For the past half century, an increase in the unemployment rate of half a point from a new low has been an unfailing indicator that a recession is either underway or very close:
There are no exceptions to this rule. If it doesn't happen, no recession. If it does happen, recession.
This month marks a half point increase from the April low of 3.4%. Just sayin'.
"for the last half century......."
THAT is the crux of the problem right there.
What is different since 50 years ago?
The BIGGEST thing is the boomers are leaving or have left the work force
The birthrate has dropped as well
There are far fewer replacement workers for those who have or are contemplating retiring
We COULD have a recession without high unemployment which would be odd.
Earn as much as possible (Salary depends on the amount of work done. Easy to apply. In this remote work from home position, we you will be responsible for managing social media platforms, handling requests, providing Feedback and messages to customers responsible),
For use. …. https://possiblework37.blogspot.com
Is it just me, or are there two prior "new low"s on the chart? That... doesn't seem like definitive evidence of anything.
I've seen you make this point a couple of times, and every time I wonder - is there a date you'd be willing to pick, where if a recession hadn't happened by then, you'd admit publicly you were wrong?
This!
Obviously, SOME day, there will be a recession. But Kevin has been predicting one for a long time now! At some point, you have to admit you were wrong, don't you?
I love Kevin, but he was sure inflation would be gone by spring 2022.
At least 2. It's also hard to tell exactly how much 0.5% is on the chart, but it sure looks like there are.
Also, some of the recessions break even the rule that's being supposed by Kevin here: there wasn't a "new low" for unemployment for any of the 3 recessions between 1979 and 2001. Likewise, the unemployment rate prior to 2008 wasn't a new low compared to 2001.
It's just a silly, made-up data point. Similar to silly, made-up baseball stats that mean nothing. "The first player in history to go 2-for-4 in 7 games in a row, with alternating strikeouts and walks!" It means nothing.
I don't read Kevin to be claiming that all recessions immediately follow X, merely that X ~always indicates an impending recession.
If X always indicates recessions, his sample size is 2 prior to COVID. And there are reasons to believe that the economy is just structurally different today than in 1979 or 2001.
If he means for there to be a sample size greater than 2, than the "always" doesn't work because there are multiple instances of in-between-recession unemployment low points in there that don't have a recession following them.
Guessing, but I think he means a “local” low, not a new absolute low in unemployment. Like a local minimum…
But I’m not sure if the point still stands: that you only know it’s a rise of .5% after a low, in retrospect? What if unemployment data is noisy and .5% increases happen but then it decreases again (more than .5% down, creating a new local low).
I think it *isn't* that noisy, and the .5% increase really has been a signal that something’s coming, in the past. But I dunno. Also, I don’t think it matters now, because I believe *this time is different* and there won’t be recession. For reasons.
Anyway, as WryCooter (great name) says below, Mr Drum is referring to the SAHM rule that is making the rounds in the news today.
The Sahm rule is that the 3 month average goes up by 0.5%. It has happened early in every recession during the time period, but not elsewhere.
Current level is 0.33% and Claudia Sahm on Twitter called it not imminent.
https://fred.stlouisfed.org/graph/?g=1aXkY
I'm old enough to remember the rule that neither the Cubs nor the Red Sox would ever win a world series, thanks to the curse of the billy goat (Cubs) and the curse of the Bambino (Red Sox). No exceptions. Except, of course, until they actually won.
Throughout history, there are things that never happened until they happened. Or, as John Lennon once said, :There's nothing you can do that can't be done."
But don't we recognize "new lows" only in retrospect? Between January and March 1986, for example, unemployment jumped from 6.7 to 7.2 percent, hovered around that zone for a few months before declining down to 5 percent three years later.
Something similar happened in 1962.
I don't see how that's any sort of infallible indicator.
What is the explanation for the second recession in the early 80’s?
Interest rates increased by the Fed.
I see what looks to be a similar size increase in the 2003-04 area of the graph, yet no corresponding recession.
I'll let XKCD handle this one.
xkcd.com/1122/
Exactly where my thoughts went. My second favorite after the statistical significance/jellybeans one: https://xkcd.com/882/
Although unemployment was goosed this month by the UAW strike.
I'm betting this time is different. Right now the Fed has tons of room to cut interest rates. They can really turn on the spigot if needed. And there's (probably) not any financial crisis on the horizon that's going to push us into some structural recession.
Commercial real estate is pretty shaky right now.
They can really turn on the spigot if needed.
I'm not sure how valid this is. In the first place, it's far from clear when, exactly, Jay Powell's concerns about inflation will have sufficiently subsided. I think it's perfectly plausible if we do get a recession, Fed easing will come slowly and begrudgingly in the beginning. Also, while no doubt it's true the Fed now has robust capacity to engage in easing, there's unlikely to be a guarantee the effects of said easing kick in quickly enough to save Joe Biden's bacon (which for me is really what counts).
But sure, a briskly recovering economy may greet President Trump in 2025!
Hey Kev,
You have misintepreted the Sahm Rule which requires a 0.50% rise over a THREE MONTH MOVING AVERAGE.
Here's Claudia's explanation: https://open.substack.com/pub/stayathomemacro/p/quick-update-on-the-sahm-rule?r=8dgpo&utm_campaign=post&utm_medium=web
There's one other thing that KD is not taking into accout
As the Late Boomers retire and then start to die off in substantial numbers we are gonna see the largest transfer of wealth in history
That cannot be dismissed
It will have a HUGE impact on the economy
Are we, though? And will it?
cmayo
The wealth transfer is just staring and MUCH of that wealth is in real estate.
It will take time to unwind and get distributed to families
But it WILL happen and the affect will be huge
I don't think so, I think there's going to be a ton of theft by the medical establishment for care of aging baby boomers in their final years also healthy boomers are living longer into their old age so they are spending more of their net worth before they die.
"Late Boomers" are now in their early 60s. The effect you mention will no doubt have a big impact, but it's not going to be felt in a major way for a while (even "peak" boomers are only 70). Certainly this isn't something that's going to palpably affect the business cycle over the next couple of years.
The past 50 years is doing some lifting here. In October of 1962 the unemployment rate was 5.4%. In February of 1963 the unemployment rate was 5.9%. The next recession happened in ... 1970. Kevin is close to stating the Sahm rule which "posits the start of a recession when the three-month moving average of the unemployment rate rises by a half-percentage point or more relative to its low during the previous 12 months." Fred has a graph of the Sahm rule, which has been gradually rising. But is not yet at the level that would predict a recession.
Josh
Also consider this
The birthrate in 1957 was 122 children per 1000 women
In 2022 it was down to 56.1 or more than half
Immigration has not made up the difference so TODAY we just don't have as many people to replace anyone who is leaving as we used to
Out military is also struggling to meet recruiting goals in every branch of the services.
I suspect a slowdown is coming within the next 50 years but it may be we have to get used to a new norm rather than rely on what happened 50 years ago !!
I suspect a slowdown is coming within the next 50 years
You're really going out on a limb!
KD has reached the point where he's so entrenched in this particular argument that there's no point in discussing it with him any further, it's either data points that reinforce him or snark all the way down.
I think you're missing the point of the Fed targeting a soft landing.
So n = 6.
Color me unimpressed.