Government data

US JOLTS quits rate, September 2026

What will the U.S. Bureau of Labor Statistics JOLTS Table 4 preliminary Total quits rate, seasonally adjusted, be for September 2026 in the first official print?

Forecast

1.9%
1.77%–2.03%
80% prediction interval
gpt-5.5Run record ↗
1.71.822.1January 2026July 2026Nov 20261.9%
historyforecast path80% interval
Probability distribution

Cumulative probability

Chance that the outcome is at or below each value.

0%25%50%75%100%1.57%1.9%2.23%Forecast value

Shaded band: 80% interval (1.77%–2.03%). Dashed line: point estimate (1.9%).

Derived from the point estimate and 80% interval; the agent did not report a full distribution.

Analysis

§Forecast for BLS JOLTS quits rate, September 2026 first print

Resolver: BLS JOLTS Table 4, Total quits rate, seasonally adjusted, preliminary September 2026 first print. The registered target contract sets resolutionDate to 2026-11-02; the official BLS JOLTS release calendar fetched this run lists September 2026 for release on 2026-11-03 at 10:00 a.m. ET, so I keep the cell tied to the registered resolutionDate while noting the calendar discrepancy.

▸ Reported tool use: official.lookupmodel report
Fetched BLS JOLTS release calendar for the target reference month.
↳ BLS schedule lists July 2026 release on 2026-09-01, August 2026 release on 2026-09-29, September 2026 release on 2026-11-03, and October 2026 release on 2026-12-01.
▸ Reported tool use: official.lookupmodel report
Fetched BLS archived JOLTS Table 4 first-print releases for January through March 2026.
↳ First-print seasonally adjusted Total quits rates from BLS Table 4: January 2026 = 2.0 percent with 3,137 thousand quits; February 2026 = 1.9 percent with 2,974 thousand quits; March 2026 = 2.0 percent with 3,171 thousand quits.
▸ Reported tool use: official.lookupmodel report
Fetched BLS archived JOLTS Table 4 first-print releases for April through July 2026.
↳ First-print seasonally adjusted Total quits rates from BLS Table 4: April 2026 = 1.9 percent with 2,977 thousand quits; May 2026 = 1.9 percent with 3,065 thousand quits; June 2026 = 2.0 percent with 3,232 thousand quits; July 2026 = 1.9 percent with 3,056 thousand quits.
▸ Reported tool use: official.lookupmodel report
Fetched current BLS JOLTS Table 4 and summary for same-variant context.
↳ Current BLS July 2026 Table 4 shows Total quits rate = 1.9 percent, Total quits level = 3,056 thousand, June 2026 revised quits rate = 2.0 percent, and July quits changed by -157 thousand from June.

The reference class is same-variant BLS JOLTS Table 4 Total quits rate, seasonally adjusted, first print. Base rate anchor: the seven fetched first prints since January 2026 are 2.0, 1.9, 2.0, 1.9, 1.9, 2.0, and 1.9 percent, giving a mean of 1.943 percent and a median of 1.9 percent.

Level effect favors 1.9 percent because the latest first print was 1.9 and the current revised Table 4 also places July at 1.9. Momentum effect is near zero: the sequence alternates 1.9 and 2.0 with no sustained climb or fall. One-off July industry weakness argues modestly below 2.0, while the policy-mechanism context of low quits as a worker-confidence measure argues against a rapid jump.

Prior/update/interval: persistence prior on the same-variant first-print quits rate uses the January-July 2026 historical sample. Adjustments are level 1.90, momentum 0.00, one-off July weakness -0.02, and mean reversion +0.02, implying point = 1.90 percent. Successive changes are -0.1, +0.1, -0.1, 0.0, +0.1, -0.1; sample sigma = 0.098 percentage point, so 1.28*sigma = 0.125. I use an 80 percent half-width of 0.13, giving 1.90 - 0.13 = 1.77 and 1.90 + 0.13 = 2.03; because the official print is rounded to one decimal, this effectively covers first-print outcomes of 1.8, 1.9, and 2.0 percent.

Counter-considerations: upside risk is a September rebound in voluntary quits from better worker confidence or service-sector churn, which would land above the interval if the first print rounds materially above 2.0 percent. Downside risk is a broader hiring slowdown or weaker labor-market confidence, which would land below the interval if the first print rounds materially below 1.8 percent. The outside the interval cases require a larger-than-recent monthly move.

Review disposition: accepted the blocking resolver critique by restoring resolutionDate to the registered 2026-11-02 contract and recording the BLS calendar discrepancy in reasoning; accepted the optional rounded-outcome clarification for the interval.

Key drivers

  • recent quits-rate persistence near 1.9 percent
  • no clear first-print momentum since January 2026
  • cool but not collapsing labor turnover
  • September forecast made before August JOLTS release

Tool evidence

This run has no captured tool responses. “Reported tool use” in the analysis is the model’s account, not an archived tool response.

