thesisPrototype

Government data

US CPS production occupations employment, September 2026

What will BLS first report in CPS Employment Situation Table A-19 for September 2026 employed people age 16 and over in Production occupations, not seasonally adjusted, in millions?

Forecast

7.7M
7.4M–8M
80% prediction interval
gpt-5.5Run record ↗
7.37.67.98.2January 2026August 2026Oct 20267.7M
historyforecast path80% interval
Probability distribution

Cumulative probability

Chance that the outcome is at or below each value.

0%25%50%75%100%7M7.7M8.4MForecast value

Shaded band: 80% interval (7.4M–8M). Dashed line: point estimate (7.7M).

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

Analysis

§September 2026 CPS production occupations employment forecast

Framing and exact resolver: this is the CPS household-survey Production occupations row, not seasonally adjusted, Total age 16 years and over, reported by BLS in thousands in Employment Situation Table A-19 and converted to millions by multiplying by 0.001. The matching BLS/FRED source code for the same Production occupations employment level is LNU02032213; FRED/ALFRED are used only as public history mirrors, while resolution uses the BLS A-19 page.

▸ Reported tool use: official.lookupmodel report
Fetch current BLS CPS Employment Situation Table A-19 row for Production occupations from https://www.bls.gov/web/empsit/cpseea19.htm.
↳ BLS Table A-19 reports Production occupations at 7,716 thousand for August 2026 and 7,482 thousand for August 2025, equal to 7.716 million and 7.482 million after the 0.001 transform.
▸ Reported tool use: history.lookupmodel report
Fetch recent BLS-source LNU02032213 history from FRED/ALFRED history pages and current series page.
↳ The BLS-source LNU02032213 history gives January 2026 7,786 thousand, February 2026 7,742 thousand, March 2026 7,685 thousand, April 2026 7,883 thousand, May 2026 7,912 thousand, June 2026 7,759 thousand, July 2026 8,121 thousand, and August 2026 7,716 thousand.
▸ Reported tool use: official.lookupmodel report
Verify the Employment Situation release date from the BLS schedule page for September 2026 and compare with the registered target window.
↳ The BLS schedule lists Reference Month September 2026 with Release Date Oct. 02, 2026 and Release Time 08:30 AM; the registered sourceBinding window says 2026-10-06 to 2026-10-14, so the official calendar is earlier by 4 to 12 days.
▸ Reported tool use: history.lookupmodel report
Check the same current BLS/FRED mirror values used for the volatility reference class in thousands and transformed millions.
↳ The fetched 2026 values are 7,786, 7,742, 7,685, 7,883, 7,912, 7,759, 8,121, and 7,716 thousand; transformed by 0.001 they are 7.786, 7.742, 7.685, 7.883, 7.912, 7.759, 8.121, and 7.716 million.

Base rate / reference class: the strongest simple prior is latest-level persistence from the same CPS production occupations series, anchored on August 2026 at 7.716 million. The recent reference class is January-August 2026 exact-series values of 7.786, 7.742, 7.685, 7.883, 7.912, 7.759, 8.121, and 7.716 million, with the year staying mostly in a 7.7-8.1 million band but noisy month-to-month movement.

Level, momentum, one-off, and policy effects: level starts at the August 2026 print of 7.716 million. Momentum is mixed because July's spike to 8.121 was followed by August's fall to 7.716. One-off CPS sampling noise is large for this occupation subgroup, so I do not extrapolate the full July-to-August drop. Policy mechanism is neutral; no discrete public policy event changes the September CPS occupation count.

Prior/update/interval: persistence prior = 7.716 million from August 2026; historical sample = January-August 2026 values 7.786, 7.742, 7.685, 7.883, 7.912, 7.759, 8.121, and 7.716 million; adjustment components = +0.000 million seasonal because no fetched same-series August-to-September transition was used, +0.000 one-off, +0.000 policy, yielding point = 7.716 million. Interval method = RMS successive changes from the fetched 2026 history, chosen as the available exact-series realized-volatility sample in the draft despite being thin for a noisy CPS subgroup: changes are -0.044, -0.057, +0.198, +0.029, -0.153, +0.362, and -0.405 million, so sigma = sqrt((0.044^2 + 0.057^2 + 0.198^2 + 0.029^2 + 0.153^2 + 0.362^2 + 0.405^2)/7) = 0.228 million. The 80% half-width is roughly 1.28*sigma = 1.28*0.228 = 0.292 million; 7.716 - 0.292 = 7.424 and 7.716 + 0.292 = 8.008.

