Catastrophic Health Expenditure and Health Insurance: Evidence from COVID-19

Abstract

This paper examines catastrophic health expenditure (CHE) by household insurance type in the United States before, during, and after the COVID-19 pandemic, using Medical Expenditure Panel Survey data from 2016 to 2023. Prior research documents differences in out-of-pocket spending and financial burden across insurance types, but the U.S. experience during COVID remains underinvestigated. This paper addresses whether COVID-19 increased households’ exposure to CHE, and whether this risk differed by insurance type. Household-level linear probability models and logit regressions are estimated to assess whether insurance type is associated with differential CHE risk at the 10% and 25% of household income thresholds, and difference-in-difference models are constructed to test whether these effects changed during the COVID period (2020) and the post-COVID recovery period (2021-2023). Education-stratified models are estimated to examine heterogeneity across household socioeconomic characteristics. Insurance type is found to be a strong and heterogeneous predictor of CHE: Medicaid provides the most consistent protection at both thresholds and across education levels, Medicare provides significant protection at the severe 25% threshold, private non-group insurance is associated with elevated CHE risk at the moderate 10% threshold, and private group insurance is statistically indistinguishable from uninsured status at both thresholds. No insurance type experienced a statistically significant differential change in CHE probability during the acute pandemic period, consistent with the temporary suppression of healthcare utilization and the equalization of cost-sharing exposure through federal relief policies. In contrast, insurance type effects on CHE diverged significantly in the post-COVID period, with the divergence distributed across all insurance types and concentrated among higher-educated households. Analysis at the household level reveals that uniform coverage across all household members determines financial protection for Medicaid and Medicare households, a finding not captured by individual-level frameworks. These results suggest that individual enrollment in these programs does not guarantee household-level security, with implications for outreach and insurance design.