Sources and resolution

Official source
U.S. Bureau of Labor Statistics Job Openings and Labor Turnover Survey Table 4
Resolution date
November 2, 2026· outcome not recorded
Resolution rule
Resolve to the preliminary Total quits rate, seasonally adjusted, in BLS JOLTS Table 4 for September 2026 as first published by BLS; use the published percent rounded to one decimal and do not incorporate later revisions, annual benchmark updates, errata, or same-day corrections unless the registered target rule is amended.
Run details

The analysis is the model’s written report. Tool-use descriptions in that report are model claims; the activity artifacts contain the execution record.

thesis.analyst · gpt-5.5 · fast · v2.5.11

pre-submit review · completed

Draft is mostly publishable, but it violates the ledger target contract by changing the resolutionDate from 2026-11-02 to 2026-11-03.

  • blocking resolver: The cell's resolutionDate is 2026-11-03, but the canonical ledger target contract requires 2026-11-02 byte-for-byte.
  • info optional_suggestion: Add one sentence that the 1.77-2.03 interval effectively covers rounded first-print outcomes of 1.8, 1.9, and 2.0 percent.

disposition accepted: Review disposition: accepted the blocking resolver critique by restoring resolutionDate to the registered 2026-11-02 contract and recording the BLS calendar discrepancy in reasoning; accepted the optional rounded-outcome clarification for the interval.

disposition not applicable: Review disposition: accepted the blocking resolver critique by restoring resolutionDate to the registered 2026-11-02 contract and recording the BLS calendar discrepancy in reasoning; accepted the optional rounded-outcome clarification for the interval.

Activity artifacts

Complete original trace

§Forecast for BLS JOLTS quits rate, September 2026 first print

Resolver: BLS JOLTS Table 4, Total quits rate, seasonally adjusted, preliminary September 2026 first print. The registered target contract sets resolutionDate to 2026-11-02; the official BLS JOLTS release calendar fetched this run lists September 2026 for release on 2026-11-03 at 10:00 a.m. ET, so I keep the cell tied to the registered resolutionDate while noting the calendar discrepancy.

▸ Reported tool use: official.lookupmodel report
Fetched BLS JOLTS release calendar for the target reference month.
↳ BLS schedule lists July 2026 release on 2026-09-01, August 2026 release on 2026-09-29, September 2026 release on 2026-11-03, and October 2026 release on 2026-12-01.
▸ Reported tool use: official.lookupmodel report
Fetched BLS archived JOLTS Table 4 first-print releases for January through March 2026.
↳ First-print seasonally adjusted Total quits rates from BLS Table 4: January 2026 = 2.0 percent with 3,137 thousand quits; February 2026 = 1.9 percent with 2,974 thousand quits; March 2026 = 2.0 percent with 3,171 thousand quits.
▸ Reported tool use: official.lookupmodel report
Fetched BLS archived JOLTS Table 4 first-print releases for April through July 2026.
↳ First-print seasonally adjusted Total quits rates from BLS Table 4: April 2026 = 1.9 percent with 2,977 thousand quits; May 2026 = 1.9 percent with 3,065 thousand quits; June 2026 = 2.0 percent with 3,232 thousand quits; July 2026 = 1.9 percent with 3,056 thousand quits.
▸ Reported tool use: official.lookupmodel report
Fetched current BLS JOLTS Table 4 and summary for same-variant context.
↳ Current BLS July 2026 Table 4 shows Total quits rate = 1.9 percent, Total quits level = 3,056 thousand, June 2026 revised quits rate = 2.0 percent, and July quits changed by -157 thousand from June.

The reference class is same-variant BLS JOLTS Table 4 Total quits rate, seasonally adjusted, first print. Base rate anchor: the seven fetched first prints since January 2026 are 2.0, 1.9, 2.0, 1.9, 1.9, 2.0, and 1.9 percent, giving a mean of 1.943 percent and a median of 1.9 percent.

Level effect favors 1.9 percent because the latest first print was 1.9 and the current revised Table 4 also places July at 1.9. Momentum effect is near zero: the sequence alternates 1.9 and 2.0 with no sustained climb or fall. One-off July industry weakness argues modestly below 2.0, while the policy-mechanism context of low quits as a worker-confidence measure argues against a rapid jump.

Prior/update/interval: persistence prior on the same-variant first-print quits rate uses the January-July 2026 historical sample. Adjustments are level 1.90, momentum 0.00, one-off July weakness -0.02, and mean reversion +0.02, implying point = 1.90 percent. Successive changes are -0.1, +0.1, -0.1, 0.0, +0.1, -0.1; sample sigma = 0.098 percentage point, so 1.28*sigma = 0.125. I use an 80 percent half-width of 0.13, giving 1.90 - 0.13 = 1.77 and 1.90 + 0.13 = 2.03; because the official print is rounded to one decimal, this effectively covers first-print outcomes of 1.8, 1.9, and 2.0 percent.

Counter-considerations: upside risk is a September rebound in voluntary quits from better worker confidence or service-sector churn, which would land above the interval if the first print rounds materially above 2.0 percent. Downside risk is a broader hiring slowdown or weaker labor-market confidence, which would land below the interval if the first print rounds materially below 1.8 percent. The outside the interval cases require a larger-than-recent monthly move.

Review disposition: accepted the blocking resolver critique by restoring resolutionDate to the registered 2026-11-02 contract and recording the BLS calendar discrepancy in reasoning; accepted the optional rounded-outcome clarification for the interval.

calibrated forecast · 80% CI
1.9%[1.77% · 2.03%]
Target metadata

Data point: bls.jolts.quits_rate.2026-09.first_print

More government data forecasts