Upside risk is a rebound in the September CPS sample, stronger goods-sector labor demand, or occupational reclassification back into production work, which would land above the interval at more than 8.008 million. Downside risk is another adverse CPS sample, deeper manufacturing weakness, or a larger seasonal September decline, which would land below 7.424 million and outside the interval.

Review disposition: accepted the blocking resolver critique by keeping the cell tied to the registered target resolutionDate of 2026-10-14 while noting the BLS calendar discrepancy; accepted the update critique by removing the unsupported -0.160 million August-to-September adjustment; accepted the interval critique by explicitly labeling the seven-change 2026 volatility basis as thin but intentional given the draft's available exact-series history.

Key drivers

  • August 2026 A-19 level
  • Recent CPS occupation volatility
  • Persistence without unsupported seasonal adjustment
  • No discrete policy mechanism

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 Employment Situation, CPS Table A-19
Resolution date
October 14, 2026· outcome not recorded
Resolution rule
Resolve to the first official BLS value published for September 2026 in CPS Employment Situation Table A-19, row 'Production occupations,' Total, 16 years and over, not seasonally adjusted, reported in thousands. Multiply the published thousands value by 0.001 and report millions rounded to three decimals. Use the first official print only, with no later revisions or later correction exceptions. The official BLS Employment Situation calendar fetched in the draft listed September 2026 for October 2, 2026 at 08:30 AM ET; this conflicts with the registered target window ending 2026-10-14, so the publication cell keeps the registered target resolutionDate and states the discrepancy rather than silently changing the target.
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

The draft is mostly coherent, but it has a blocking target-contract mismatch on resolution date and an unsupported August-to-September adjustment.

  • blocking resolver: The draft sets resolutionDate to 2026-10-02, while the registered ledger target requires the publication cell to stay tied to the target contract whose expectedReleaseWindow ends 2026-10-14.
  • warning update: The -0.160 million 'typical August-to-September NSA softening' adjustment is material but not supported by the cited January-August 2026 sample, which contains no August-to-September transitions.
  • warning interval: The 80% interval is computed from only seven 2026 month-to-month changes, which is an explicit method but a thin volatility sample for a noisy CPS occupation subgroup.
  • info optional_suggestion: Clarify whether the August 2026 value is itself a first print or current table value if that matters for trace wording.
  • info optional_suggestion: Keep FRED/ALFRED framed only as history mirrors, as the draft already does, because BLS is the resolver.

disposition accepted: Review disposition: accepted the blocking resolver critique by keeping the cell tied to the registered target resolutionDate of 2026-10-14 while noting the BLS calendar discrepancy; accepted the update critique by removing the unsupported -0.160 million August-to-September adjustment; accepted the interval critique by explicitly labeling the seven-change 2026 volatility basis as thin but intentional given the draft's available exact-series history.

disposition accepted: Review disposition: accepted the blocking resolver critique by keeping the cell tied to the registered target resolutionDate of 2026-10-14 while noting the BLS calendar discrepancy; accepted the update critique by removing the unsupported -0.160 million August-to-September adjustment; accepted the interval critique by explicitly labeling the seven-change 2026 volatility basis as thin but intentional given the draft's available exact-series history.

disposition accepted: Review disposition: accepted the blocking resolver critique by keeping the cell tied to the registered target resolutionDate of 2026-10-14 while noting the BLS calendar discrepancy; accepted the update critique by removing the unsupported -0.160 million August-to-September adjustment; accepted the interval critique by explicitly labeling the seven-change 2026 volatility basis as thin but intentional given the draft's available exact-series history.

disposition not applicable: Review disposition: accepted the blocking resolver critique by keeping the cell tied to the registered target resolutionDate of 2026-10-14 while noting the BLS calendar discrepancy; accepted the update critique by removing the unsupported -0.160 million August-to-September adjustment; accepted the interval critique by explicitly labeling the seven-change 2026 volatility basis as thin but intentional given the draft's available exact-series history.

disposition not applicable: Review disposition: accepted the blocking resolver critique by keeping the cell tied to the registered target resolutionDate of 2026-10-14 while noting the BLS calendar discrepancy; accepted the update critique by removing the unsupported -0.160 million August-to-September adjustment; accepted the interval critique by explicitly labeling the seven-change 2026 volatility basis as thin but intentional given the draft's available exact-series history.

