The labor market, daily — jobs report preview, claims, JOLTS, wages

Nowflation payrolls forecast for 2026-10: +102k, release 2026-11-06 8:30 AM ET — on the record before every jobs Friday, graded after.

Weekly claims are the highest-frequency labor read; JOLTS openings and the quits rate lead wages; the Sahm rule and the curve are the recession triggers — all on one page with the payrolls track record against the Street.

What this page measures

The labor market is read here at three speeds: weekly jobless claims (the fastest official series there is), the monthly jobs report and JOLTS, and the quarterly-paced wage trackers. Above them sits Nowflation's own payrolls forecast, put on the record before every jobs Friday and graded after. For the October 2026 report (released November 6, 2026) the model calls +102k and a prior print of +29k.

How the payrolls forecast is built

Model nfp_momentum_claims_v1.0: a monthly payroll change regressed on recent payroll momentum and the change in initial claims, refit every month on prior data only (intercept 75.5k, momentum coefficient 0.354, claims coefficient -4.08 per thousand claims, fitted on 125 months). Walk-forward over the last 24 prints it misses by 93.2k on average against 93k for a naive momentum guess, and beats the naive guess 50% of the time. Walk-forward refit each month on prior data only. PAYEMS history is latest-vintage (revised), so this grades the method, not real-time data vintages.

MonthActual (k)Model (k)Naive (k)
April 2026+148+104+73
May 2026+63+86+69
June 2026+31+83+142
July 2026-10+184+81
August 2026+133+74+28
September 2026+29+117+51

The latest readings and how to read them

Nonfarm payrolls rose +29k in September 2026. Unemployment printed 4.2% for September 2026 against 4.1% expected. The Sahm rule — the three-month average unemployment rate minus its twelve-month low — stands at 0.00pp for September 2026, below the +0.50pp trigger that has accompanied the start of every recession since 1970; it confirms a downturn rather than predicting one, and a trigger is a warning, not proof: it crossed the trigger in July 2024, peaked at +0.57pp in August 2024 and was back below it by October 2024, with no recession. Initial claims were 199,000 in the week of October 10, 2026 and continuing claims 1,710,000; initial claims turn first when hiring turns, continuing claims say how long people stay out. JOLTS: 7,079 thousand openings in August 2026, a quits rate of 1.9% and a hires rate of 3.3%. Quits lead wages — workers quit when they can find better pay — so a falling quits rate is the earliest sign wage growth will cool. Wages: the Atlanta Fed tracker (median, same-person, three-month average) shows 3.9% growth (September 2026) and Indeed's posted-wage index 2.5% (August 2026). Set against the Gauge at 3.45%, that is the real-wage arithmetic the real-wages page carries.

The recession signals on this page

Jobs-report track record — the print against the Street

ReportReleasedActual (k)Consensus (k, FMP calendar estimate)Surprise (k)
September 2026October 2, 2026+29+90-61
August 2026September 4, 2026+162+56+106
July 2026August 7, 2026-23+80-103
June 2026July 2, 2026+57+110-53
May 2026June 5, 2026+172+85+87
April 2026May 8, 2026+115+62+53

First prints are revised twice in the following two months and again at the annual benchmark; the revision tracker keeps the receipts. The labor market matters to inflation through wages: services priced mostly on labor are the supercore slice the Fed watches, and the Fed's own dashboard of these series is on the Fed Watch page.

Terms on this page

Vintage (vintage-true data)
A vintage is a statistic exactly as it stood on a given date, before later revisions; a vintage-true backtest uses only the data that existed the day before each release, so it cannot borrow information from the future. Nowflation stores its own history vintage-true and never silently revises a published value.
Nonfarm payrolls (NFP)
The BLS establishment-survey count of jobs excluding farms, the self-employed and the military; the monthly change is the headline of the jobs report.
Sahm rule
A recession indicator by Claudia Sahm: when the three-month average unemployment rate rises 0.50 points above its low of the prior twelve months, the rule signals that a recession has likely begun. Every US recession since 1970 crossed that line, but so have months with no recession, so a trigger is a warning, not proof.
JOLTS
The Job Openings and Labor Turnover Survey — openings, hires, quits and layoffs — published about five weeks after the month.
Walk-forward backtest
A test that refits the model each month using only data available then and scores the resulting forecast, so nothing from the future leaks into the score.