What Do Initial and Continued Claims Suggest About Timing of a Recession? – Mish Talk

January 28, 2023

*Initial and continued unemployment claims data from the BLS, months calculation by Mish*

Current data is easily within a timeframe that recession may have started. In half of the last eight recessions, a recession had already started on the current numbers.

Continued Claims Monthly Average

*Continued unemployment claims monthly average from the BLS, chart by Mish*

The Covid pandemic massively distorted initial and continued claims. I will return to that point in a bit.

Continued and Initial Claims Monthly Average

Initial and continued claims in numbers, data from BLS, chart by Mish

Chart Notes

Initial claims bottomed in March and have gone up in a weak fashion since
Continued claims bottomed in May and have risen at a steady pace

Once someone has lost their job, it is taking longer to find another.

Alternate Point of View

Tweet: https://twitter.com/BobEUnlimited/status/1618614575448096768

“Looking at the labor market, we should expect recession to begin when claims are ~418k given the past couple cycles and today’s working age population size. Even a relatively fast deterioration would take nearly a year to get there.”

I have a number of issues with that theory given all the other recession-supporting evidence, but here’s a look in isolation.

Initial and Continued Claims at Recession Start

Comparison to Prior Recessions

The working age population theory goes right out the window given that continued claims were 2.87 million in 1980 but only 2.80 million 28 years later in 2008.
Covid dramatically distorted all the numbers. One month into recession, initial claims were 4,663,250. Continued claims were 17,032,000.

I suggest a Covid snapback in several ways.

First, I think recession starts with a far lower than average initial claims number and a lower continued claims number.

Demographics supports that idea as well. People (baby boomers) are retiring at an unprecedented clip. That alone has kept up the demand for jobs.

Second, because of the above points, I think we will enter and exit the next recession with a far lower than normal rise in unemployment.

We never did fill all the leisure and hospitality jobs from the 2020 recession and we are not going to have people walking away from homes as in the Great Recession.

So don’t expect this recession to look anything like either of the last two.

The Covid recession was short and amazingly steep. Look for this one to be long and very shallow. By long, I mean in or flirting with recession for years as opposed to continuous recession.

Scroll to Continue

Case for Recession Now

In 5 of the last 7 recessions, industrial production gave a lead time of zero to 3 months.

For discussion please see A Better Definition of Money and Lacy Hunt’s Thoughts on When a Recession Will Start

Consumer Spending Hits Brick Wall

Real Personal Consumption Expenditures from BEA, chart by Mish

Earlier today I noted Personal Spending Hits a Solid Brick Wall in December Despite Rise in Income

Brick Wall

Consumers literally hit the brick wall then went into reverse in November and December.
Real PCE fell 0.2 Percent in November and 0.3 percent in December.
Real PCE Goods were negative 0.9 percent in both months.
Real PCE Services rose 0.2 percent in November and was flat in December.

Real PCE Consecutive Month Declines

Real PCE declined 2 consecutive months starting March of 2020 when the Covid recession began
Real PCE declined 2 consecutive months starting November of 2012
Real PCE declined 4 consecutive months starting February of 2009, in the Great Recession
Real PCE declined 4 consecutive months starting September of 2008, in the Great Recession
Real PCE declined 2 consecutive months starting June of 2008, in the Great Recession
Real PCE declined 2 consecutive months starting August of 2005

The series only dates to January of 2001. Of the 6 prior occurrences, four were in recessions.

Cornucopia of Weakness

Industrial production, housing, retail spending, hours worked, money supply, the yield curve, and even employment levels all scream recession.

The only balancing weight I can find is jobs. And there I expect negative revisions because of the discrepancy with employment and QCEW data.