The monthly US employment report is one of the two or three biggest scheduled market events of any month. It is also more complicated than the single number that gets reported, and the complications frequently change the conclusion.
August's report came in hotter than expected, which raised the odds of a Federal Reserve rate rise and pressured markets. This tutorial explains what is actually in the release and how to read it.
Two surveys, not one
The most important structural fact: the report combines two entirely separate surveys, conducted differently, measuring different things.
The establishment survey asks businesses how many people are on their payroll. It produces the headline payrolls figure. The sample is very large, which makes it statistically reliable.
The household survey telephones households and asks who is working. It produces the unemployment rate. The sample is much smaller and therefore noisier.
Why they disagree
They measure genuinely different things, so divergence is normal rather than an error.
The establishment survey counts jobs. Someone working two part-time roles is counted twice. The self-employed are not counted at all.
The household survey counts people. Someone with two jobs counts once. The self-employed are included.
So a report can show strong payroll growth alongside a rising unemployment rate without contradiction. When they diverge persistently over several months, that divergence often signals a turning point neither survey shows clearly alone.
The headline number
Non-farm payrolls is the change in the number of jobs, excluding farm work, government varies by category, and a few other exclusions.
This is what moves markets in the first seconds, because automated systems read it immediately. It is also the figure most likely to be revised substantially later.
The unemployment rate
This measures the share of the labour force that wants work and cannot find it.
The crucial subtlety is the phrase "labour force". People who have stopped looking for work are not counted as unemployed - they are outside the labour force entirely.
This means the unemployment rate can fall for a bad reason: people giving up and leaving the labour force rather than finding jobs. Checking the participation rate alongside it prevents this misreading.
The participation rate
The share of the working-age population either working or looking for work.
Rising participation alongside steady unemployment is genuinely healthy - more people entering the workforce and being absorbed. Falling participation flatters the unemployment rate while describing a weaker economy.
Average hourly earnings: the line that matters
If you read only one figure while the Federal Reserve is fighting inflation, make it this one.
The reasoning is direct. The stubborn part of inflation is services, and services are mostly labour. A repair, a haircut, a medical appointment or a restaurant meal is largely someone's time being sold. When pay rises faster than productivity, businesses either accept thinner margins or raise prices, and most choose the latter.
This is why an employment report functions as an inflation report in disguise, and why strong wage growth is currently treated as bad news by markets.
Revisions: the most ignored part
Every report restates the previous two months. These revisions receive almost no coverage and are frequently large enough to reverse the story.
A headline beating expectations by 40,000 jobs, accompanied by downward revisions of 60,000 across prior months, is a net negative. The market's first reaction will usually be to buy the beat; the correction arrives minutes or hours later when the detail is absorbed.
Revisions also tend to run in one direction near turning points. When the economy is deteriorating, initial estimates get revised down repeatedly. A run of downward revisions is one of the better early warnings available to an attentive reader.
The sector breakdown
Where jobs were added matters, not just how many.
Growth concentrated in a single sector is less encouraging than broad-based hiring. Some categories are considered more cyclically sensitive and therefore more informative about the direction of the economy.
A report showing gains spread across many industries is a healthier signal than the same total driven by one category.
Hours worked
The average working week is a quiet leading indicator that gets very little attention.
Employers reduce hours before cutting staff, because hiring and firing is expensive and disruptive. A falling average working week alongside stable payrolls often precedes actual job losses.
Conversely, rising hours suggest employers are stretching existing staff and may need to hire.
The order to read it in
A practical sequence for the morning of the release:
- Payrolls versus expectations. Drives the immediate reaction.
- Revisions. Do they support or undermine the headline?
- Average hourly earnings. The inflation-relevant line.
- Unemployment rate and participation together. Neither means much alone.
- Hours worked. The quiet leading indicator.
- Sector breadth. Broad or narrow?
Why the reaction depends on the regime
The same data produces opposite reactions depending on what the central bank is worried about.
