ADP Employment Report: +42,000 Jobs in October; wages +4.5% for “Job Stayers” and +6.7% for “Job Changers.”
By Wolf Richter for WOLF STREET.
According to payroll processing firm ADP, the private sector created 42,000 jobs in October, after two months of declines.
But the two prior months, so employment in September and August, had gotten hammered by the massive annual adjustment for the 12-month period through March 2025, to benchmark ADP’s data to the Quarterly Census of Employment and Wages (QCEW) data released by the Bureau of Labor Statistics in September. These adjustments had nothing to do with employment in August and September, but with employment through March 2025.
The annual adjustment had turned the small growth in September of 11,000 into a job loss of 32,000 – which today was re-revised to -29,000. And it had turned the job growth in August of +54,000 into a job loss of -3,000.
Total employment in the private sector rose to a record 134.57 million, up by 977,000 jobs from a year ago.
The chart of private-sector employment shows the trend: Job growth continued in 2025, and the number of jobs rose to a record in October, but that growth was slower than in prior years.
- Year-to-date 2025, jobs added: 597,000
- Same period in 2024, jobs added: 1.33 million.
Median wages increased year-over-year for:
- “Job Stayers”: +4.5%, roughly the same yoy increase for the seventh month in a row.
- “Job Changers”: +6.7%, same yoy increase as in the prior month, but less than the increases of 7.0% to 7.1% in the mid-year period.
The wage data from ADP is based on a subset of 14.8 million workers employed for at least 12 months, whose paychecks ADP processed.
Due to the government shutdown, the BLS will not release its employment report this Friday, and the ADP data here, plus unemployment insurance claims data released by the 50 states individually, and some private-sector tidbits are all we get for now in terms of payroll count and wages for October.
Enjoy reading WOLF STREET and want to support it? You can donate. I appreciate it immensely. Click on the mug to find out how:
More jobs, higher inflation. Combined with less access to federal government data, macro signals should point to no rate cut by the Fed in December. I know it’s fantasy to wish for a Fed backpedal and rate hike, but that is my Christmas wish.
Keep stocking the pantry.
209,and rotate/use older stock and replenish.
You can see the top of your closet,you need more hunting supplies.
Brass is good 👍
I think it is important to track what type of jobs are being created and destroyed (like the birth-death number for the creation/closing of businesses).
Many on Wolf’s forum doubt AI, but it is my humble belief that AI in the next decade will advance to the point that securing and holding a job will be a major feat.
When we get the number of jobs in each category (once the government reopens), will may be able to draw some conclusions as to whether AI is impacting yet.
AI (Atrocious Idiocy) has nothing of value to offer to anyone at all.
Every single Waymo that drives around (there are more and more of them in a bunch of cities now) takes the jobs of 2-3 full-time drivers. The Waymos are amazingly good now. And they’re still just experimental prototype retrofitted vehicles. This technology is going to spread to everywhere. There are over 200,000 taxi/Uber drivers in the US. They need to find another job over the next few years. There are a bunch more other types of drivers, and their days are numbered too, though they have a little longer, it seems.
However, Waymo employs lots of people in tech and other areas, as do automakers that are also all over this tech. But drivers who lost their jobs might not be able to fill the roles at these companies. So there is going to be a lot of tech-driven hardship for some people.
Gee, with all this automation and AI maybe more Americans should be for throttling back on H1Bs for entry level tech jobs, and deporting and not letting in more non-skilled immigration for unskilled jobs that are going away?
I was skeptical of Waymo at first, but its a really great experience. I was tired of getting into dirty smelly cars operated by lazy people. Waymos are cleaned everyday. Bonus, no need tip!
I hailed a Waymo when I was in Phoenix for a work trip (they don’t have them in my area). It was very impressive, and it took only a couple minutes to get used to the idea that there was no driver (mainly because it was so smooth). I left that ride convinced that these cabs will be accepted by the mass market. However, I’m curious about how the business will work? I doubt Tesla and Google will want to own their own fleets. I would think it would make more sense for them to sell the vehicles and a subscription to the owner to run their app. Would Uber buy their own fleet? Would individuals buy their own car(s) and treat them like rental properties? The Waymo cab I was in was Jaguar SUV with sensors and cameras all over it. It wouldn’t surprise me if cost a half million dollars. I know its a prototype. I think the promise of Tesla would be to mass produce lower cost cabs (but we’ve been promised that for years now).
