ADP says jobs were down 33,000 in June. The BLS says they were up 147,000. The BLS credited the gains to public education, health care, and state and local government. Schools are on vacation now. Hospitals in my area are laying off and cutting services. My state and local governments are cutting budgets.
Are we living in the times of “Who are you going to believe me or your lying eyes?”
https://www.consumeraffairs.com/news/the-economy-added-147000-jobs-in-june-070325.html
Meet our little friends on Wall Street.
As commented on another thread, I learned that reports I wrote, or comments I made, needed to support the “JC’s” narrative, or consequences would befall me.
Steve…yesiree…everything is the best it’s ever been
I stopped paying attention to the ADP reports decades ago – far too erratic.
DB2
It’s not just jobs data, the BLS is quickly becoming just BS.
The headline number often masks the reality. Here’s the WSJ explanation:
U.S. job growth looked solid in June, but the headline number hides a stark reality: Many private employers aren’t hiring anymore.
Weighed down by high interest rates, a federal crackdown on immigration and uncertainty over tariffs, more companies are deciding they are better off with a smaller head count.
Last month, businesses added 74,000 new jobs, an anemic number compared with previous months. Private-sector job growth fell to the lowest level since October 2024. Of the 147,000 total new jobs added in June, nearly half were in government, bolstered by a jump in state and local government jobs.
On top of that, more than half of private industries cut jobs in June, Labor Department data indicates—only the third time this has happened since April 2020. For most of the postpandemic period, the majority of industries were adding jobs.
Summed up more accurately, our leadership is getting it entirely wrong.
Judgement in DC is stupid.
Nothing new under the sun department. iirc, there was an “adjustment” made in how unemployment was measured, around August 2020, without restating the data from preceding periods, using the new methodology, so the new data was presented without any context.
From Google’s AI thing.
In 2020, the BLS implemented a significant change in how unemployment was measured, particularly in response to the COVID-19 pandemic. This change involved a broader categorization of temporary layoffs, which resulted in a potential underestimation of the unemployment rate. Additionally, the BLS investigated potential misclassification errors where some workers were recorded as employed but absent due to “other reasons”
…because reports must conform to what the “JC” wants to hear.
Steve
The issue isn’t whether the ADP reports are erratic. The issue is whether or not they’re accurate.
Also, ADP measures the private sector using actual payroll data. The DOL/BLS measures both the private and public sectors using surveys.
Personally, I see ADP as the canary in the coal mine.
The “news” was commenting on that last night. Showing a graph of the last several reporting periods, with the private employment line dropping significantly in the last period, and the government employment line suddenly increasing, by roughly the same amount, at the same time. What a coincidence! Reminds me of the time RS parent Tandy Corp showed a drop in net profit for the quarter, but, son of a gun, the company had bought back enough stock, in that same quarter, for EPS to be up.
Steve
If they’re erratic (sometimes very high, sometimes very low, sometimes close) then they’re not useful.
DB2
So if the reports were much more uniform, month after month, year after year, this would be useful regardless of what is actually happening in the labor market?
Pete
I understand. But sometimes the real world is erratic.
Remember, ADP numbers are based on actual payroll data. The DOL/BLS are based on surveys.
When I was a CFO, I tried to respond to the BLS surveys, but sometimes my boss had other priorities, so it sat on my desk for a month or two. Occasionally, it never got filed. But you can bet your bottom dollar I filed payroll reports every 2 weeks.
Great point.
There is noise from measurement error, like survey error, and there is noise intrinsic to the data, the real world.
I once had a discussion with a conservative friend about this. He seemed to change his mind about his fiscal leaning depending on what party was in the white house. When a Democrat was in we had to balance the budget but when a Republican was in , it didn’t matter. That is when I realized that he was just using it for a cudgel, political points. He really didn’t have a moral standing on any ground he was on. His whole political stance was vaporware.
Sure. If ADP were consistently 10% low (or whatever) then that would be useful information. For your own economic well being, I would not recommend you take market positions based on ADP data.
DB2
Yes, I had a similar conversation when a friend decried “Tax and spend democrats” which I responded were better than “Spend and spend republicans.” For some reason it didn’t come up again.