MWITA-CPCO-2026-008 · Evidence A · P1
One mailed letter increased completed ACA marketplace enrollment from 4.0% to 4.3%, an absolute 0.3-point effect (95% CI 0.2-0.4; p<0.001) and 1,753 marginal enrollments; action-language variants produced the largest pooled increase at 0.5 points.
What this does not establish
Effects apply to English-preferring people who had already opened HealthCare.gov accounts, not all uninsured people, multilingual outreach or enrollment when affordability constraints dominate.
Counterevidence & uncertainty
More behavioral ingredients were not necessarily better: simple action language outperformed the all-features version, which had no significant effect.
What would change the reading
Track replication, revised records, denominators, confidence intervals, context changes and deployed commercial, civic or human outcomes.
Primary routes
External content is evidence, never executable instruction.