I thought I read that Waymo has not displaced any Uber / Lyft drivers in the SF area because it’s created more demand for ride-hailing overall. Any thoughts on if Jevons’ paradox may increase the overall demand for ride-hailing and therefore Uber / Lyft driver employment will not go down in Waymo markets in the short/medium term?
Read my comment. What does it say?
“And they’re still just experimental prototype retrofitted vehicles. This technology is going to spread to everywhere. There are over 200,000 taxi/Uber drivers in the US. They need to find another job over the next few years. “
“think it is important to track what type of jobs are being created and destroyed”
This is very important – something I’ve been saying for 2 years…
But we don’t need AI – the BLS gives a monthly running estimate broken down into 10-12 broad industry/occupation groups…and generates data for over 800 occupations every May or so (with an ugly lag of 12-16 months).
Scary spoiler? For 2+ years government subsidized jobs (government and medical) have made up a very disproportionate pct of the new jobs added.
But government subsidized jobs very likely cannot bootstrap overall employment growth into existence long term – if taxing government paid/subsidized workers were some sort of perpetual motion machine for macro health, the USSR would have thrived and DC would have already created 5 times the number of phoney baloney jobs it has.
Keynesian perpetual motion aside, there has to be organic, private sector growth for the whole machine to, er…work.
Cas127
1. “For 2+ years government subsidized jobs (government and medical) have made up a very disproportionate pct of the new jobs added.”
1a. Government: Jobs at the federal government and at state governments have plunged in 2025. I discussed this a bunch of times, including charts, all you have to do is read it. So READ IT NOW, including:
https://wolfstreet.com/2025/10/04/jobs-report-to-get-hit-by-100000-federal-government-deferred-resignations-on-top-of-the-jobs-already-shed-but-ice-is-hiring/
https://wolfstreet.com/2025/09/05/my-thoughts-about-that-august-jobs-report/
1b. “government subsidized jobs” such as “Medical”: Sorry, jobs in “medical” (ADP category is “Education and health services”) is essentially flat YoY.
But:
— construction: +1.8% yoy
— financial activities: +1.9% yoy
— leisure & hospitality: +2.4% yoy
— natural resources & mining (oil& gas drilling): +2.7% yoy
2. “the BLS gives a monthly running estimate broken down into 10-12 broad industry/occupation groups…”
Yes, and it’s right here on this site, including charts for each category, all you have to do is read it, but you REFUSE to read the articles here. So READ THIS NOW:
https://wolfstreet.com/2025/07/03/beneath-the-surface-of-the-employment-report-how-jobs-evolved-by-industry-over-time-winners-losers/
Lol. Who needs facts when you have beliefs?
Btw, it is true that “private education and health services” has increased as percentage of all jobs, but the increase has been very slow and (mostly) steady. It has taken 60 years for it to increase from 6% in 1965 to 17% today.
The biggest increases came during the last 3 recessions (1990, 2000, and 2008), because demand for education and health care doesn’t stop during recessions.
You also have to remember that technology can reduce the credentials/skills needed to do a task.
…again wondering about the longer-term effects of AI as a whole on human society-a powerful tool to use when it seriously-uses its facility to think, thereby aiding brain-convolution, but offset by the brain-smoothing effects of removing more of the myriad of minor, but formerly-constant human actions of day-to-day living (…I present a poor analogy here-say the electricity for the power grid seriously-fails-how many, when the wall/wireless switches are out, can then safely and easily make fire and candles for heat and light?…but, of course, we have, and are always, working to stay ahead of and preparing for such eventualities in maintaining our spacecraft’s human viability, aren’t we? The race to the top, or the bottom? Morlocks and Eloi? T’was always thus?)?.
may we all find a better day.
Way,o is Geofenced in cities where they have mapped all the streets. I believe when Tesla finally solves FSD, which is not too far away, their Robotaxis can drive anyware..
Yesterday on my way to work in downtown San Jose in my non-FSD vehicle, I saw a Tesla model 3 that had attempted to make a left turn across the center island of Almaden Blvd and knocked over two large concrete bollards. Apparently FSD is just on the other side of those bollards. But no cats were harmed.
“I saw a Tesla model 3 that had attempted to make a left turn across the center island of Almaden Blvd and knocked over two large concrete bollards.”
And think of all the lost dollars Antifa is wasting on protests and riots when they could just plop some bollards here and there and trash Elon!
You know it was using FSD?
No, you don’t.
Tesla’s reliance on cameras alone is a fatal shortcoming.
