GUBMENTPlain talk · policy frontier
Filings / Childcare / Sources / Scorecard Rationale & Reconciliation
GBMT-1 · Research record · No. 1

Scorecard Rationale & Reconciliation

childcare/research/scorecard/rationale.md
This is a working research document from the childcare filing, published as written — including the parts later corrected. It is the underlying record for Whitepaper No. 1, not a summary of it.

Headline result

Three designs occupy the top four under every objective weighting: public option (a8), Head Start scaled (a3), and K–12 extension (a4). The fourth seat is not stable: supply-first (a6) holds it under equal / labor-supply / child-dev weightings; caregiver-choice (a11) takes it under equity-first. Federal fallback (a9) keeps the holdout-proofing cell (federalism=5) but leaves the all-weightings top set after the independent re-score removed unevidenced above-neutral scores.

This replaces the pass-1 claim of a flip-free top-4 (a3/a6/a8/a9). Full reconciliation: rescore-log.md.

What the stable three share: supply-side or direct-provision money (ws09), standards housed inside federal spending programs where legislation is needed (ws11), and either holdout-proofing (a3/a8) or demonstrated red-state durability at the pre-K / school-age tier (a4). They are components of a segmented design, not rival universals — a4's coverage_02=1 is definitional.

Notable placements. Supply-first (a6) remains the compensation/capacity ramp and ranks top-4 under three of four weightings, dragged by durability=2 (annual-appropriations fragility). Caregiver-choice (a11) is the only architecture at 5 on both coverage_02 and preference_fit — and now enters the equity top-4. Demand-side allowance (a5) and federalized Tri-Share (a10) rank last under every weighting. The cash comparator is mid-table (burden relief without supply) — it does not rank last; that earlier claim contradicted rankings.txt.

Key cell citations (evidence tags)

Reconciliation log (independent structurally-blinded re-score, 2026-08-07)

Full board re-scored via batch API with evidence inlined and scorecard unreachable. 47 cells moved; 25 judgment splits kept.

Material headline effects:

Sample re-score (deviation #6) remains in git history; it is superseded as the independence claim.

H14.2 adjudication (per criteria)

Segmented-beats-uniform: supported, strengthened. Best per-band instruments (a4 for 5–12 and strong on 3–4; a3/a8 for holdout-proof delivery; a6+a11 blend for 0–2) each outscore uniform demand-side and employer-split designs. Formal criteria met (≥2 of 3 bands, ≥3 of 4 weightings). The re-score made the segmentation visible in the ranks rather than only in the narrative.

← All Childcare research documents Sources digest Read the whitepaper