The Payroll Benchmark Revision Lands at 10 a.m. Friday. That Is About the Same Minute Warsh Is Expected at the Jackson Hole Podium.

by Nora Halloran · August 25, 2026

Senior Analyst, US Market Current

The Bureau of Labor Statistics has scheduled the Current Employment Statistics Preliminary Benchmark for the nation on Friday, Aug. 28, at 10 a.m. Eastern time, with March 2026 as the reference month. The agency's release calendar shows two companion items at the same hour that morning: the state and area version of the preliminary benchmark, and County Employment and Wages for the first quarter of 2026 — the quarterly census that supplies the administrative employment counts the benchmark is measured against.

That is the same hour Federal Reserve Chair Kevin Warsh is expected to speak at the Kansas City Fed's Jackson Hole symposium. TechTimes reported on Aug. 21 that Warsh's keynote was expected around 10 a.m. ET Friday, with precise timing awaiting the Kansas City Fed's agenda, which the outlet said was expected Wednesday evening. As of Tuesday the Kansas City Fed's symposium page still listed only the dates, Aug. 27-29, and the topic, "Financial Innovation: Implications for Payments and Policy." No program or speaker list had been posted.

The collision matters because the benchmark revision is the single largest scheduled adjustment to the payroll record in the calendar year, and, if the expected timing holds, it would land in the same window as the most closely watched Fed communication of the quarter.

What the benchmark does is narrow. Once a year, BLS re-anchors its survey-based payroll estimates to a near-complete count of jobs drawn from state unemployment insurance tax records. The preliminary estimate published in August is the difference between the March survey level and that administrative count. It is not a new monthly jobs number, and it does not by itself change any published monthly print. The final benchmark is what gets folded into the official series, and that happens the following February with the January employment report.

Last year's sequence shows how wide the gap between preliminary and final can be. On Sept. 9, 2025, BLS put the preliminary benchmark revision for March 2025 at minus 911,000 jobs, or minus 0.6%, with the private sector accounting for minus 880,000, or minus 0.7%. The final benchmark, incorporated on Feb. 11, 2026 with the January 2026 estimates, came in at minus 862,000, or minus 0.5%, on a not seasonally adjusted basis — about 50,000 jobs shallower than the preliminary figure had implied. Two caveats on that pairing: the narrative text of the BLS benchmark article states minus 861,000 while its total-revision table carries minus 862,000, and the seasonally adjusted comparison in the same document is minus 898,000, or minus 0.6%.

The 2025 preliminary detail was concentrated rather than uniform. Trade, transportation and utilities was revised down by 226,000, or 0.8%; leisure and hospitality by 176,000, or 1.1%; retail trade by 126,200, or 0.8%; wholesale trade by 110,300, or 1.8%; and information by 67,000, or 2.3%. BLS attributed the overstatement principally to businesses reporting lower employment to the quarterly census than to the monthly survey, along with disparities in who responded to each.

Context for the size of these numbers comes from BLS's own history. The bureau's benchmark article states that over the prior 10 years the annual benchmark revision at the total nonfarm level averaged 0.2% in absolute terms, within an absolute range of less than 0.05% to 0.4%. Against that yardstick, the last two years are outliers on the same side: minus 598,000, or 0.4%, for the 2024 benchmark, and minus 862,000, or 0.5%, for 2025.

Going further back, the pattern is smaller and two-directional. The 2023 benchmark was minus 187,000, or 0.1%. The 2022 benchmark was positive — plus 506,000, or 0.3%. The 2021 benchmark was minus 7,000, an absolute revision BLS footnotes as less than 0.05%. The 2019 benchmark was minus 489,000, or 0.3%. A run of consecutive large downward revisions is unusual in that record, which is part of why Friday's number will get read as more than a bookkeeping entry.

There is a specific trap in interpreting the print. The benchmark measures the level of employment in a single month — March 2026 — not the pace of hiring in the months since. A large downward revision would say the March level was lower than the survey had estimated; it would not mechanically say anything about June, July or August. It also does not distribute itself evenly backward: the wedge is only allocated across prior months when the final benchmark is applied.

For markets the practical question is sequencing. If the Kansas City Fed publishes its agenda Wednesday evening and Warsh's slot lands on top of the 10 a.m. data drop, traders will be pricing a benchmark number and parsing a Fed chair's Jackson Hole remarks inside the same window, with the week's other macro material — July PCE, the second estimate of second-quarter GDP and July durable goods — already behind them from Wednesday, per Newsquawk's calendar. Newsquawk's own preview treated the payrolls revision as genuine event risk rather than a technical footnote.

Nothing about Friday changes the August employment report, which BLS has scheduled for Sept. 4 at 8:30 a.m. ET. What it changes is the base those subsequent prints are read against, and the confidence with which anyone can describe the level of employment in the American economy as of this spring.

The preliminary benchmark will be public at 10 a.m. Friday. Whether Warsh's remarks land in the same window depends on an agenda the Kansas City Fed had not published as of Tuesday. If they do, which one moves more of the curve is the open question.