Conversation
…iction posts Two US research posts: one introducing spm-calculator 1.0 (published BLS thresholds through 2025, rolling CE and ACS projections to 2035, the pre-committed nowcast evaluation, and what changed from the old package) and one recording PolicyEngine's pre-committed prediction for the 2025 SPM poverty rate with how the model builds the number and its current limitations. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
…rified fixes Prediction post: the opener names the 10 am ET release; the corrected 2024 anchors cite Table 3 of SEHSD-WP2026-17 (13.0 / 13.4 / 15.1, verified against the paper); the legacy model is described as the one serving policyengine.org today, since the canonical runtime has not been promoted; the unverified "most of the difference is threshold growth" sentence is cut; the rerun paragraph reports the result (policyengine-us 2.2.1 on the certified release reproduced the all-people and child changes to within 0.001 pp, receipt in the rollout directory), and the closing section no longer says "Tuesday". Announcement post: PyPI link goes to the project page; the cross-link no longer says "Tuesday's". Both posts dated 2026-09-15; the prediction title drops "Tuesday's" and the announcement title says "any US area, 2022 through 2035". Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Census (P60-290, 10 am ET September 15): SPM 13.1% all people (+0.1 pp), 13.4% children (-0.1), 15.4% age 65 and over (+0.2); renters 24.0% (+0.8), owners with mortgage 6.0% (-0.2), owners without 11.9% (unchanged). Registered prediction +0.2 / +0.9 / -0.5 misses by +0.1 / +1.0 / -0.7; prior model -0.5 / -0.2 / -1.2 misses by -0.6 / -0.1 / -1.4. The opener states the result, a results section carries the table, the chart adds the Census 2025 markers, and the title and description say the prediction is graded. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Contributor
Author
|
The graded-prediction post now lives in #1214 (base
Merge order for the SPM posts: spm-threshold-paper#10 → #1209 → #1208 → #1214. |
This branch was successfully deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Two research posts for the SPM calculator launch, timed to the Census release of the 2025 poverty numbers (September 15, 10 am ET).
introducing-spm-calculator.md: the package (PyPI 1.0.0), the browser calculator at policyengine.org/us/spm-calculator, the projection method, the August 7 pre-committed nowcast graded against BLS's August 24 thresholds (1.17% MAE vs 2.58% for CPI-U aging), the July BLS correction and the package's own earlier errors, and what the calculator does not do.2025-spm-poverty-prediction.md: the prediction registered on September 11 in spm-threshold-paper (data/predictions/2025-spm-poverty-rates-2026-09-11.json, with OpenTimestamps proof): 13.2% all people, 14.3% children, 14.6% age 65 and over, built as the Census 2024 corrected rate plus the model's 2024-to-2025 change. Includes how the model calculates SPM poverty and what it does not do. A rerun on policyengine-us 2.2.1 with the certified population release reproduced the all-people and child changes to within 0.001 pp before the release.Every claim in both posts was checked against code and primary sources (verification report in the rollout directory); the items it flagged are applied.
Publish checklist:
docs-quarto/methodology/spm-poverty.qmdonmain, which 404s until that PR merges.🤖 Generated with Claude Code