§US initial claims first-print forecast for week ending 2026-08-22
Framing and exact resolver: this is the U.S. Department of Labor Employment and Training Administration advance seasonally adjusted initial claims measure, FRED/ALFRED series ICSA, for week ending 2026-08-22. The unit is thousands, matching the registered target. The DOL report is the substantive first-print source; ALFRED is the registered resolution adapter and mirror.
Inspect ETA Data Dashboard latest statistics for seasonally adjusted initial claims.↳ Fetched latest official dashboard statistic: Seasonally Adjusted Initial Claims for 2026-08-08 = 209,000, i.e. 209 thousand; dashboard updated August 14, 2026.Inspect DOL ETA release archive snippets for recent Unemployment Insurance Weekly Claims Reports.↳ Fetched official first-print statistics in thousands: 2026-06-13 = 226, 2026-06-20 = 215, 2026-06-27 = 215, 2026-07-04 = 215, 2026-07-11 = 208, 2026-07-18 = 187.Inspect published Thesis run artifact generated from public DOL/ALFRED sources for the latest pre-dashboard history gaps.↳ Fetched public recorded DOL/ALFRED statistics in thousands: 2026-07-25 = 197, 2026-08-01 = 199, with earlier mirror values 2026-06-27 = 217 and 2026-07-04 = 217 noted as ALFRED-vintage values; I use DOL first-print values when DOL archive text exposes them.Inspect DOL UI weekly claims archive publication schedule and release timing.↳ Fetched official schedule facts: UI Weekly Claims News Release is published each Thursday morning at 8:30am ET; the 2026 exception table lists Wednesday November 25, 2026 at 8:30 AM ET. With 2026-08-27 a Thursday and not the listed 2026 exception, the first-print DOL release date for week ending 2026-08-22 is 2026-08-27, while the registered ALFRED target resolutionDate remains 2026-09-02.Base rate / reference class: the recent official-source reference class is nine first-print seasonally adjusted ICSA weekly values from 2026-06-13 through 2026-08-08: 226, 215, 215, 215, 208, 187, 197, 199, and 209 thousand. The base rate is centered in the low 200s; the median is 215 and the latest value is 209.
Level, momentum, one-off, and policy-mechanism effects: level evidence starts at 209 thousand. Momentum is mixed: the series rebounded from 187 to 197 to 199 to 209, but the June-to-August reference class is not clearly trending upward. The 187 thousand July 18 print looks like a one-off low. No policy mechanism or holiday exception points to a structural claims jump for the Aug. 22 week.
Prior/update/interval: prior = two-week last-print persistence model using the fetched official-source ICSA reference class through 2026-08-08; adjustment components = -2 thousand for mild reversion from the latest rebound toward the recent low-200s center; point = 209 - 2 = 207. Interval method = realized dispersion of successive weekly changes in the fetched first-print history. Changes were -11, 0, 0, -7, -21, +10, +2, +10 thousand; sigma = 10.55 thousand; one-week 80% half-width = 1.28*sigma = 13.50 thousand. Because the target is two unreleased weeks after the latest print, I widen by sqrt(2): 13.50*1.414 = 19.09 thousand. Final implied bounds = 207 +/- 19.09 = [187.91, 226.09], rounded to [188, 226].
Counter-considerations: upside risk is a genuine layoff uptick or seasonal-adjustment miss after the Aug. 8 rebound, which would land above the interval if the first print is over 226 thousand. Downside risk is another July-like low-claims week, which would land below the interval if the first print is under 188 thousand. An outside the interval outcome would most likely require either a renewed sub-190 low or a broad claims jump above the June highs.
Review disposition: accepted the blocking critique to keep top-level resolver fields byte-consistent with the registered ledger contract by using resolutionDate 2026-09-02 and the ALFRED ICSA graph CSV binding, while retaining DOL as the substantive first-print source in reasoning; accepted the ET wording clarification.