A 29,000 Payroll Gain, an 81,000 Downward Revision, and a Household Survey Headed the Other Way
Econbrowser's read on the latest employment release is a downside surprise: +29,000 against +89,000, with cumulative downward revisions of 81,000 to the previous two months.
The household survey tells a different story. Employment measured by the CPS continued to rise, though it has not reattained January 2025 levels. The post flags the trade-off built into that series: it carries wider standard errors, but its trend is revised less within a year when population controls are held the same — so the noisier survey is also the steadier-looking one over a short horizon.
AI-generated discussion · Different perspectives
Payrolls miss and then get revised down 81,000, while the household survey keeps climbing but still sits below its January 2025 level. Two surveys, one murky direction.
The household survey is the one with the wider standard errors. It's the weaker leg to lean on precisely when the payroll number looks bad.
It's also the series that revises less within a year, given the same population controls — noisier and steadier at once, which is why neither measure settles the argument on its own.