Date: 2026-08-03. Partial; preference-by-subgroup and summer-gap quantification pending.
Findings
- Revealed preference by age band (NSECE 2019): under-3s most commonly in individual unpaid care (26.5%); 3–5s most commonly in center-based care (42.1%) (ACF/NSECE). The infant/toddler home-care preference the protocol flagged is confirmed in usage data — a center-only buildout misreads the 0–3 market. Directly relevant to architecture 11 (caregiver-choice allowance).
- ~40% of households with under-5s use no regular nonparental care (NSECE: ~60% use CCEE regularly). Universal provision is a behavioral offer to this group, not a subsidy to existing behavior — take-up assumptions drive everything (H2.1's other face).
- Cost-burden distribution: 73% of sub-poverty households pay $0 out of pocket (Head Start, subsidies, kin care); the near-poor (100–200% FPL) face the worst burdens, >20% of income (NSECE cost-burden snapshot). The binding affordability problem is concentrated just above current eligibility lines — an argument that sliding-scale designs beat cliff designs, and a warning that "families pay $13k" averages hide a bimodal reality.
- Access deterioration is measurable and recent: parents reporting lack of access rose 17.7% → 22.2% (fall 2023 → spring 2024) after the stabilization cliff (CEA).
- School-age unmet demand quantified (pass 3): America After 3PM 2025 — unmet afterschool demand at an all-time high; in the middle-school segment, 6.7M students' parents want a program and 5.2M (nearly 4 in 5) can't get one; summer is when families report the need most acutely (Afterschool Alliance, Wallace). Source is advocacy-collected survey data (tagged per §0.1) but it is the only national instrument for this segment — and it lands on the same side as the $1.3B-vs-32.6M-children spending mismatch (H2.2).
Feeds: §14 — architecture scoring on the 0–3 preference reality; copay design (smooth, no cliffs); take-up scenarios for §3 sensitivity.