Assessment of Churn in Coverage Among California’s Health Insurance Marketplace Enrollees

Abstract

Individual health insurance marketplaces established through the Affordable Care Act of 2010 (ACA) fill a critical gap for those who lack other coverage options. The high degree of coverage turnover, or churn, has raised concerns about affordability and strategic behavior on the part of individuals to sign up only when they need care. To assess the role of California's ACA marketplaces in the broader health care landscape by understanding enrollment tenure and churn. This cross-sectional study analyzed survey data from a representative, probability-based sample of enrollees in California's ACA marketplace, Covered California, collected immediately after the open enrollment periods in 2018, 2019, and 2021. Administrative data from Covered California from January 1, 2014, to December 31, 2021, were used to analyze marketplace tenure. Survey data included 9571 heads of households aged 18 to 64 years who were newly enrolled or had recently terminated their plan, directly drawn from Covered California's administrative records. Administrative data included individuals up to age 65 years who enrolled in the marketplace during 2014 to 2021. New enrollment in or termination from health care coverage through California's ACA marketplace. Enrollment tenure in California's ACA marketplace, sources of coverage prior to enrolling and after terminating coverage, and demographic or plan characteristics associated with the decision to go uninsured. Median (IQR) length of coverage among 5.4 million enrollees (mean [SD] age, 38 [16] years; 17% Asian American/Native Hawaiian or other Pacific Islander, 2.5% Black or African American, 23% Latino [response options were Hispanic, Spanish, or Latino origin], 29% White, 7.5% categorized as other [including American Indian/Alaskan Native, multiple races, and other], and 21% of unknown race or ethnicity) was 14 (6-35) months, and 41% to 46% of enrollees disenrolled within 1 year, with substantial variation by subgroups. Despite this churn, only 14% (95% CI, 12%-15%) of 6474 terminating members surveyed across 3 years (2018, 2019, and 2021) reported being uninsured after leaving the marketplace, with the rest moving to job-based coverage or Medicaid. Most of those surveyed (mean [SE] percentage, 56% [0.016] individuals) reported having had employer-sponsored insurance or Medicaid prior to enrolling in the marketplace. Among subsidized renewal candidates, Latino candidates were 1.5 percentage points (95% CI, 0.8-2.3 percentage points) more likely to go uninsured compared with White candidates, and those with no expected physician visits in the coming year were 4.8 percentage points (95% CI, 2.4-7.2 percentage points) more likely to go uninsured vs those who expected physician visits. The results of this cross-sectional study of coverage churn found that ACA marketplaces served 2 distinct types of individuals, long-term enrollees but, more often, individuals with short-term-coverage needs due to a change in eligibility for other insurance. These results suggest that marketplaces are smoothing coverage disruptions and that policies to reduce gaps in coverage should be designed with this in mind.