Repository navigation
Engine-grounded SMD accounting bracket on the FY2024 QC cases - #78
Merged
Merged
Conversation
Harvested from a lane that died after writing its artifacts and before committing; the maintainer validated the artifact (5 tests pass, the Colorado row and the recoverability verdict verified) and committed with this disclosure. Recoverability: the SMD itself censors actual expenses, so the counterfactual is a bracket — Colorado between $0 and −$5.4M annual issuance, 0 to +0.04pp mechanical measured-rate change — reported beside the deployed statistical construction: two accounting answers to one lever, neither causal.
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The formula-side counterpart to the deployed simulator's SMD scenario, computed on the recorded FY2024 QC cases through the 6,081/6,081-certified administrative sequence — the number promised to the SNAP Screener/Georgetown thread.
Recoverability verdict first (the finding that shapes everything): for a case recorded at the state's standard medical deduction, actual allowable expenses are not recoverable — the SMD process itself replaces qualifying-range values with the standard (
FSMEDEXPsemantics, FY2024 techdoc p. 75; the minimodel replacement rule, p. 50). So the counterfactual is a bracket, never a point: convention (a) sets censored expenses equal to the standard (delta zero); convention (b) sets them at the $35 floor.Five of the seven parity-verified states operate an SMD (AZ, CA, CO, GA, TX; MD and NY carry zero registry amounts and are reported, not computed). Colorado: 53 claimant cases, 46 censored; the bracket runs $0 to −$5.4M in annual issuance and 0 to +0.04pp on the measured rate — mechanical, error dollars held fixed over smaller issuance, so the mechanical construction raises the rate where the deployed model's re-prediction lowered its crossing measure (−0.060pp). The memo states why the two constructions answer different questions and that neither is causal or validates the other.
Provenance disclosure: the estimation lane died after writing its artifacts and before committing; the maintainer validated (5 artifact-lock tests pass; Colorado row and recoverability verdict checked against the JSON) and committed with that disclosure in the commit message.
🤖 Generated with Claude Code