thesisPrototype

Government data

Colorado SSI recipients, July 2026

Social Security Administration SSI Monthly Statistics Table 4 number of recipients, all federally administered payments, Colorado, July 2026, non-seasonally adjusted, first print

Forecast

66k
66k–67k
80% prediction interval
gpt-5.5Run record ↗

Cumulative probability

Chance that the outcome is at or below each value.

0%25%50%75%100%66k66k67kForecast value

Shaded band: 80% interval (66k–67k). Dashed line: point estimate (66k).

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

Observed outcomeinside 80% interval
actual
66k
forecast
66k with 80% interval [66k, 67k]
error
-0k · absolute 0k
cdf score
CRPS 0.062 · PIT 0.31
source
ssa_ssi_monthly SSI Monthly Statistics, Table 4 (All Federally Administered Payments), Colorado row, Total

First print for 2026-07 captured from the edition's own page https://www.ssa.gov/policy/docs/statcomps/ssi_monthly/2026-07/table04.html through a declared headless-browser transport (ssa.gov refuses non-browser clients); the page title and table caption name the edition, the parsed row and column are label-anchored, the table's own row/column sums reproduce the published totals, and the prior edition's anchor value was re-read from its page at capture time; no Wayback Machine capture of the page exists to corroborate.

Analysis

§Forecast for Colorado SSI recipients in July 2026

Framing and exact resolver: this is SSA SSI Monthly Statistics Table 4, All Federally Administered Payments, row Colorado, Total column, for July 2026, non-seasonally adjusted and first print. The resolution page should be the stable July 2026 Table 4 URL; Table 4 is the exact variant for the target.

▸ Reported tool use: official.lookupmodel report
Opened SSA Publishing Schedule and the SSA statistics archive for SSI Monthly Statistics timing.
↳ Fetched timing evidence: SSA Publishing Schedule lists SSI Monthly Statistics frequency as Monthly; SSA statistics archive lists SSI Monthly Statistics last released June 2026 and next expected July 2026, so the July 2026 data table is expected on month-level timing with by-date 2026-08-31.
▸ Reported tool use: official.lookupmodel report
Opened SSA SSI Monthly Statistics June 2026 Table 4 for the latest same-variant Colorado anchor.
↳ Fetched June 2026 Table 4 row Colorado Total recipients of 66,417; components were aged 12,130 and blind/disabled 54,287.
▸ Reported tool use: official.lookupmodel report
Opened SSA SSI Monthly Statistics May and April 2026 Table 4 pages for recent same-variant Colorado history.
↳ Fetched Colorado all federally administered totals: May 2026 66,403 recipients and April 2026 66,539 recipients; May components were aged 12,120 and blind/disabled 54,283.
▸ Reported tool use: official.lookupmodel report
Opened SSA SSI Monthly Statistics February and January 2026 Table 4 pages for earlier 2026 same-variant history.
↳ Fetched Colorado all federally administered totals: February 2026 66,914 recipients and January 2026 67,018 recipients; February components were aged 12,188 and blind/disabled 54,726.
▸ Reported tool use: official.lookupmodel report
Used the official December 2025 Table 4 page as the year-turn anchor for Colorado.
↳ Fetched Colorado December 2025 total of 67,134 recipients, with aged 12,264 and blind/disabled 54,870.

Base rate/reference class: the reference class is recent first-print SSA SSI Monthly Statistics Colorado total-recipient levels in the same all-federally-administered, non-seasonally-adjusted Table 4 variant. The base rate is a stable mid-66-thousand to high-67-thousand Colorado caseload, with the latest available count 66.417 thousand in June.

Level, momentum, one-off, and policy mechanism: the level anchor is June 2026 at 66.417 thousand. Momentum is negative over the first half of 2026, but May-to-June was nearly flat, so the July update is kept close to persistence. I do not add a one-off shock; eligibility, terminations, aging, deaths, and administrative processing should move the count gradually at monthly frequency.

Prior/update/interval: persistence prior is June 2026 at 66.417 thousand and is the selected time-series prior; I do not run a separate model candidate because the fetched adjacent monthly sample is short and irregular. Historical sample for dispersion uses fetched same-variant adjacent changes Dec->Jan, Jan->Feb, Apr->May, and May->Jun: -0.116, -0.104, -0.136, and +0.014 thousand. Adjustment components are -0.036 thousand from a light 30% weight on the January-to-June average decline and 70% weight on latest-month persistence, +0.000 for one-off shocks, and +0.000 for identifiable policy mechanism shifts, giving point 66.417 - 0.036 = 66.381 thousand. sigma = sqrt((0.116^2 + 0.104^2 + 0.136^2 + 0.014^2) / 4) = 0.104 thousand. The raw 80% half-width is roughly 1.28*sigma = 1.28*0.104 = 0.133 thousand; I widen to 0.200 thousand, 1.50x the raw half-width, because the fast-run sample is short and skips March. Final implied bounds are 66.381 - 0.200 = 66.181 and 66.381 + 0.200 = 66.581.

