How to Read the US Jobs Report, Line by Line

On 2 October 2026, the US jobs report counted two different changes in employment for September. Employer payrolls rose 29,000. The household survey counted 406,000 more people with a job.

This guide reads that report line by line, in the order the Bureau of Labor Statistics (BLS) prints it. It covers the payroll change, the household count, the Unemployment Rate, participation, earnings and the revisions to the two months before.

The first line, in Table B-1, is the change in Nonfarm Payrolls: +29,000 in September 2026. BLS described it as little changed. Over the prior 12 months, payrolls had risen an average of 45,000 a month. Our guide to what Nonfarm Payrolls count and how the employer survey is collected covers the survey itself.

A single month carries a wide margin of error. BLS gives the 90% confidence interval on the monthly payroll change as on the order of plus or minus 122,000. That is a range with about a 90% chance of containing the true change. Applied to September, that interval runs from a loss of 93,000 to a gain of 151,000.

The chart splits September's +29,000 into three groups. Goods-producing industries added 18,000 jobs and private service-providing industries added 28,000. Together, that is the private-sector gain of 46,000. Government lost 17,000.

Summary table B prints a diffusion index, which shows how widespread the change was. It is the percentage of 250 private industries adding jobs, with industries that did not change counted as half. A reading of 50% means gains and losses balance. September read 49.0%, down from 57.6% in August.

The household survey is in the A tables, which start with Table A-1. It asks people, not employers. In September, its count of employed people rose by 406,000, from 162.746 million to 163.152 million. That is 14 times the payroll gain.

The two counts differ by definition. BLS lists four differences in its technical note:

The chart compares the two counts for February to September 2026, using the latest published figures. January is left out, because BLS introduces updated population estimates with each January report. The household count swings by hundreds of thousands. It fell 507,000 in June and rose 569,000 in August.

Payrolls moved between a loss of 156,000 and a gain of 214,000. The two counts had opposite signs in March, April and June. Over the eight months, payrolls rose 452,000 and the household count rose 55,000.

The Unemployment Rate was 4.2% in September, up from 4.1% in August. It has stayed between 4.1% and 4.3% since March.

The rate is the number of unemployed people as a percentage of the labour force. Unemployed people have no job, are available to work and have looked for work in the past four weeks, with an exception for temporary layoffs. The labour force is the employed plus the unemployed.

In September there were 7.1 million unemployed people in a labour force of 170.3 million. Our Unemployment Rate guide explains how the headline rate is measured and what the broader U-6 measure adds.

BLS puts the 90% confidence interval on a one-month change in the rate at about plus or minus 0.3 percentage points, at a rate near 6.0%. At a lower rate such as September's 4.2%, the interval is narrower. September's change of 0.1 points is a third of the interval stated for 6.0%.

Two more household lines show who is in the labour force. The participation rate is the labour force as a percentage of the civilian population aged 16 and over. The employment-population ratio is the number employed as a percentage of that population. The gap between them is the unemployed, as a percentage of the population.

The chart covers September 2025 to September 2026. The participation rate fell from 62.5% to a low of 61.4% in July, then recovered to 61.8%. The employment-population ratio fell from 59.7% to 58.9%, then recovered to 59.2%.

Both lines are household survey figures: the solid line is the participation rate and the dashed line is the employment-population ratio. Each has a grey dotted stretch across October 2025, a month with no household figure.

Over the same year, the Unemployment Rate fell from 4.4% to 4.2%. The employment-population ratio ended the year 0.5 points lower. So the rate fell while a smaller share of the population was working.

The participation rate ended the year 0.7 points lower, so people left the labour force. People outside the labour force are not counted as unemployed. The number of unemployed people fell from 7.6 million to 7.1 million.

The release also prints three smaller household lines:

Earnings come from the payroll survey, in Tables B-3 and B-8. Average hourly earnings of all private employees rose 5 cents, or 0.1%, to $37.81 in September. Over 12 months, they are up 3.0%.

For production and nonsupervisory employees, the average was $32.60, up 7 cents. The average workweek stayed at 34.4 hours.

Hours belong beside earnings, because weekly pay is the hourly figure multiplied by hours worked. At $37.81 and 34.4 hours, that is about $1,301 a week. BLS prints the figure directly, as $1,300.66, in Summary table B.

Each jobs report revises the two months before it. In this one, July was cut by 31,000, from +21,000 to −10,000. August was cut by 29,000, from +162,000 to +133,000. Combined, the two months were 60,000 lower than first reported.

BLS prints the revisions in a paragraph at the end of its summary of the payroll survey. The tables label the latest two months as preliminary.

The revised months feed the three-month average. July at −10,000, August at +133,000 and September at +29,000 average 51,000 a month. Before the revisions, the same three months would have averaged 71,000. Our guide to why economic data gets revised covers how large revisions usually are.

The jobs report has six lines: the payroll change, household employment, the Unemployment Rate, participation, earnings and the revisions. Three come from the payroll survey and three from the household survey. Read together, they show where the two surveys agree and where they do not.

This explainer is part of the YX Insights Academy. Each one takes a single idea and checks it against real data.

The same approach runs through everything else we publish:

Good places to start on the website:

The payroll survey counts jobs on employers' payrolls, while the household survey counts people. The household survey includes the unincorporated self-employed, farm workers and people on unpaid leave. The payroll survey leaves them out and counts a person once for each payroll they are on. September 2026 showed +29,000 against +406,000.

BLS states them in a paragraph at the end of the payroll survey section of its summary. The change for each of the two prior months is given with its old and new value. In the data tables, the latest two months are marked preliminary. In September 2026, July and August were cut by 60,000 combined.

The diffusion index shows how many industries are adding jobs. BLS calculates it as the percentage of 250 private industries with rising employment, plus half of those with no change. A reading of 50% means gains and losses balance. It was 49.0% in September 2026.

The participation rate is the labour force, people working or looking for work, as a percentage of the population aged 16 and over. The employment-population ratio counts only those with a job. In September 2026 the rate was 61.8% and the ratio was 59.2%, so the 2.6-point gap is the unemployed as a share of the population.

DISCLAIMER: This article is strictly educational. Any information or analysis in this note is not an offer to sell or the solicitation of an offer to buy any securities. Nothing in this note is intended to be investment advice and nor should it be relied upon to make investment decisions. Any opinions, analyses, or probabilities expressed in this note are those of the author as of the note's date of publication and are subject to change without notice.