Date: 2026-08-03.
The natural experiment ran: ARPA stabilization as supply-side evidence
The $24B stabilization program and its expiry constitute the cleanest US evidence on how public money moves this market (CEA analysis):
- During: costs fell while funds flowed; 220k providers sustained
- After: prices resumed rising; parents lacking access 17.7% → 22.2% in two quarters; provider net losses accelerated; effects measurably smaller in the 11 states + DC that self-funded stopgaps (NWLC)
Refinement of H9.1: pass-through is instrument-dependent. Supply-side operating grants suppressed prices (money reached the cost side); Australia's demand-side subsidy saw fees outrun subsidy growth (ws15). H9.1 as written ("new spending absorbed by price increases") holds for demand-side instruments and inverts for supply-side operating support. This asymmetry is the single most actionable market finding for §14: architecture 5 (demand-side allowance) carries the inflation risk; supply-side instruments demonstrably do not, at $24B scale.
- Chain occupancy at ~71% (ws12) adds a wrinkle: short-run slack exists in existing centers in the markets chains serve — demand subsidies there would fill seats before raising prices; in desert/rural markets with no slack, they can only raise prices. Pass-through is also geography-dependent.
Feeds: §14 scorecard (pass-through row now has evidence directionality); §16 (stabilization-style operating grants are the proven fast instrument).