How do estimates of ACA improper payments break down between fraud, error, and eligibility documentation mismatches?

1. What the headline numbers mean: billions estimated, but different methods produce different totals

Congressional staff and some outside researchers present the largest headline totals—up to roughly $27 billion per year in improper APTC payments—based on extrapolations tying millions of “improper” enrollees to subsidy overpayment, while Paragon and related analyses place the plausible range of APTC improper payments roughly between $15 billion and $25 billion and estimate 4–6.4 million people receiving more than they are owed ^{[2] [1] [5]}.

2. Fraud: proven vulnerabilities, but no consensus on its dollar share

Investigations and covert tests have shown the system can be gamed—GAO-backed claims cited by Ways and Means say fake identities with bogus or never‑issued Social Security numbers were accepted and nearly all remained subsidized in follow‑up checks—evidence that deliberate fraud is possible and has occurred ^[2]. That said, government consolidated improper‑payment reporting and watchdog commentary emphasize that the bulk of agency improper‑payment estimates are not classified as fraud and that “very little fraud” appears in aggregation of improper‑payment totals, meaning the fraction clearly attributable to intentional deception remains disputed and not well quantified in public data ^{[3] [6]}.

3. Administrative error and eligibility mismatches: the bulk of measurable improper payments

Federal reporting shows improper payments across programs are driven primarily by overpayments and documentation problems rather than adjudicated fraud, and HHS and GAO note that many Marketplace improper payment estimates arise from missing or inconsistent eligibility documentation and failures to reconcile tax returns with advance credits—examples include assertions that $21 billion in subsidies lacked evidence of tax reconciliation in 2023—pointing to paperwork, data‑matching gaps, and process failures as major contributors ^{[3] [2]}.

4. The “innocent misestimate” phenomenon: income estimates, reconciliations, and policy design

KFF and other program analysts highlight that many consumers are beneficiaries of a system that relies on self‑reported projected income: differences between estimated and actual annual income can make an honest enrollee appear “improperly enrolled” or trigger improper payment classification even absent intent to deceive, and policy changes expanding subsidies increased enrollments and the scope for such mismatches ^{[4] [7]}.

5. Rulemaking, proposed fixes, and contested remedies

CMS’s 2025 Marketplace Integrity rule projects savings—up to $12 billion in 2026—by tightening enrollment integrity and administrative checks, indicating regulators view procedural tightening as a lever to reduce improper payments; critics, however, warn stricter paperwork and verification will also raise barriers for eligible low‑income enrollees and that different analysts (and political actors) emphasize either fraud or program complexity depending on policy goals ^{[8] [4]}.

6. Bottom line and limits of the evidence

The best current reading from the sources is that tens of billions in improper APTC payments likely exist annually (commonly cited $15–$27 billion), but public data do not provide a definitive split: documented system vulnerabilities and covert tests show fraud is real and potentially large in localized instances, while GAO and program auditors indicate most improper‑payment estimates across federal programs stem from overpayments and documentation/eligibility mismatches rather than adjudicated fraud—therefore any precise percentage breakdown between fraud, honest error, and documentation mismatches cannot be confidently asserted from available public reports ^{[2] [1] [3] [6] [4]}.