KARACHI: Friday's United States jobs report moved markets on three continents, as our payrolls report records. This explainer sets out how the Bureau of Labor Statistics produces the number, and why the first version is never the final one.
Two surveys, two headlines
The headline jobs figure — nonfarm payrolls, up 165,000 in October — comes from the establishment survey, a monthly sample of employers' payroll records, excluding farm work. The unemployment rate — 4.1% in October — comes from a different instrument entirely, the household survey, which asks people whether they are working or looking for work. Because the surveys measure different things, they can diverge in a single month without either being wrong.
Both figures are seasonally adjusted: the bureau removes the predictable calendar patterns — summer hiring, holiday retail, school terms — so that the monthly change reflects the economy rather than the season. The unrounded unemployment rate in October was 4.141%, against 4.1753% the month before, the release's detail showed — a reminder that the one-decimal headline rounds a continuous measure.
Why revisions are built in
Employers return their payroll data over several weeks, and the bureau publishes its first estimate from the responses in hand, then revises it twice as late returns arrive and seasonal factors are re-estimated. On Friday, August was revised up by 53,000 to 186,000 and September by 31,000 to 60,000 — a combined 84,000 — in the same release that printed October's first estimate.
That is why market coverage treats the level and the revisions together: October's beat against the 65,000 consensus, plus an 84,000 stronger past, described a firmer labour market than either number alone. It is also why a single month's first print is the least reliable version of itself.
Why the Fed — and Karachi — watch it
The Federal Reserve's mandate covers employment as well as prices, so the report is one of the two or three releases that can reprice its expected path in an afternoon — as Friday's move in December odds showed, in our rates report. The next release, for November, is scheduled for Friday, December 4.
This explainer describes the release and its conventions, and is educational content, not financial advice.
Key takeaway: Payrolls are two surveys and a standing revision schedule: Friday's 165,000 and the 84,000 upward revision were one report, read whole.