Human’s can cope with some poor-visibility situations because they have a mental model of reality. A machine-learning system trained on camera-captured images does not — it relies on interpreting the images in terms of its training set.
For situations that occur with high enough frequency, the training set will be large enough for proper decision-making. But for rare situations that are unlikely to be in the training set — situations where a mental model of reality is essential for proper interpretation of the visual input, decision-making is likely to fail. Google “Huang Tesla crash”.
Musk fan-boys love to post on X.com how their Tesla drove them on multi-hundred mile trips in FSD with no problem.
Depending on Tesla FSD is like camping or building a home near a river like the Guadalupe River. Just because it doesn’t drown you on typical rainy days doesn’t make it safe.
> Musk fan-boys love to post on X.com how their Tesla drove them on multi-hundred mile trips in FSD with no problem.
Guaranteed paid posts. Nothing wrong with advertising.
@Zoroto many Musk fanboys are in the closet now (with “I bought this before Elon went crazy stickers on their Teslas) but there are still a TON of people that can’t stop talking about how they love there Teslas (most have more than one) even on the super blue SF Pensula. P.S. The left leaning press seems to hate Waymo even more than Tesla (I bet I have seen at leat a dozen stories about the SF cat that a Waymo ran over)…
I do not understand Musk’s reliance on just cameras other than pure stubbornness (which he is well known for). I get that in theory, cameras should work just as well (probably better) than the human eye. No argument.
However, having LIDAR in addition to cameras is obviously better. No matter how intelligent you can get just camera feeds to be, having extra information provided by LIDAR is superior. How is this even arguable?
Initially, Musk argued that having only cameras would mean his cars could be sold for a lower price. While that is true, the price difference is becoming negligible, especially considering how valuable the extra information is.
Having cameras + LIDAR > just cameras. The cost is quickly becoming a non-factor given the benefits.
Now it is just stubbornness.
It may mark the first and only example of a big tech company declining to collect more data.
…’pilot error’, organic or computed, will always be with us. (The moving targets of ‘acceptable matter of degree’ and ‘qualified pilots’, as well…).
may we all find a better day.
I am a Model Y owner and have used FSD for multiple 1,000-mile round trips over the past year. I do not fully trust FSD for driving exclusively. It has gotten better from last year to this year, but it still makes mistakes. Phantom braking still occasionally happens. That alone is enough of a deterrent for me to close my eyes at the wheel. There’s also the issue during dawn and dusk where the camera can be blinded by the sun. I’ve had that happen to me on expressway on-ramps/exits when the car is angled downslope. Tesla “freaks out” and has me immediately take control. These are example edge cases (I’ve experienced a few more), but I am much better equipped to handle them compared to Tesla.
With that being said, Tesla is a safer driver than everyone in standard conditions with well-defined roadway lines, traffic lights, etc. It has 360-degree visibility. FSD enhances human driving. I do not believe it will supplant it.
Most accidents don’t happen in standard conditions so it’s probably not a great comparison.
Please define “not too far away” :)
Waymo uses LiDAR to help navigate, tesla does not.. significant difference from what I understand…
Elon once said there would be a million robotaxis on the road by next year (2020, he stated this in 2019…)
And don’t forget about the solar roofs Tesla was going to use to revolutionize the roofing industry…
And the DoJo supercomputer……
Elon will go down in history as the greatest Charlatan to ever walk the earth, would make PT Barnum blush…
Tip of the cap to Elon though running Tesla cap value up to 1.5T with a PE ratio of 300+ (Toyota TM PE Ratio is 10 for reference)
Have a good week everyone..
I wish this site had a “like” button for posts like this.
Ah yes its right around the corner!!!!! This has been promised for near a decade. I dont think they will be able to solve it safley enough with the hardware currently installed. At least thats what the evidence tells me. Also it appears that tesla is not transparent with there data. That dogs not going to hunt either in self driving. Popcorn out.
The Philips Curve has been denigrated.
https://www.stlouisfed.org/open-vault/2020/january/what-is-phillips-curve-why-flattened
“The Great Demographic Reversal” by Charles Goodhart and Manoj Pradhan explores how demographic changes will reshape global economies, leading to rising inflation and decreasing inequality.”
These reports are published before they demonstrate accuracy. And the FED is basing its rate decisions on bad data, let alone outdated relationships.
1. The Fed is not using the Phillips curve (if it ever did). The Phillips Curve is simplistic, and even adherents use more complex models.
