September Jobs Report 2026

Job Numbers Came In Strong. Here Is Why You Should Hold Them Loosely.

In August, US employers added 162,000 jobs and unemployment held at 4.1%. Economists had forecast 53,000 to 56,000. The report also flipped July from a reported loss of 23,000 jobs into a gain of 21,000, and revised June and July up by a combined 55,000.

That is a good month. It is also three times what anyone expected, which is the part worth sitting with.

I am going to say something here that I would rather not have to say. The federal labor data is thinner than it used to be. Not falsified. Thinner.

Here is what happened to the heft. Three things, and none of them are conspiracies.

Fewer people answering. The survey behind the unemployment rate hit a record-low 64% response rate in November 2025, worse than its pandemic low.

Fewer people counting. BLS lost one in five staff between 2024 and 2026, on a budget that buys 20% less than it did in 2009.

One month gone for good. The October 2025 household survey was never conducted during the shutdown and never will be, which breaks year-over-year math for October 2025 and October 2026.

I am not going to pretend the monthly report is precise.

But I am also not going to join the crowd that has decided the numbers are garbage. Because the evidence does not support that either.

Every August, BLS checks its monthly estimates against actual state unemployment tax records. Last year that check knocked 911,000 jobs off the books and set off a year of open distrust. This year, released on August 28, the check found that the data was off by 79,000, or one tenth of one percent. The ten-year average error is two tenths of a percent.

Read that again. After all the noise, this year's numbers were more accurate than the historical norm. The New York Times reached the same conclusion, and the Cleveland Fed studied the full revision series back to 2010 and found no sign that anything structural had broken.

The failure mode here is imprecision, not manipulation. Those two require different responses. Imprecision means use wider error bars and slower measures. Manipulation would mean stop reading entirely. Only one of those is warranted, and it is the boring one.

Which brings me to the part almost nobody covers.

Is Jobs Data More Reliable Outside the US?

The UK went further than we did. In 2024 the Office for National Statistics stripped its own Labour Force Survey of accredited official statistic status because response rates collapsed. Their budget office still has to use it. Academics at LSE have started substituting a think tank's estimates built from payroll tax records instead.

Britain formally demoted its jobs data. We have not done that. Keep that in mind next time someone tells you the US numbers are uniquely broken.

Long-Term Unemployment Is Rising in Every Major Economy

Look at the headline rates and almost nothing is happening. US 4.1%, unchanged. Canada 6.4%, unchanged. Germany 6.4%, unchanged. The euro area 6.4%, and the full EU has now sat at 6.1% for six straight months. Mexico 2.9%, unchanged since June. The UK 4.9%, down a tenth.

Six economies, five statistical agencies, five different methodologies. All flat. If you only read headlines you would conclude the global labor market is fine.

Now look one layer down: the same thing is wrong everywhere.

In the US, 1.9 million people have been jobless 27 weeks or longer, 27.0% of everyone unemployed. In Canada it is 24.0% of 1.5 million. In Mexico, the number of people searching six months or more went from 90,482 to 137,545 in a single year. EU youth unemployment is 15.1%, and the UK's is 16.2%, near a twelve-year high. Britain has lost 78,000 payrolled employees over the year while its unemployment rate improved. Germany is at its highest unemployment rate since 2020, with the unadjusted count above 3 million for a second month.

Mexico is the clearest case of why the headline misleads. A 2.9% unemployment rate sounds like full employment until you learn that 56.2% of Mexican work is informal. You cannot be counted as unemployed if you were never counted as employed.

Here is why this matters more than any single print. That pattern of flat headlines and stuck workers underneath shows up across five agencies with five different sets of measurement problems. When a finding survives that much methodological noise, I trust it more than I trust one Friday number from one agency.

Nobody is getting fired. Everybody is stuck. Six economies agree.

Is Rust-Out the New Burnout?

No. It is not a new term. It was coined in 1975. 

Here’s what it means. Not too much work. Too little that matters.  Underused instead of overloaded. Skills going quietly rusty. A chronic sense of being underchallenged and stuck. Bored. 

This is what a frozen labor market feels like from the inside. People sitting in jobs they have outgrown, with nowhere obvious to go.

Five things job seekers can do

Watch your own industry, not the headline. August's gains came mostly from restaurants and school hiring. Leisure and hospitality added 62,000 while the information industry lost jobs, and ADP showed manufacturing shedding 17,000 in the same month. A strong national print can sit on top of a bad quarter in your sector.

Understand that moving still pays, and staying still costs. Base pay for people who changed jobs rose 4.7% over the year. For people who stayed, 3.0%. That gap is the entire financial argument for a move, and it compounds every year you do not make one.

Stop waiting for a promotion to justify the search. Promotions fell in 10 of 11 sectors last year, and the promotion rate is down 25% from its 2022 peak. The ladder is not slow this year. It is missing rungs.

Budget for a longer search than you think. In August, 1.9 million Americans had been jobless for 27 weeks or longer, more than one in four of everyone unemployed. Low firing does not mean fast hiring. If you land in a search, plan for months, not weeks.

Build your exit before you need it. Quits hit their lowest level since August 2020. The door is narrow, which means the work of getting through it starts while you still have a paycheck. Network on a calendar, not on an emergency.

Five things career coaches can do

Screen for rust-out, not just burnout. Most intake forms only ask about overwhelm. Ask the other half. Bored most of the time. Work feels meaningless. Not learning or growing. Stuck but not overwhelmed. Those are the signs.

Name the structure so the client stops blaming themselves. Rust-out is a mismatch between a person and their work, not a personal failing. Clients who understand they are stuck for structural reasons act. Clients who think they are lazy stall.

Retire the stretch-assignment prescription as a default. The standard advice for a bored employee is to go ask for more challenging work. In a market where promotions fell across 10 of 11 sectors, that advice puts the burden on the person with the least leverage in the room.

Stop handing clients the monthly headline. They are getting whiplashed by a number that came in three times higher than forecast, from a survey with a record-low response rate. Give them duration and youth unemployment instead. Those move slowly and they are what a client's search actually runs into.

Know where your scope ends. Job boredom is not just a mood. Employees with higher job boredom showed reduced heart rate variability, and chronic workplace boredom is linked to depression and anxiety. Those presentations overlap with rust-out. Have referral relationships ready, and use them.

One thing employers can do

Do not let your staff rust in place.

Your retention risk right now is not overwork. It is boredom. Bored employees are 2.1 times more likely to be thinking about leaving.

And the cheapest fix is the one most companies skip. Only 42% of employees got any upskilling in the past year. Those who got none were 140% more likely to feel insecure in their jobs.

Here is the number I would put on a wall. 82% of employees feel secure in their current role. Only 62% feel secure about their future with the company. That 20-point gap is the exact place people quit from.

So stop measuring whether your people are busy. Start measuring whether they are learning. Rotate the administrative load. Ask what strengths went unused this quarter. Protect time for the work that actually requires their judgment.

A frozen labor market is doing you a favor right now. Your bored people cannot leave. That is temporary, and they will remember which employers used the pause to invest in them and which ones just enjoyed the low turnover numbers.

‍