Counter-considerations: upside risk would be a July administrative rebound or delayed entries in aged recipients, which would land above the interval if Colorado exceeded 66.581 thousand. Downside risk would be unusually heavy terminations, disability-recipient attrition, or returned-check/eligibility cleanup effects reflected before first publication, which would land below the interval if Colorado came in under 66.181 thousand. An outside the interval result would imply a larger processing swing than the recent adjacent-month reference class.

Review disposition: accepted the June Table 4 provenance correction, renamed the June anchor consistently, dropped the non-resolving Table 10/Table 15 context, explicitly selected persistence as the time-series prior, and shrank the downward update closer to June persistence because May-to-June was nearly flat.

Key drivers

  • June 2026 same-variant Table 4 Colorado count was 66.417 thousand before the target month
  • Recent Colorado SSI recipient counts drifted lower from January to June
  • May-to-June momentum was nearly flat after earlier declines
  • Persistence dominates because the latest monthly move was small
  • Whole-recipient SSA counts are converted to thousands

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
Social Security Administration SSI Monthly Statistics
Resolved
August 23, 2026
Resolution rule
Resolve to the first published July 2026 value in SSA SSI Monthly Statistics Table 4, row Colorado, column Total number of recipients for all federally administered payments. SSA publishes whole recipient counts; convert to thousands by dividing by 1,000 and report to the same 0.001-thousand precision implied by whole recipients. Use the first official print only and ignore later revisions after the first July 2026 table is posted. SSA's public schedule and archive give month-level monthly timing rather than a specific day; 2026-08-31 is the latest expected by-date for the July 2026 monthly publication.
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 · v2.2.0

pre-submit review · completed

The draft is close, but publication should fix a provenance inconsistency around the latest June anchor and tighten the model-prior/update explanation.

  • blocking resolver: The resolver correctly targets July 2026 SSA SSI Monthly Statistics Table 4, but the latest June evidence/sourceContext cites Table 10/Table 15 even though the same-variant June Table 4 page exists and contains Colorado total recipients of 66,417.
  • warning model_prior: The trace uses last-print persistence as the prior, but does not clearly state whether any formal time-series/model candidate was run or ruled out for this repeated monthly series.
  • warning update: The -0.050 thousand downward adjustment is plausible but only loosely tied to current evidence, since May-to-June was nearly flat and the rationale is residual first-half momentum.
  • info optional_suggestion: If available, include March 2026 Table 4 or explain its absence so the volatility sample does not appear selectively gapped.
  • info optional_suggestion: Rename the June historicalContext label from federal-payment to same-variant Table 4 total recipients for consistency with the resolver.

disposition accepted: Review disposition: accepted the June Table 4 provenance correction, renamed the June anchor consistently, dropped the non-resolving Table 10/Table 15 context, explicitly selected persistence as the time-series prior, and shrank the downward update closer to June persistence because May-to-June was nearly flat.

disposition accepted: Review disposition: accepted the June Table 4 provenance correction, renamed the June anchor consistently, dropped the non-resolving Table 10/Table 15 context, explicitly selected persistence as the time-series prior, and shrank the downward update closer to June persistence because May-to-June was nearly flat.

disposition accepted: Review disposition: accepted the June Table 4 provenance correction, renamed the June anchor consistently, dropped the non-resolving Table 10/Table 15 context, explicitly selected persistence as the time-series prior, and shrank the downward update closer to June persistence because May-to-June was nearly flat.

disposition not applicable: Review disposition: accepted the June Table 4 provenance correction, renamed the June anchor consistently, dropped the non-resolving Table 10/Table 15 context, explicitly selected persistence as the time-series prior, and shrank the downward update closer to June persistence because May-to-June was nearly flat.

disposition not applicable: Review disposition: accepted the June Table 4 provenance correction, renamed the June anchor consistently, dropped the non-resolving Table 10/Table 15 context, explicitly selected persistence as the time-series prior, and shrank the downward update closer to June persistence because May-to-June was nearly flat.

