Both nonfarm payroll employment (+29,000) and the unemployment rate (4.2 percent) changed little in September, the U.S. Bureau of Labor Statistics reported today. (Oct 2 )
Not only that, the change in total nonfarm payroll employment for July was revised down by 31,000, from +21,000 to -10,000, and the change for August was revised down by 29,000, from +162,000 to +133,000. With these revisions, employment in July and August combined is 60,000 lower than previously reported. - https://www.bls.gov/news.release/pdf/empsit.pdf
That is an excellent question--and is going to be a central theme when I digest the Employment Situation Summary (that will be Monday's article!)
Obviously, at least one of the two reports is way wide of the mark (and both have histories of being so).
There are lies, damned lies, and statistics, and then we get the BLS job numbers, which at least occasionally wind up in category all their own.
What makes the jobs numbers per the BLS particularly intriguing is that the same report which only saw 29,000 jobs created saw the employment level increase by ~400K, and the cohort not in the labor force decrease by ~350K.
At the same time, 60,000 jobs from prior months were just "erased" by revisions.
This report is wonky on multiple levels.
(This, of course, is why I compare BLS math to Lou Costello's demonstration of 13 going into 28 7 times. Vaudevillian lunacy is the only comparison that fits)
The Fed does use these numbers, and Wall Street takes them seriously as well. That means we can discount the numbers, we can critique the numbers, but we really can't afford to ignore them. Too many other economic actors do not ignore them, for better or worse.
Even with the best of reporting, however, there is always going to be a healthy dose of estimating in these numbers. We should expect that there will be errors, and that means sometimes there will be ginormous errors.
Ideally, what can be communicative even with the inaccuracy are the overarching trends. If the reports are generated consistently, in theory the potentials for error and inaccuracy should be largely the same month to month, which means trends and changes are going to be relatively reliable even if the individual months are wide of the mark.
That's what makes the BLS undershooting projections troubling. It's not as if these numbers exist in a vacuum. Not only is there the ADP report, but there are also the private-sector estimates in the PMI data which points towards job growth. Presumably industrial production is ramping up in the US, and that will be putting pressure on employers to hire.
Personal consumption expenditures were up in August, so if people are spending at the very least leisure, hospitality and retail should be getting pressure to add workers.
There have been signals pointing towards a larger jobs print than this.
I know you’ve been concerned about a jobs recession for a long time, Peter, so if the new data reassures you (for now), today’s numbers are indeed great news. And great numbers are just what Republicans need mere weeks before the midterms!
In the sectors that saw jobs decline, such as Information and Profession Services, is there any explanation for it? Are analysts blaming AI, or some worrisome erosion in the economy?
Enlighten me, Peter, what am I missing here?
ADP says 90,000 jobs
The BLS says 29,000 jobs were created
Both nonfarm payroll employment (+29,000) and the unemployment rate (4.2 percent) changed little in September, the U.S. Bureau of Labor Statistics reported today. (Oct 2 )
Not only that, the change in total nonfarm payroll employment for July was revised down by 31,000, from +21,000 to -10,000, and the change for August was revised down by 29,000, from +162,000 to +133,000. With these revisions, employment in July and August combined is 60,000 lower than previously reported. - https://www.bls.gov/news.release/pdf/empsit.pdf
So where is the job growth????
That is an excellent question--and is going to be a central theme when I digest the Employment Situation Summary (that will be Monday's article!)
Obviously, at least one of the two reports is way wide of the mark (and both have histories of being so).
There are lies, damned lies, and statistics, and then we get the BLS job numbers, which at least occasionally wind up in category all their own.
What makes the jobs numbers per the BLS particularly intriguing is that the same report which only saw 29,000 jobs created saw the employment level increase by ~400K, and the cohort not in the labor force decrease by ~350K.
At the same time, 60,000 jobs from prior months were just "erased" by revisions.
This report is wonky on multiple levels.
(This, of course, is why I compare BLS math to Lou Costello's demonstration of 13 going into 28 7 times. Vaudevillian lunacy is the only comparison that fits)
With revisions everywhere does it even matter anymore, except the Fed bases it's policies and interest moves on all of these so-called facts.
Zero Hedge had an interesting article that I'm linking as well about this today: "Can We All Just Admit This Is Just Absurd?!" - https://www.zerohedge.com/markets/can-we-all-just-admit-just-absurd
The Fed does use these numbers, and Wall Street takes them seriously as well. That means we can discount the numbers, we can critique the numbers, but we really can't afford to ignore them. Too many other economic actors do not ignore them, for better or worse.
Even with the best of reporting, however, there is always going to be a healthy dose of estimating in these numbers. We should expect that there will be errors, and that means sometimes there will be ginormous errors.
Ideally, what can be communicative even with the inaccuracy are the overarching trends. If the reports are generated consistently, in theory the potentials for error and inaccuracy should be largely the same month to month, which means trends and changes are going to be relatively reliable even if the individual months are wide of the mark.
That's what makes the BLS undershooting projections troubling. It's not as if these numbers exist in a vacuum. Not only is there the ADP report, but there are also the private-sector estimates in the PMI data which points towards job growth. Presumably industrial production is ramping up in the US, and that will be putting pressure on employers to hire.
Personal consumption expenditures were up in August, so if people are spending at the very least leisure, hospitality and retail should be getting pressure to add workers.
There have been signals pointing towards a larger jobs print than this.
Thanks for the insights.
I know you’ve been concerned about a jobs recession for a long time, Peter, so if the new data reassures you (for now), today’s numbers are indeed great news. And great numbers are just what Republicans need mere weeks before the midterms!
In the sectors that saw jobs decline, such as Information and Profession Services, is there any explanation for it? Are analysts blaming AI, or some worrisome erosion in the economy?