thesisPrototype

Government data

US computer and math employment, July 2026

How many people were employed in computer and mathematical occupations in July 2026, not seasonally adjusted, in the BLS CPS first print?

Overdue · was due Aug 7, 2026

The resolver routes this target's reference, but its last run printed no line naming this target.

NO_REPORT · resolver status as of Oct 7, 2026

Forecast

7M
6.6M–7.4M
80% prediction interval
gpt-5.6-solRun record ↗

Cumulative probability

Chance that the outcome is at or below each value.

0%25%50%75%100%5.9M7M8.1MForecast value

Shaded band: 80% interval (6.6M–7.4M). Dashed line: point estimate (7M).

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

Analysis

§July 2026 computer and mathematical employment forecast

The target is the total, age 16 and over, for Computer and mathematical occupations in CPS Table A-19, reported in thousands and not seasonally adjusted. Resolution uses the strict August 7 first print and the table's 0.001 conversion to millions.

▸ Reported tool use: official.lookupmodel report
Fetch the latest BLS CPS Table A-19 observation and matched year-earlier value.
↳ BLS Table A-19 reports 6,950 thousand employed in June 2026 and 6,602 thousand in June 2025 for Computer and mathematical occupations.
▸ Reported tool use: official.lookupmodel report
Fetch recent BLS annual-average occupation levels from CPS Table 9.
↳ BLS CPS Table 9 reports 6,386 thousand in 2024 and 6,711 thousand in 2025; the 2025 figure is an 11-month average because October data were not collected.
▸ Reported tool use: official.lookupmodel report
Fetch earlier official CPS annual-average reference-class observations.
↳ Official CPS Table 9 archives report 5,688 thousand in 2021, 6,171 thousand in 2022, and 6,502 thousand in 2023.
▸ Reported tool use: official.lookupmodel report
Verify the official release date from the BLS CPS calendar and latest Employment Situation notice.
↳ The BLS calendar lists the July 2026 Employment Situation for August 7, 2026, and the June release specifies August 7, 2026 at 8:30 a.m. ET; this is 1 day after the registered expected window end.

The outside-view base rate is gradual growth with substantial CPS noise: annual-average employment rose from 5.688 million in 2021 to 6.711 million in 2025, while June 2026 was 0.348 million above June 2025.

Level, momentum, one-off, and mechanism effects: the 6.950 million June level supplies the persistence anchor; positive multi-year growth adds 0.050 million; a judgmental small offset for weak technology hiring subtracts 0.020 million; no identified July-specific policy or classification change warrants another adjustment.

Prior/update/interval: persistence model prior = June 2026's 6.950 million; historical sample = 2021-2025 annual averages of 5.688, 6.171, 6.502, 6.386, and 6.711 million. Their successive changes are +0.483, +0.331, -0.116, and +0.325 million, giving trend sigma = 0.236 million. Because annual averages smooth the monthly CPS target, add a monthly sampling/noise proxy of 0.239 million, the absolute gap between June 2026's 6.950 million and the latest 2025 annual average of 6.711 million. Combining these independent components gives sigma = sqrt(0.236^2 + 0.239^2) = 0.336 million. Adjustments are +0.050 million for longer-run momentum and -0.020 million for soft hiring, yielding 6.950 + 0.050 - 0.020 = 6.980 million. The 80% half-width is 1.28*sigma = 1.28*0.336 = 0.430 million, implying 6.550 to 7.410 million.

Upside risk comes from faster AI-related hiring, labor-force re-entry, or favorable CPS sampling and would land above the interval if employment exceeds 7.41 million. Downside risk comes from layoffs, weak hiring, or adverse sampling and would land below the interval if employment is under 6.55 million; either outcome would be outside the interval.

Review disposition: Accepted the blocking interval critique by adding an explicit monthly sampling/noise component to the annual-change volatility estimate and widening the bounds. Retained the August 7 release-date discrepancy and labeled the weak-hiring offset as judgmental.

Key drivers

  • June 2026 level of 6.950 million
  • positive longer-run occupational employment trend
  • CPS monthly sampling volatility
  • judgmental small offset for weak technology hiring

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
August 7, 2026· outcome not recorded
Resolution rule
Resolve to the total number of employed people age 16 and over in the 'Computer and mathematical occupations' row of CPS Employment Situation Table A-19 for July 2026, not seasonally adjusted, as first published by BLS on August 7, 2026. Convert thousands to millions by multiplying by 0.001 and retain three decimal places. Ignore subsequent revisions. The registered source window ending August 6 conflicts with the official BLS calendar and June release notice, both of which specify August 7.
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-5.6-sol · fast · v2.2.0

pre-submit review · completed

Draft is mostly publishable on target identity and reasoning structure, but the interval method needs a stronger target-relevant volatility basis before publication.