Activity artifacts

Complete original trace

§September 2026 CPS production occupations employment forecast

Framing and exact resolver: this is the CPS household-survey Production occupations row, not seasonally adjusted, Total age 16 years and over, reported by BLS in thousands in Employment Situation Table A-19 and converted to millions by multiplying by 0.001. The matching BLS/FRED source code for the same Production occupations employment level is LNU02032213; FRED/ALFRED are used only as public history mirrors, while resolution uses the BLS A-19 page.

▸ Reported tool use: official.lookupmodel report
Fetch current BLS CPS Employment Situation Table A-19 row for Production occupations from https://www.bls.gov/web/empsit/cpseea19.htm.
↳ BLS Table A-19 reports Production occupations at 7,716 thousand for August 2026 and 7,482 thousand for August 2025, equal to 7.716 million and 7.482 million after the 0.001 transform.
▸ Reported tool use: history.lookupmodel report
Fetch recent BLS-source LNU02032213 history from FRED/ALFRED history pages and current series page.
↳ The BLS-source LNU02032213 history gives January 2026 7,786 thousand, February 2026 7,742 thousand, March 2026 7,685 thousand, April 2026 7,883 thousand, May 2026 7,912 thousand, June 2026 7,759 thousand, July 2026 8,121 thousand, and August 2026 7,716 thousand.
▸ Reported tool use: official.lookupmodel report
Verify the Employment Situation release date from the BLS schedule page for September 2026 and compare with the registered target window.
↳ The BLS schedule lists Reference Month September 2026 with Release Date Oct. 02, 2026 and Release Time 08:30 AM; the registered sourceBinding window says 2026-10-06 to 2026-10-14, so the official calendar is earlier by 4 to 12 days.
▸ Reported tool use: history.lookupmodel report
Check the same current BLS/FRED mirror values used for the volatility reference class in thousands and transformed millions.
↳ The fetched 2026 values are 7,786, 7,742, 7,685, 7,883, 7,912, 7,759, 8,121, and 7,716 thousand; transformed by 0.001 they are 7.786, 7.742, 7.685, 7.883, 7.912, 7.759, 8.121, and 7.716 million.

Base rate / reference class: the strongest simple prior is latest-level persistence from the same CPS production occupations series, anchored on August 2026 at 7.716 million. The recent reference class is January-August 2026 exact-series values of 7.786, 7.742, 7.685, 7.883, 7.912, 7.759, 8.121, and 7.716 million, with the year staying mostly in a 7.7-8.1 million band but noisy month-to-month movement.

Level, momentum, one-off, and policy effects: level starts at the August 2026 print of 7.716 million. Momentum is mixed because July's spike to 8.121 was followed by August's fall to 7.716. One-off CPS sampling noise is large for this occupation subgroup, so I do not extrapolate the full July-to-August drop. Policy mechanism is neutral; no discrete public policy event changes the September CPS occupation count.

Prior/update/interval: persistence prior = 7.716 million from August 2026; historical sample = January-August 2026 values 7.786, 7.742, 7.685, 7.883, 7.912, 7.759, 8.121, and 7.716 million; adjustment components = +0.000 million seasonal because no fetched same-series August-to-September transition was used, +0.000 one-off, +0.000 policy, yielding point = 7.716 million. Interval method = RMS successive changes from the fetched 2026 history, chosen as the available exact-series realized-volatility sample in the draft despite being thin for a noisy CPS subgroup: changes are -0.044, -0.057, +0.198, +0.029, -0.153, +0.362, and -0.405 million, so sigma = sqrt((0.044^2 + 0.057^2 + 0.198^2 + 0.029^2 + 0.153^2 + 0.362^2 + 0.405^2)/7) = 0.228 million. The 80% half-width is roughly 1.28*sigma = 1.28*0.228 = 0.292 million; 7.716 - 0.292 = 7.424 and 7.716 + 0.292 = 8.008.

Upside risk is a rebound in the September CPS sample, stronger goods-sector labor demand, or occupational reclassification back into production work, which would land above the interval at more than 8.008 million. Downside risk is another adverse CPS sample, deeper manufacturing weakness, or a larger seasonal September decline, which would land below 7.424 million and outside the interval.

Review disposition: accepted the blocking resolver critique by keeping the cell tied to the registered target resolutionDate of 2026-10-14 while noting the BLS calendar discrepancy; accepted the update critique by removing the unsupported -0.160 million August-to-September adjustment; accepted the interval critique by explicitly labeling the seven-change 2026 volatility basis as thin but intentional given the draft's available exact-series history.

calibrated forecast · 80% CI
7.7M[7.4M · 8M]
Target metadata

Data point: bls.cps.employed_people_by_occupation.production.september_2026.first_print

More government data forecasts