Bloomberg consensus is 4%. This implies near trigger for the Sahm Rule (in real time):
Figure 1 (updated): Real time Sahm Rule (blue), and consensus based on Bloomberg as of 7/4/2024 (tan square). Source: FRED, Bloomberg, and author’s calculations (corrected 7/4 7:22pm h/t Joseph).
If the unemployment rate instead hits 4.1%, then the Sahm rule would not indicate a recession (corrected 7:30). If the consensus is hit in June, and the unemployment rate stays constant at 4% through July then we’ll still be below threshold.
on the other hand, as Calculated Risk concluded yesterday, “Heavy truck sales are solid.”
Addendum: Macroduck brings my attention to Claudia Sahms’ caveat re: this time around.
I’m sure Menzie knows this, but for more casual followers of the Great Claudia, a reminder:
“Why My Recession Rule Could Go Wrong This Time”
https://www.bloomberg.com/opinion/articles/2023-11-07/one-highly-accurate-recession-indicator-could-be-wrong-this-time?srnd=undefined&sref=YAR8Qcu4
Immigration may be causing enough change in the U3 rate to knock the Sahm rule off kilter.
Given that recent inflation performance has been quite good and that some lagged effects of tight monetary policy have yet to arrive, the FOMC should ignore Sahm’s modest honesty and cut rates. Now. Rising recessionary risk isn’t necessary to tame inflation and shouldn’t be tolerated.
We civilians may take comfort from the possibility that the Sahm rule will be wrong; the Fed should not.
Also from Sahm, on a quite different topic:
https://macromomblog.com/2021/07/29/after-the-rain/
Larry Summers is an even bigger schmuck than I thought he was.
“me me me me” “I, I. I, I I” it strikes me as very narcissistic, While she’s right about Summers, it has the same smell everything she writes.says does. self-obsession
An economist who is self obsessed? Uh – what’s new?
@ pgl
Uhm, if you are saying, being self-“navel-gazing” isn’t particular to female economists, but is also a problem rampant among male economists, I am forced to agree with you. Fair point.
“For god sake’s, John Taylor and I are the only two people who have a “Rule” in economic policy.”
Well – Art Okun has a “law” which is better than a “rule”. Or am I misreading Chevron deference?
That’s a great post by Dr. Sahm. I did not know of that backstory until reading it. One of the (many) problems with the “purge the rottenness out of the system” types like Larry and Mellon is that they regard people who disagree with them as part of the rottenness that needs to be purged.
Imagine trying to shoot down the career of a young professional just because she disagrees with you. Down deep, Larry must know he’s not as good as he pretends to be.
Moses, look at #10 on Playboy’s 1975 “best drummer” poll:
https://forum.amcorner.com/proxy.php?image=https%3A%2F%2Fthecarpentersdailynews.files.wordpress.com%2F2022%2F01%2Fimg_9755.jpg%3Fw%3D584&hash=d6087cf45e9d6e69d40f63421c067fb5
While you’re at it, check #12 on the “best vocal group” list.
https://forum.amcorner.com/proxy.php?image=https%3A%2F%2Fthecarpentersdailynews.files.wordpress.com%2F2022%2F01%2Fimg_9756.jpg%3Fw%3D582&hash=6b79d4c6611e0f21bbd6165217c2cb31
Seems like music fans thought Karen ranked higher as a drummer than as a singer, but just barely. Bonham reportedly blew a gasket. Which seems a little uppity considering that he was ranked higher than Art Blakey, Max Roach, Elvin Jones, Jack De Johnette, Shelly Manne…
“If the unemployment rate instead hits 4.1%, then the Sahm rule would indicate a recession. If the consensus is hit in June, and the unemployment rate stays constant at 4% through July then we’ll be at the threshold.”
The low in the last 12 months for the 3-month average is 3.6 so the threshold is a 3-month average of 4.1. Not a single month but a 3-month average of 4.1. If June and July are both 4.0, along with May’s 4.0, then the three month average is 4.0. So, close but not the threshold. The threshold is 4.1.
And then in July, the 12-month low moves up to 3.7 so the threshold becomes 4.2.
You’re quibbling over the meaning of “threshold”? I’m sure we’re all impressed. So here’s how Merriam-Webster defines it in this context:
“a level, point, or value above which something is true or will take place and below which it is not or will not”.
Taking the first condition “…above which something is true…” Menzie is right. The signal is given above 4.0%. Taking the second condition “…below which it is not…” and you are correct. You insist on correcting Menzie when he is not wrong. Not only is he not wrong, and it’s his blog, but given the point he is making, his 4.0% is certainly correct – 4.0% is the number “above which” the Sahm rule signals recession.