  • blocking interval: The 80% interval is based on four changes in annual-average employment, which is a smoothed trend sample and does not directly measure month-to-month CPS Table A-19 first-print volatility for a July level forecast.
  • info optional_suggestion: Keep the August 7 official release-date discrepancy explicit because it conflicts with the registered expectedReleaseWindow ending August 6.
  • info optional_suggestion: Tie the weak-tech-hiring adjustment to a cited public labor-market indicator or state that it is a judgmental small offset.

disposition accepted: Review disposition: Accepted the blocking interval critique by adding an explicit monthly sampling/noise component to the annual-change volatility estimate and widening the bounds. Retained the August 7 release-date discrepancy and labeled the weak-hiring offset as judgmental.

disposition not applicable: Review disposition: Accepted the blocking interval critique by adding an explicit monthly sampling/noise component to the annual-change volatility estimate and widening the bounds. Retained the August 7 release-date discrepancy and labeled the weak-hiring offset as judgmental.

disposition not applicable: Review disposition: Accepted the blocking interval critique by adding an explicit monthly sampling/noise component to the annual-change volatility estimate and widening the bounds. Retained the August 7 release-date discrepancy and labeled the weak-hiring offset as judgmental.

Activity artifacts

Complete original trace

§July 2026 computer and mathematical employment forecast

The target is the total, age 16 and over, for Computer and mathematical occupations in CPS Table A-19, reported in thousands and not seasonally adjusted. Resolution uses the strict August 7 first print and the table's 0.001 conversion to millions.

▸ Reported tool use: official.lookupmodel report
Fetch the latest BLS CPS Table A-19 observation and matched year-earlier value.
↳ BLS Table A-19 reports 6,950 thousand employed in June 2026 and 6,602 thousand in June 2025 for Computer and mathematical occupations.
▸ Reported tool use: official.lookupmodel report
Fetch recent BLS annual-average occupation levels from CPS Table 9.
↳ BLS CPS Table 9 reports 6,386 thousand in 2024 and 6,711 thousand in 2025; the 2025 figure is an 11-month average because October data were not collected.
▸ Reported tool use: official.lookupmodel report
Fetch earlier official CPS annual-average reference-class observations.
↳ Official CPS Table 9 archives report 5,688 thousand in 2021, 6,171 thousand in 2022, and 6,502 thousand in 2023.
▸ Reported tool use: official.lookupmodel report
Verify the official release date from the BLS CPS calendar and latest Employment Situation notice.
↳ The BLS calendar lists the July 2026 Employment Situation for August 7, 2026, and the June release specifies August 7, 2026 at 8:30 a.m. ET; this is 1 day after the registered expected window end.

The outside-view base rate is gradual growth with substantial CPS noise: annual-average employment rose from 5.688 million in 2021 to 6.711 million in 2025, while June 2026 was 0.348 million above June 2025.

Level, momentum, one-off, and mechanism effects: the 6.950 million June level supplies the persistence anchor; positive multi-year growth adds 0.050 million; a judgmental small offset for weak technology hiring subtracts 0.020 million; no identified July-specific policy or classification change warrants another adjustment.

Prior/update/interval: persistence model prior = June 2026's 6.950 million; historical sample = 2021-2025 annual averages of 5.688, 6.171, 6.502, 6.386, and 6.711 million. Their successive changes are +0.483, +0.331, -0.116, and +0.325 million, giving trend sigma = 0.236 million. Because annual averages smooth the monthly CPS target, add a monthly sampling/noise proxy of 0.239 million, the absolute gap between June 2026's 6.950 million and the latest 2025 annual average of 6.711 million. Combining these independent components gives sigma = sqrt(0.236^2 + 0.239^2) = 0.336 million. Adjustments are +0.050 million for longer-run momentum and -0.020 million for soft hiring, yielding 6.950 + 0.050 - 0.020 = 6.980 million. The 80% half-width is 1.28*sigma = 1.28*0.336 = 0.430 million, implying 6.550 to 7.410 million.

Upside risk comes from faster AI-related hiring, labor-force re-entry, or favorable CPS sampling and would land above the interval if employment exceeds 7.41 million. Downside risk comes from layoffs, weak hiring, or adverse sampling and would land below the interval if employment is under 6.55 million; either outcome would be outside the interval.

Review disposition: Accepted the blocking interval critique by adding an explicit monthly sampling/noise component to the annual-change volatility estimate and widening the bounds. Retained the August 7 release-date discrepancy and labeled the weak-hiring offset as judgmental.

calibrated forecast · 80% CI
7M[6.6M · 7.4M]
Target metadata

Data point: bls.cps.employed_people_by_occupation.computer_mathematical.july_2026.first_print

More government data forecasts