Skip to main content

Investigating an Alleged "Real-Time Recession Indicator" That Lags Miserably

Let's take a look at the Sahm rule and statements by its author that the economy is not in recession.
  • Author:
  • Publish date:
Recession Indicator by Claudia Sahm, Data From St. Louis Fed, Chart by Mish

Recession Indicator by Claudia Sahm, Data From St. Louis Fed, Chart by Mish

The Claim vs the Reality

"The Sahm rule is an empirical regularity, not a proposition. It says that when the 3-month moving average of the unemployment rate is 0.5 percentage point above its low over prior 12 months, we are in a recession. Since 1970, hits every one of them."

"Currently, the Sahm rule is 0.03, which is 0.47 percentage point below the trigger. We are not in a recession now."

Let's take a look at the data, the claim, and the trigger. 

Real Time Sahm Rule Recession Indicator 1960

Real Time Sahm Rule Recession Indicator 1960

In the above chart and all those that follow, the first number is the recession indicator at the time the recession started and the second number is the first trigger at 0.50 or higher.

Real Time Sahm Rule Recession Indicator 1970

Real Time Sahm Rule Recession Indicator 1970

Real Time Sahm Rule Recession Indicator 1973

Real Time Sahm Rule Recession Indicator 1973

Real Time Sahm Rule Recession Indicator 1980

Real Time Sahm Rule Recession Indicator 1980

Real Time Sahm Rule Recession Indicator 1982

Real Time Sahm Rule Recession Indicator 1981

Real Time Sahm Rule Recession Indicator 1990

Real Time Sahm Rule Recession Indicator 1990

Real Time Sahm Rule Recession Indicator 2022

Real Time Sahm Rule Recession Indicator 2022

Last 9 Recessions Recap

  • 1960-05-01: 0.07!
  • 1970-01-01: 0.30
  • 1973-12-01: 0.03!
  • 1980-02-01: 0.30
  • 1981-08-01: -0.13!
  • 1990-08-01: 0.17
  • 2001-04-01: 0.40
  • 2008-01-01: 0.40
  • 2020-03-01: 0.30

Not Recession Because?

Allegedly we are not in recession now because the indicator is 0.03.

In 1973 the economy was in recession at 0.03. In 1981 the economy was in recession at -0.13. 

In 33% of the recessions, the indicator was 0.17 or less, once negative. 

7 out of the last nine recessions started with the trigger at 0.30 or less. 

Not Real Time

There is nothing about the indicator that is real time. 

I will grant Sahm that when 0.50 is "triggered" the economy is likely in recession. But how useful in practice is that?

Scroll to Continue

RECOMMENDED ARTICLES

By the time the indicator signals recession, everyone but the Fed and White House already is painfully aware

This is like "predicting" rain when the raindrops hit your head. There is no knowledge gained by this indicator. 

Logic Error 

More problematic is Sahm's logic error. Did you catch it?

"Currently, the Sahm rule is 0.03, which is 0.47 percentage point below the trigger. We are not in a recession now."

Mathematically, If A then B does not imply if Not A then Not B.

There are also problems on the back end.

"The Sahm rule is an empirical regularity, not a proposition. It says that when the 3-month moving average of the unemployment rate is 0.5 percentage point above its low over prior 12 months, we are in a recession. Since 1970, hits every one of them."

When did the 1990 recession end by that logic?

Has Recession Started?

I happen to believe so. But I am not the one who decides. The NBER does. And I might be wrong. 

However, the alleged "real time" indicator provides no useful information nor any reason to say that a recession has not started!

Models Don't Think

One can say jobs are not signaling recession. But jobs are a very lagging indicator and so is the 0.50 trigger based off unemployment rates. 

100% of the time a recession started before the "real time" trigger says it did! 

In 1990 and 2022 the recession was half over before the trigger hit. 

Housing Bust and Cyclicals the Recession Key

Existing home sales have fallen six consecutive months and 25.9 percent since January. There is no instance where that has happened and the economy was not in recession. 

In case you missed it, please see Cyclical Components of GDP, the Most Important Chart in Macro

Also see A Big Housing Bust is the Key to Understanding This Recession

If the data follows the path I expect, we will have a third quarter of negative GDP with real final sales falling since May.

Don't Count on GDI Either

For discussion, please see On an Income Basis the Economy is Humming, GDP says No, Which is Believable?

Maybe the NBER says I am right but perhaps I am wrong. Regardless, what good is a "real time" measure that does not predict rain until everyone can see the rain is falling?

Even sillier, the model's creator now says it is not raining when perhaps it is. We will not know until months from now whether it is raining today!

This post originated on MishTalk.Com.

Thanks for Tuning In!

Please Subscribe to MishTalk Email Alerts.

Subscribers get an email alert of each post as they happen. Read the ones you like and you can unsubscribe at any time.

If you have subscribed and do not get email alerts, please check your spam folder.

Mish