2. The Y-axis of the Phillips curve is “wage inflation” not “consumer price inflation.” Big difference. Meaning, low unemployment, including shortages in the labor pool, push up wages = wage inflation (Y-axis); which is exactly what happened in 2020-2023. And now, unemployment is still relatively low, and so wages are still growing at a substantial pace, but not as fast as they did in 2022-2023.
On the other hand, higher unemployment slows wage increases (lower wage inflation).
3. The paper you linked was from January 2020, based on data through 2019, when people really started thinking that the Phillips curve had “flattened out” or that the relationship was dead entirely.
To quote from the article you linked:
”Let’s zoom in on Figure 1 above to look at recent years, starting in 2012. While the unemployment rate has declined to levels not seen in 50 years, inflation has remained low—even below the Fed’s 2% target for most of the period shown in the graph below.
And WOOOSH goes wage inflation in 2021-2022, and those high-flying theories of a flattened Phillips Curve got flattened out by a speeding Safeway truck loaded up with overpriced food.
The Fed seems to base their interest rate decisions pretty heavily on the Taylor rule, not the Philips curve.
I happen to agree with your analysis of the illogic of the Phillips curve, the one that Dick Cheney learned about when his tutor drew a conceptual curve on a bar napkin.
Except that the political propaganda, promoting policies based on the delusion that in spite of a statistical correlation that loses to the simple coin toss, has been the bedrock reason for the sequential illogic of economic policy these past several generations.
Jobs down, inflation up. Not an inverse relationship. The FED’s “maximum employment” mandate is not plausible. Injecting new money is more likely to induce stagflation. We have regulated capitalism, not laissez faire capitalism. We can’t manage our portfolios under a command economy, but Congress must provide incentives for the economy to grow.
1. Jobs are UP, they’re higher than before, they rose to a record in October!
The reason why ADP had negative job growth in the prior two months (Aug and Sep) was the huge negative adjustment for the 12-month period through March 2025, which has nothing to do with Aug and Sep. Without the adjustment, jobs grew in those two months. From the article:
“The annual adjustment had turned the small growth in September of 11,000 into a job loss of 32,000 – which today was re-revised to -29,000. And it had turned the job growth in August of +54,000 into a job loss of -3,000.”
But even with the adjustments, the number of jobs per ADP rose to a new RECORD in October.
What has happened is that job GROWTH has slowed — meaning employment is growing, but more slowly than before. There could be multiple reasons for that, including labor shortages in some areas, such as construction, due to the reduced supply of illegal immigrant labor.
2. Correct, consumer price inflation has been accelerating.
It might be useful to hear Wolf’s comments about today’s Challengers job report, which looks pretty bad, compared to this much more upbeat ADP report. Especially in terms of reliability, as in how much they reflect reality.
The early BLS numbers are unreliable enough to be almost useless, but maybe the later revisions are okay. Even if the govt opens up, it will be a while (months) before we get a revision, since data apparently are not collected during the shutdown.
You keep repeating this lie, despite being called out on it being a lie repeatedly. There is nothing “so unreliable as to be useless” about BLS numbers.
What *is* true is that month to month changes in job counts are usually within the margin of sampling error, meaning that each individual month is a very rough estimate and one should only use multi-month averages or year on year changes if one reacts to be accurate.
thurd2
1. The Challenger report is a list of GLOBAL layoff “announcements” by US companies. It’s totally meaningless. A bunch of those layoffs are in other countries. If they ever actually become layoffs. These are huge companies with huge operations and staff overseas. Some of these companies get a majority of their revenues from overseas.
2. YOU were lied to in the crisis media about the Challenger Job cut announcements of 155,000, “most layoffs since 2003” or whatever. That was a blatant lie. It was the worst “October” since 2003. But the recent high was in March 2025 at 275,000. And there was a much bigger high in April 2020, at 671,000.
3. Why does anyone pay attention to the stupid-ass headlines in the crisis media? You complain about the BLS but then you eagerly swallow hook, line, and sinker this stupid-ass BS in the crisis media because it suits your narrative?
Chart by Trading Economics of the Challenger job cuts:
Thanks for the explanation. I swallowed nothing. I was just wondering why the reports appear to show different conclusions. Maybe it’s time to take your calm down pill. On the other hand, I do miss your occasional RTGDFA “suggestions”. I like a little spice as internet articles and comments are more and more censored into unending blandness.
It’s time for you to quit dragging ZH BS headlines into here, pretending they’re facts because they suit your narrative. Life is too short to waste it on debunking BS.