Activity artifacts

Complete original trace

§Forecast for Colorado SSI recipients in July 2026

Framing and exact resolver: this is SSA SSI Monthly Statistics Table 4, All Federally Administered Payments, row Colorado, Total column, for July 2026, non-seasonally adjusted and first print. The resolution page should be the stable July 2026 Table 4 URL; Table 4 is the exact variant for the target.

▸ Reported tool use: official.lookupmodel report
Opened SSA Publishing Schedule and the SSA statistics archive for SSI Monthly Statistics timing.
↳ Fetched timing evidence: SSA Publishing Schedule lists SSI Monthly Statistics frequency as Monthly; SSA statistics archive lists SSI Monthly Statistics last released June 2026 and next expected July 2026, so the July 2026 data table is expected on month-level timing with by-date 2026-08-31.
▸ Reported tool use: official.lookupmodel report
Opened SSA SSI Monthly Statistics June 2026 Table 4 for the latest same-variant Colorado anchor.
↳ Fetched June 2026 Table 4 row Colorado Total recipients of 66,417; components were aged 12,130 and blind/disabled 54,287.
▸ Reported tool use: official.lookupmodel report
Opened SSA SSI Monthly Statistics May and April 2026 Table 4 pages for recent same-variant Colorado history.
↳ Fetched Colorado all federally administered totals: May 2026 66,403 recipients and April 2026 66,539 recipients; May components were aged 12,120 and blind/disabled 54,283.
▸ Reported tool use: official.lookupmodel report
Opened SSA SSI Monthly Statistics February and January 2026 Table 4 pages for earlier 2026 same-variant history.
↳ Fetched Colorado all federally administered totals: February 2026 66,914 recipients and January 2026 67,018 recipients; February components were aged 12,188 and blind/disabled 54,726.
▸ Reported tool use: official.lookupmodel report
Used the official December 2025 Table 4 page as the year-turn anchor for Colorado.
↳ Fetched Colorado December 2025 total of 67,134 recipients, with aged 12,264 and blind/disabled 54,870.

Base rate/reference class: the reference class is recent first-print SSA SSI Monthly Statistics Colorado total-recipient levels in the same all-federally-administered, non-seasonally-adjusted Table 4 variant. The base rate is a stable mid-66-thousand to high-67-thousand Colorado caseload, with the latest available count 66.417 thousand in June.

Level, momentum, one-off, and policy mechanism: the level anchor is June 2026 at 66.417 thousand. Momentum is negative over the first half of 2026, but May-to-June was nearly flat, so the July update is kept close to persistence. I do not add a one-off shock; eligibility, terminations, aging, deaths, and administrative processing should move the count gradually at monthly frequency.

Prior/update/interval: persistence prior is June 2026 at 66.417 thousand and is the selected time-series prior; I do not run a separate model candidate because the fetched adjacent monthly sample is short and irregular. Historical sample for dispersion uses fetched same-variant adjacent changes Dec->Jan, Jan->Feb, Apr->May, and May->Jun: -0.116, -0.104, -0.136, and +0.014 thousand. Adjustment components are -0.036 thousand from a light 30% weight on the January-to-June average decline and 70% weight on latest-month persistence, +0.000 for one-off shocks, and +0.000 for identifiable policy mechanism shifts, giving point 66.417 - 0.036 = 66.381 thousand. sigma = sqrt((0.116^2 + 0.104^2 + 0.136^2 + 0.014^2) / 4) = 0.104 thousand. The raw 80% half-width is roughly 1.28*sigma = 1.28*0.104 = 0.133 thousand; I widen to 0.200 thousand, 1.50x the raw half-width, because the fast-run sample is short and skips March. Final implied bounds are 66.381 - 0.200 = 66.181 and 66.381 + 0.200 = 66.581.

Counter-considerations: upside risk would be a July administrative rebound or delayed entries in aged recipients, which would land above the interval if Colorado exceeded 66.581 thousand. Downside risk would be unusually heavy terminations, disability-recipient attrition, or returned-check/eligibility cleanup effects reflected before first publication, which would land below the interval if Colorado came in under 66.181 thousand. An outside the interval result would imply a larger processing swing than the recent adjacent-month reference class.

Review disposition: accepted the June Table 4 provenance correction, renamed the June anchor consistently, dropped the non-resolving Table 10/Table 15 context, explicitly selected persistence as the time-series prior, and shrank the downward update closer to June persistence because May-to-June was nearly flat.

calibrated forecast · 80% CI
66k[66k · 67k]
Target metadata

Data point: ssa.ssi.recipients.colorado.2026-07.first_print

More government data forecasts