Reading the Jobs Report: Payrolls, Unemployment and Revisions

On a Friday morning each month — usually the first Friday — the Bureau of Labor Statistics publishes The Employment Situation, better known as the jobs report. At 8:30 a.m. Eastern time it can move every major market. The report is easier to read once you know that it combines two separate surveys that sometimes disagree.

Survey 1: the establishment survey (payrolls)

The Current Employment Statistics survey asks a large sample of businesses and government agencies — covering roughly a third of all nonfarm payroll jobs — how many people are on their payrolls. It produces:

• Nonfarm payrolls: the headline monthly change in the number of jobs (“employers added 150,000 jobs”).
• Average hourly earnings: the main wage-growth measure in the report.
• Average weekly hours.
• Jobs by industry: health care, manufacturing, government, and so on.

“Nonfarm” means it excludes farm workers, along with the self-employed and private household employees.

Survey 2: the household survey (unemployment rate)

The Current Population Survey interviews about 60,000 households. It asks people about their own situation, which allows it to count the self-employed and to measure:

• The unemployment rate: unemployed people actively looking for work, as a share of the labor force.
• The labor force participation rate: the share of the adult population working or looking for work.
• U-6, a broader measure that adds discouraged workers and people working part time who want full-time work.

Because the household sample is much smaller, its monthly changes are noisier. That is why payrolls and the unemployment rate can give conflicting signals in a single month: they are measuring different things with different methods.

Revisions matter

The first payroll estimate is based on incomplete responses. BLS revises each month’s figure in the next two reports as more businesses respond, and those revisions can be tens of thousands of jobs. A strong headline accompanied by large downward revisions to the previous two months is weaker than it looks.

Once a year, BLS also benchmarks payrolls to near-complete counts from state unemployment-insurance tax records. The preliminary estimate of this benchmark revision, published in late summer, can shift the level of employment by hundreds of thousands of jobs.

How to read the report

1. Compare payrolls with consensus — but look at the three-month average too, which smooths out noise.
2. Add the revisions to the prior two months.
3. Check the unemployment rate and whether it rose because people lost jobs or because more people started looking (participation rising).
4. Look at wages. Average hourly earnings matter to the Fed as a signal of inflation pressure. Month-over-month changes around 0.3% are roughly consistent with annual wage growth near 3.5%–4%.
5. Look at the mix. Are gains broad, or concentrated in one or two sectors such as health care and government?

Why markets react

Like inflation data, the jobs report matters mainly because of the Federal Reserve, whose mandate includes maximum employment. Strong job and wage growth can mean higher rates for longer; a weakening labor market raises the odds of cuts. As a result, “good news” for workers is sometimes “bad news” for stocks, and vice versa — a paradox that disappears once you see that the market is reacting to expected Fed policy, not to employment itself.

Other labor indicators

The jobs report is only monthly, so analysts also track:

• Weekly initial jobless claims, published every Thursday by the Department of Labor — a timely signal of layoffs.
• JOLTS (Job Openings and Labor Turnover Survey), which reports openings, hires and quits.
• The ADP private payroll report, published two days before the official report — though its month-to-month correlation with the BLS figure is loose.

Taken together, these give a fuller picture than any single Friday headline.