When inflation is the concern, as it is now, a strong report is bad for markets because it makes rate rises more likely. When growth is the concern, a strong report is good news.
September 2026 is firmly in the first regime. Inflation has been above target for more than five years, and the Fed chair has publicly committed to bringing it down. Every data point is being read through that filter.
We cover this in detail in why good news is bad news.
The seasonal adjustment caveat
The figures are seasonally adjusted to remove predictable patterns - holiday hiring, seasonal industries, academic year effects.
The adjustment is necessary but imperfect. When behaviour patterns shift, historical seasonal factors become less accurate, producing figures that overstate or understate what actually happened.
This is another reason single-month readings deserve caution and three-month averages are more informative.
What the report cannot tell you
A few honest limitations.
It describes the past, with a delay of several weeks. It cannot distinguish good jobs from poor ones. It says nothing about job security or satisfaction. And it is a statistical estimate built from sampling, revised repeatedly, and imperfect by construction.
The summary
Read the revisions before celebrating the headline. Read average hourly earnings before deciding what the report means for interest rates. Read the unemployment rate and participation together or not at all.
Doing that puts you ahead of most coverage, which typically reports one number and moves on - and it explains why the market's second reaction to a jobs report is so often the opposite of its first.
The birth-death adjustment
A technical detail that occasionally becomes controversial and is worth understanding.
The establishment survey samples existing businesses. It cannot directly count jobs at companies that have just been founded, nor can it immediately detect jobs lost at businesses that have closed without responding.
To account for this, statisticians apply a model estimating net job creation from business formation and closure. This adjustment can add or subtract a meaningful number from the headline figure.
The model works reasonably well in stable periods and less well at turning points, precisely because business formation and failure rates change fastest when the economy shifts. This is one reason initial estimates tend to be revised most heavily around inflection points.
You do not need to follow this closely. It is worth knowing simply because it explains part of why revisions can be so large, and why a single month's headline deserves less confidence than its precision implies.
Other employment indicators worth watching
The monthly report is not the only source, and some alternatives arrive more frequently.
- Weekly jobless claims. The number of people newly filing for unemployment benefits. Published weekly, so far more current than the monthly report. A sustained rise is one of the earliest reliable signals of labour market deterioration.
- Job openings data. How many positions employers are trying to fill. The ratio of openings to unemployed people is a good measure of how tight the market is, and therefore of wage pressure.
- Private payroll estimates. Published by payroll processors ahead of the official report. Correlation with the official figure is imperfect, which limits their usefulness as a predictor.
- Quit rates. How many people voluntarily leave jobs. People quit when they are confident of finding something better, so a high quit rate signals worker confidence and wage pressure.
The quit rate is particularly underrated. It reflects how workers themselves assess their prospects, which is information no survey of employers can capture.
Why the jobs report may matter less this month
A note on relative importance in the current environment.
With inflation the dominant concern and the August consumer price index due on 11 September, ahead of a Federal Reserve decision, the inflation data currently outranks employment data in the market's attention.
That said, the employment report retains a specific significance: it contains the wage figure, and wages drive the services inflation that has proved most stubborn. So the jobs report is partly an inflation report, and the wage line is where those two stories meet.
If the labour market were to deteriorate visibly, the ranking would change quickly. A sharp rise in claims or a weak payrolls figure would shift the committee's focus from prices to employment, which would change how markets interpret every subsequent release.
This article is general information, not financial advice. Do your own research or speak to a licensed professional before making money decisions.
General information, not advice. This article is published to everyone who reads it and takes no account of your circumstances, so it is not a personal recommendation. Trader Suite is not authorised or regulated by the FCA. Trading and investing involve a substantial risk of loss, and you should seek independent advice before acting. Our articles are researched and drafted with AI assistance and reviewed before publishing. Full risk disclosure.
TraderSuite Team
TraderSuite builds indicators and automated strategies for NinjaTrader 8. Our articles are written by the team, researched and drafted with AI assistance, and reviewed before publishing.






