all three tiers requires a fundamentally different workflow.
On FRED:
- Search for SAHMREALTIME -- find the series among 800K+ results
- Search for ICSA -- navigate to the right weekly series (not the 4-week average, not the advance number)
- Search for JOLTS data -- pick from dozens of sub-series (openings, quits, hires, separations, by industry)
- Repeat for U6RATE, AWHMAN, TEMPHELPS, W875RX1, INDPRO
- Download each series individually as CSV
- Align the frequencies (weekly claims vs. monthly unemployment vs. monthly JOLTS)
- Build your own charts and comparison views
- Set up manual calendar reminders for release dates
- Repeat every month
Total time: 45-60 minutes per update cycle, assuming you already know every series code.
On DataSetIQ:
- Search "Sahm Rule recession" -- SmartFind returns SAHMREALTIME as the top result with an IQ Score
- Search "labor market deterioration" -- get a curated set of related series ranked by relevance and data quality
- Use multi-series comparison to overlay initial claims, unemployment, and JOLTS data on one chart
- Set up Live Data Pulse alerts on SAHMREALTIME and ICSA so you're notified the moment new data drops
- Generate an Advanced Research Brief on any series for instant AI analysis of the latest reading
Total time: under 5 minutes, with automated monitoring going forward.
| Task | FRED Workflow | DataSetIQ Workflow |
|---|---|---|
| Find SAHMREALTIME | Search 800K+ series, filter manually | SmartFind returns top result instantly |
| Compare 5+ labor indicators | Download CSVs, align in Excel | Multi-series comparison, one click |
| Get alerted on new data | Manual calendar reminders | Live Data Pulse, automatic |
| Analyze latest reading | Read the raw number yourself | Advanced Research Brief with AI context |
| Total setup time | 45-60 minutes per cycle | Under 5 minutes, then automated |
Search labor market recession signals on DataSetIQ
How to Interpret the Sahm Rule in 2026
As of the latest BLS data, the Sahm Rule indicator reads approximately 0.20pp -- well below the 0.50pp threshold. The unemployment rate has stabilized in the 4.0-4.2% range, initial claims remain below 230,000, and JOLTS openings have leveled off rather than continuing to decline.
Current Sahm Rule Dashboard Readings (April 2026)
| Indicator | Latest Reading | Signal |
|---|---|---|
| Sahm Rule | ~0.20pp | No trigger |
| Initial Claims (4-wk avg) | ~225,000 | Normal range |
| Continued Claims | ~1.85M | Slightly elevated |
| U-6 Underemployment | ~7.5% | Stable |
| JOLTS Openings | ~7.5M | Normalizing |
| Quits Rate | ~2.1% | Pre-pandemic normal |
| Temp Employment | Declining | Watch zone |
| Avg Weekly Hours (Mfg) | ~40.1 hours | Stable |
The picture is one of a labor market that has cooled from overheated (2022) to balanced (2026), without tipping into outright deterioration. The Sahm Rule is not close to triggering. However, the continued decline in temporary employment and the persistence of elevated continued claims are worth monitoring. If temp employment losses accelerate and initial claims begin trending above 250,000, the Sahm Rule could move toward its threshold within 3-6 months.
Key Takeaways
- The Sahm Rule triggers when the 3-month average unemployment rate rises 0.50pp above its 12-month low -- it confirmed every U.S. recession from 1970 to 2020 with zero false positives during that period
- The 2024 trigger was driven largely by labor supply expansion (immigration) rather than demand-side weakness, highlighting that structural shifts can distort the signal
- Initial jobless claims (ICSA) are the highest-frequency confirmation indicator -- weekly data with a one-week lag, compared to the Sahm Rule's monthly frequency
- JOLTS quits rate and temporary employment are the most underappreciated leading indicators for labor market turning points
- A single indicator is never enough -- pair the Sahm Rule with breadth measures (U-6, JOLTS) and leading indicators (temp employment, average hours) for a complete picture
- As of April 2026, the Sahm Rule reads well below its trigger threshold, but temp employment declines warrant continued monitoring
Frequently Asked Questions About the Sahm Rule
What is the Sahm Rule and how does it work?
The Sahm Rule is a recession indicator created by economist Claudia Sahm. It triggers when the three-month moving average of the U.S. unemployment rate rises 0.50 percentage points or more above its lowest point in the prior 12 months. It detected every U.S. recession from 1970 to 2020 with zero false positives. You can track it via FRED series SAHMREALTIME.
Has the Sahm Rule ever been wrong?
From 1970 through 2019, the Sahm Rule produced zero false positives. However, it triggered in July 2024 without an official recession following, likely because immigration-driven labor supply growth -- rather than demand-side weakness -- pushed unemployment higher. This was its first false signal in over 50 years.
What is the current Sahm Rule reading in 2026?
As of April 2026, the Sahm Rule indicator reads approximately 0.20 percentage points, well below the 0.50pp recession trigger threshold. The U.S. unemployment rate has stabilized in the 4.0-4.2% range, and the labor market shows cooling without deterioration.
How is the Sahm Rule different from the yield curve as a recession predictor?
The yield curve (10Y-3M spread) is a *leading* indicator that can signal recession risk 6-12 months in advance but produces occasional false positives. The Sahm Rule is a *coincident* indicator that confirms a recession has likely begun within 1-3 months of the start, with historically zero false positives (prior to 2024). The two complement each other: the yield curve warns, the Sahm Rule confirms.
What datasets should I monitor alongside the Sahm Rule?
Pair the Sahm Rule (SAHMREALTIME) with initial jobless claims (ICSA) for weekly confirmation, JOLTS quits rate (JTSQUR) for worker confidence, temporary employment (TEMPHELPS) as a leading indicator, and U-6 underemployment (U6RATE) for labor market breadth. Search all labor market indicators on DataSetIQ.
Start Monitoring Sahm Rule and Labor Market Signals
Building a real-time labor market deterioration dashboard shouldn't require memorizing 15 FRED series codes and manually downloading CSVs every month. DataSetIQ consolidates all of these indicators -- Sahm Rule, initial claims, JOLTS, temporary employment, and more -- into a single searchable workspace with AI-powered analysis and automated alerts.
Set up Live Data Pulse on SAHMREALTIME and ICSA, and you'll know within hours when the data moves -- not days later when the headlines catch up.
Explore Sahm Rule and labor market datasets on DataSetIQ
No credit card required. 14-day free trial on paid plans.
*This analysis reflects conditions as of the publication date. Labor market data is revised frequently -- always verify current readings before making investment or policy decisions.*
Related Articles

7 Economic Indicators That Actually Predict Recessions (Backtested Across 50 Years)
Most recession indicators sound convincing but fail real-world testing. These seven indicators actually predicted every U.S. recession since 1970 — with clear lead times and minimal false alarms. Learn the exact framework used by macro hedge funds and central banks.
Nowcasting GDP: How Economists Estimate Growth Before the Official Number Drops
GDP data arrives weeks late. Nowcasting models estimate growth in real time using high-frequency indicators. Here's how the framework works, which datasets drive the estimates, and how to build your own tracker.
Complete Guide to Unemployment Rate Data: BLS, OECD & More
2008 took 6 years to recover. 2020 took 18 months. The data tells a fascinating story about two very different economic crises.