Labor Market Data and Payrolls

On the first Friday of most months, the U.S. releases a single report that can move stocks, bonds, and currencies within seconds of hitting the wire. It's built from two separate surveys stitched together, and understanding what each one actually counts explains why the headline number sometimes seems to contradict the rest of the report.

Two surveys, two questions

The establishment survey asks businesses how many people are on their payroll — this is where the headline "nonfarm payrolls" number comes from, the net change in jobs across the economy excluding farm work, private households, and a few other categories. It's a large, fairly reliable sample, but it can't tell you anything about who is unemployed, since it only counts people who currently have a job.

The household survey asks individuals directly whether they worked, looked for work, or neither. This produces the unemployment rate:

$$u = \frac{\text{unemployed}}{\text{labor force}}$$

In words: of everyone either working or actively looking for work — the labor force — what fraction is not currently working. Someone who stops looking for a job entirely drops out of the labor force and doesn't count as unemployed at all, which is why the unemployment rate can fall for a bad reason (people giving up the search) as well as a good one (people finding jobs).

Worked example

A month's household survey shows 160 million people employed and 6 million unemployed (actively looking but without a job), for a labor force of 166 million: $u = 6/166 \approx 3.6\%$. The following month, 200,000 people stop looking for work and drop out of the labor force entirely, while employment and unemployment counts otherwise hold steady. New labor force: 165.8 million, unemployed still 6 million: $u = 6/165.8 \approx 3.62\%$ — barely moved, but if those 200,000 had instead simply found jobs, unemployment would have fallen to roughly $5.8/165.8 \approx 3.50\%$. Same headline direction is possible from two very different underlying stories, which is exactly why the labor force participation rate is checked alongside the unemployment rate, not instead of it.

What this means in practice

Rates markets treat payrolls as one of the clearest windows into whether a central bank will hike, hold, or cut, because labor market strength is a core input to most policy mandates. A strong payrolls print with rising wages can push bond yields up on the expectation that policy stays tighter for longer; a weak print does the opposite. Revisions matter too — the initial print for a given month is often revised in the following two releases as more complete data comes in, sometimes by enough to reverse the market's initial read.