Fragmented Insurance and Billing Frictions: Understanding Denied Health Insurance Claims

• In Event: Administrative Frictions in US Health Insurance: Evidence on Coverage Loss, Claim Denials, and Eligibility Churn

Abstract

Billing and insurance-related activities are a major contributor to the high administrative costs of the US health care system, with estimates suggesting administrative costs represent at least 15 percent of total US health care spending. Yet the economics of claim denials—one of the most visible and controversial manifestations of these billing frictions—remains poorly understood. This paper provides new evidence on the nature and origins of denied health insurance claims, arguing that the prevalence of denials presents an economic puzzle and studying the role of insurance fragmentation for why they persist in equilibrium.

Using detailed administrative data from the Massachusetts All-Payer Claims Database covering 2013–2020, we document three striking facts about claim denials across nearly the entire private insurance landscape. First, denials are widespread and persistent: roughly 13-15 percent of claim lines are denied across all major market segments, including employer-sponsored insurance, the individual market, Medicaid, and Medicare Advantage. Second, contrary to a simple cost-containment narrative, denials are substantially more common for low-price and routine services than for expensive procedures and are not disproportionately targeted at sicker enrollees. Third, many denials are predictable ex ante: approximately 28 percent of denied claim lines involve insurer-procedure combinations where the insurer denies over 80 percent of such claims, a pattern that is difficult to rationalize under standard models of provider behavior.

To explain these patterns, we develop a model of insurer billing and show that a standard single-insurer auditing framework cannot generate the observed patterns. However, a model with fragmented insurance— multiple insurers with varying coverage rules, combined with providers’ limited ability to tailor billing strategies across payers—can match all three empirical facts.

We then provide three pieces of evidence that insurance fragmentation drives denials: First, fragmented enrollment produces sharp spikes in denial rates when individuals switch insurers, with denials rising approximately 5 percentage points at the time of the switch (a roughly 35 percent increase) driven largely by claims submitted to the wrong insurer. Second, fragmented rules produce strongly bimodal denial patterns across insurers for many routine procedures, with some insurers denying nearly all claims while others rarely do; these bimodal procedures account for at least one-third of all denied claims. Finally, we show that denial rates are lower for providers with greater billing experience and for vertically integrated provider–insurer pairs, consistent with coordination and learning partially mitigating the costs of fragmentation.

Together, our results highlight how fragmented insurance systems and heterogeneous coverage rules generate substantial administrative inefficiencies, with implications for policies aimed at reducing the administrative burden of US health care.

Author

• Mark Shepard, Harvard University
  Presenting Author