Your pretense of being the arbiter of the Sahm rule is tiresome.
FWIW, I just finished my in-depth write-up of the report, and I agree with your math. We are probably going to hear lots of bad commentary in the next few days that The “Sahm Rule” has been triggered. It was not, although it was awfully close!
More importantly, note that Sahm herself has highlighted the problem that the post-pandemic immigration creates for her rule. Jason Furman of Harvard weighs in this morning, as earlier this week, with the exact same analysis. I also wrote up a similar analysis in a comment here several weeks ago (Basically, the Household Survey has badly undercounted immigrants in the last several years. While they all landed jobs in the white hot economy of several years ago, some of them aren’t now. Hence the unemployment rate has been rising steadily, even though layoffs are pretty somnolent.
“Your pretense of being the arbiter of the Sahm rule is tiresome.”
I’m not arbitrating anything. I’m simply pointing out the math, unless you think stating that 2 plus 2 equals four is up for arbitration. The math is quite simple and there is no reason not to be accurate.
““a level, point, or value above which something is true or will take place and below which it is not or will not”.
Taking the first condition “…above which something is true…” Menzie is right. The signal is given above 4.0%. Taking the second condition “…below which it is not…” and you are correct. You insist on correcting Menzie when he is not wrong.
What is this verbal gobbledygook? You say that if the threshold is 4.1 that saying above 4.0 is true. Well by that logic above 3.0 is also true and above 2.0 is also true. You are really taking Bruce Hall insanity now.
And, belatedly, I see that Menzie Chinn has corrected his original posting so it seems that you are in disagreement with him. Are you starting to feel a little stupid now.
I also was making the important point that the Sahm threshold is not a fixed number. The unemployment rate has been slowly increasing over the last 12 months which means the Sahm threshold also increases slowly since it looks at the low over the last 12 months. The threshold will be 4.1 for June and 4.2 for July.
To reach the threshold, the unemployment rate doesn’t just have to increase by some fixed amount. It has to accelerate to overtake the slowly rising threshold.
joseph: Yes, I understood that. I didn’t use the three month average for lowest unemployment rate in the last year, I used originally the lowest monthly rate. My mistake.
Thanks for correcting the error. I have great respect for the work you do — both academic research and training the next generation of young economists. And on this blog you try to impart some understanding of economics to the general public, which is a great service.
I only pointed out the error because Claudia Sahm has been quite vocal about bad actors (not you) misusing the Sahm Rule for political duplicity.
*You meant to say “MacroducK” pointed out the caveat. A happy mistake for me.
Moses Herzog: Corrected – thanks.
Menzie, you can make these kind of mistakes ALL year long, and I’ll forgive you. : )
So what does the model say about 4.05%, which is the number in the June jobs report?
Jake formerly of the LP: FRED uses unemployment rate out to only one decimal point, so using that number, I have updated the graph. Under threshold
NFP rises by over 200 thousand and yet the unemployment rate rose? Let’s look at the details. The employment to population rate did not fall. Rather the labor force participation rate rose.
The 3-month average is 4.0%.
The June rate of 4.1% is the result of a 116,000 rise in employment in the household survey, combined with a 227,000 rise in the labor force. The participation rate rose by 0.1 ppt.
The increase in the labor force is where immigration shows up and is, of course, an estimate.
Indeed. The employment to population remained at 60.1% where it has basically been for 2 years.
A 50 basis-point drop in the funds rate is still the favorite for year end, with the priced-in odds of lower rates just a bit higher in response to today’s data.
Tens down about 5.5 bps. Not helping stocks so far.
Didn’t Powell say he might lower the Federal Funds rate in September if we continue to have good news on inflation.?
We have asked you to stop writing stupid trash but you just can’t help yourself – can you? FYI troll – everyone here except you knows the household survey also showed a decent increase in employment which is why the employment to population ratio stayed at 60.1%. FYI dumbass – even a moron could have figured out that the labor force participation rate rose. But not you.
“Declines in household data and lfpr should be together like synchronized swimmers.”
Can you be more stupid? It is evident you have no clue what any of this means. And yet you babble on. BTW – the labor force participation rate is from the household survey. Dude – try leaning the basics before you make any more really asinine and stupid comments.
One shouldn’t get too exercised about any one month jobs report. The confidence interval for change in number of jobs is huge, plus or minus 130,000, so we can only say that the change in employment was somewhere between 72,000 and 340,000 at 90% confidence. The change from the previous month was not statistically significant according to the BLS.
And the unemployment rate typically has a confidence interval of about 0.40. The June BLS report indicated an actual increase in the unemployment rate of 0.09 but that a change of 0.20 is required for statistical significance according to the BLS.