Navigating Medicaid: Experimental Evidence on Administrative Burden and Coverage Loss

Millions of people lose access to safety-net programs each year because they do not complete the required administrative processes. In a large-scale field experiment, we randomized outreach offering personalized assistance during the Medicaid renewal window to identify the impact of transaction costs on coverage loss. Our intervention – which provided information and assistance – simplified the renewal process without affecting eligibility, ensuring that avoidable coverage loss among eligible beneficiaries accounts for any effects we find. Sending pre-recorded calls offering free one-on-one assistance from health insurance navigators increased Medicaid renewal by 1 percentage point, a 1.5% increase relative to the control arm. However, the phone number on file was likely no longer valid for at least one-fifth of beneficiaries, suggesting some people did not receive their calls. Receiving the call increased successful renewals the most for tribal members (by 8 percentage points or 13%) and children (by 3 percentage points or 4%); receiving the call also increased renewals for people with income below the sample median (by 2 percentage points or 4%) and people with chronic disease (by 3 percentage points or 4%). Together, these data suggest that some eligible people, including those who need care, lose Medicaid because they struggle with the required administrative processes. The findings do not support the typical rationale for allowing barriers or costs to enrollment—namely, that they efficiently discourage participation by those who do not need or value the benefits.