Brief: attack the inquiry's conclusions as hard as the evidence allows. Each attack gets a response; where the attack lands, the conclusion is amended, not defended.
Attack 1: "Workforce binds" may be an artifact of studying underpaying programs
Every case cited (Canada, Quebec, Germany, UK) set public wage grids or funded rates low, then discovered shortage. A US design paying parity from day one has no clean precedent among the failures — the finding may be "underpriced labor binds," not "labor binds." Response: partially lands. The DC PEF shows parity-level funding fixes retention quickly, supporting the attack. But the FTE model shows the required flow (~800k hires/yr at ramp, 2.8M steady-state vs 1.05M today) is a labor-market-scale movement no wage level has been shown to produce on schedule. Amendment: restate the finding as "workforce binds at any historically attempted wage; at parity wages the binding constraint becomes flow/pipeline, which is untested at scale." H4.2 call unchanged; confidence in "money can't fix it" language reduced — money at parity is genuinely untested.
Attack 2: The ARPA price finding is a pandemic artifact from an interested source
The CEA (White House) analyzed its own administration's program during pandemic-distorted price dynamics. Cost declines during 2021–23 may reflect demand suppression, not grant efficacy; the two-source rule is not met by NWLC (advocacy) corroboration. Response: lands on sourcing, not direction. The expiry-side evidence (access deterioration, program losses accelerating post-cliff) is harder to attribute to pandemic dynamics in 2024. But the claim "supply-side grants suppress prices" rests heavily on CEA analysis. Amendment: downgrade H9.1's supply-side inversion from supported to supported-moderate, single-source-family; queue independent replication (academic literature via OpenAlex) in pass 3. Scorecard a6 passthrough=5 stands on the expiry evidence but is flagged.
Attack 3: Chain occupancy at 71% shows demand softness, not staffing limits
KinderCare's enrollment fell while prices rose — in served markets, at current prices, demand is the margin. This cuts against "shortage everywhere" framing and against subsidy urgency in metro markets. Response: lands as segmentation. Both are true in different geographies: deserts (rural, 70%+) are supply-starved; chain-served metros show price-rationed demand with slack. This strengthens the case against uniform national instruments and for the geography-differentiated design (ws05, H5.1) — but the inquiry should stop implying universal physical shortage. Amendment applied to framing.
Attack 4: Supply-first is politically fatal, and the scorecard hides it
Voters see nothing for 3–5 years; the durability score (2) is buried among 14 dimensions. GA/OK went demand-first universal and survived — the actual US successes did the opposite of the recommendation. Response: partially lands. GA/OK delivered visible universal seats fast at modest quality — that is a demand-first success at the 4-year-old tier, where capacity existed. The recommendation's supply-first logic applies to 0–2 and compensation, where capacity does not exist. Amendment: ws16 sequence re-stated as "visible-first where capacity exists (school-age, pre-K), supply-first where it doesn't (0–2)" — which the segmentation already implied but the sequencing draft under-stressed. The pure supply-first architecture's political fragility is real and now explicit.
Attack 5: The stable top-4 reflects correlated scoring, not evidence
All four winners embody the analyst's pass-1 conclusions (supply-side, federal, direct). If the scores encode the priors, stability across weightings measures internal consistency, not truth. Response: cannot be fully rebutted from inside. This is the deviation-#2/#6 problem: single-analyst scoring. Mitigations in place (anchored scales authored first, cited cells, sample re-score) reduce but do not eliminate it. Standing requirement: independent re-score by a second reviewer before the scorecard is presented as a conclusion rather than a draft. The result should be stated as "the evidence as synthesized here ranks…" — not "the answer is."
Net effect on conclusions
Directional conclusions survive; three confidence downgrades applied (parity-wage untested at scale; ARPA price effect single-source-family; metro vs. desert framing). The segmented, sequenced recommendation survives all five attacks and is strengthened by attacks 3 and 4.