Policy One-Pager Health & Immigration Coverage Take-Up
Working Paper · 2026
Fear as a barrier to coverage: how the public-charge rule chilled Medicaid for citizen children.
A federal rule that never legally applied to U.S.-citizen children still cut their Medicaid enrollment — and the effect outlived the rule itself.
Jonathan Palisoc
Ph.D. Candidate, Health Services Org. & Policy
University of Michigan
The bottom line
When the 2019 public-charge rule made Medicaid use count against immigrants seeking green cards, fear spread far beyond the people the rule actually targeted. U.S.-citizen kids in mixed-status families — legally exempt — lost coverage anyway. The drop began with the September 2018 announcement, persisted after the rule was reversed, and deepened during the 2023–24 Medicaid “unwinding.”
−2.0 pp
Medicaid enrollment drop among low-income citizen children in mixed-status families, vs. citizen-only families. p < 0.001 · N = 1.6M
01
A fear cycle that outlived the rule.
Sep 2018 Rule announced. Coverage drops before any law changes.
Feb 2020 Rule takes effect after Supreme Court stay.
Mar 2022 Biden formally reverses the rule.
Apr 2023 Medicaid “unwinding” — effect deepens.
02
Who feels it: 5.5 million citizen kids.
Roughly 5.5 million U.S.-citizen children live with at least one undocumented household member. Their Medicaid eligibility is identical to other citizen children — but enrollment runs through parents who may avoid government systems out of fear. Each square below represents roughly 20,000 children; brick squares mark the population whose Medicaid take-up the rule plausibly chilled.
Citizen kids in mixed-status families (~5.5M)All other low-income citizen kids
03
Three findings that sharpen the story.
−2.6 pp
Chill is larger when seven states that counter-programmed with their own coverage are excluded.
p < 0.001 · drop-7 spec.
−1.3 pp
Medi-Cal drop among naturalized citizens in California — a group the rule never applied to.
p = 0.030 · spillover
+20.5 pp
Undocumented adults more likely than naturalized citizens to decline a benefit citing immigration fears.
p < 10⁻³⁰ · restricted CHIS
A
Why it matters.
Medicaid covers about half of low-income U.S. children; childhood enrollment is linked to better health, schooling and adult earnings.
The rule changed zero eligibility rules for these kids — only the perceived cost of asking for coverage.
In 2026, a new administration has signaled renewed interest in expanding public-charge determinations.
B
What we did.
1.6M citizen children under 200% FPL across 2016–2024 American Community Survey data, compared in a difference-in-differences design: kids in mixed-status vs. citizen-only families, before vs. after Sept. 2018.
A supporting analysis uses restricted-access CHIS data on 220,925 California adults — separating green-card holders from other noncitizens, a split federal surveys don’t allow.
C
Why believe the result.
No pre-trend. The two groups tracked closely before Sept. 2018 (joint test p = 0.92).
Mechanism aligns. Effects concentrate where immigration risk is highest — mixed-status families, non-expansion states, undocumented adults.
Caveats. California Medi-Cal expansions block a clean within-state estimate for adults; pooling all noncitizen parents likely understates the true effect.
→
So what?
Immigration policy is de facto health policy. Restrictive signals move millions of eligible citizens off coverage, and the fear they trigger doesn’t lift when the rule is rescinded. Future rules need an explicit chilling-effects mitigation plan.
Jonathan Palisoc · Ph.D. Candidate, Health Services Organization & Policy, University of Michigan Working paper, May 2026. Cite as: Palisoc, J. (2026). Fear as a Barrier to Coverage.