Association Between Having an Automatic Reenrollment Option and Reenrollment in the Health Insurance Marketplaces

Of the 11.4 million US health insurance marketplace enrollees in 2019, 3.4 million were automatically reenrolled based on their marketplace coverage in 2018.¹ Marketplace enrollees are automatically reenrolled in their current health plan the following year unless they actively change their enrollment status by discontinuing their coverage or selecting a new plan. Enrollees who actively select a plan have been reported to make better plan choices²; however, requiring enrollees to make a plan selection each year may be associated with their becoming uninsured.³ In January 2019, the Centers for Medicare & Medicaid Services requested public comments on eliminating automatic reenrollment for marketplace enrollees.⁴ While evidence suggests that administrative barriers to reenrollment are associated with reductions in Medicaid coverage,⁵ it is unknown whether elimination of automatic reenrollment is associated with decreases in reenrollment in the marketplaces.

We obtained 2014-2017 individual enrollment data from California’s marketplace, Covered California, through a public records act request.⁶ These data identify whether households had the option to automatically reenroll in Covered California. Households enrolled as of December 31 in a given year were able to automatically reenroll in their plan or a similar plan in the following year unless their insurer exited Covered California. Two insurers exited Covered California during the study period. Contra Costa exited Contra Costa County in 2015. United HealthCare exited other counties in 2017 (Figure). These exits divided Covered California households into groups that could automatically reenroll and groups that could not automatically reenroll. Our sample consists of 123 244 households in geographic areas and years that experienced insurer exit (rating area 5 in 2015; rating areas 1, 9, and 11-13 in 2017). This study was deemed exempt from review and approval by the University of Pittsburgh institutional review board.

We used multivariate linear regression to examine the association between household reenrollment and whether the household could automatically reenroll and adjusted for household characteristics, including the age of the oldest household member, household size, whether the household received a premium tax credit subsidy, the postsubsidy premium of the lowest-cost available plan, and indicators for geographic areas and years. We clustered SEs by geographic areas using the wild cluster bootstrap method to address the small number of clusters. Analyses were conducted in Stata SE 15 statistical package (StataCorp LLC). Statistical significance was defined as a 2-sided P < .05.

Of the 781 households (0.63%) that could not automatically reenroll in Covered California because of insurer exit, unadjusted and adjusted reenrollment rates were 21.4% and 21.5%, respectively (Table). Both the unadjusted and adjusted reenrollment rates among the 122 463 households with the option to automatically reenroll were 51.2%. Losing the option to automatically reenroll was associated with a 30-percentage point decrease in enrollment both with adjusting for household characteristics (95% CI, 9.4%-52.0%; P < .001) and without (95% CI, 14.2%-46.8%; P < .001).

Elimination of automatic reenrollment would likely be associated with decreases in the number of enrollees who remain insured through the marketplaces. As an opt-out policy³ similar to that used in other health insurance markets such as Medicaid,² automatic reenrollment may be associated with increases in continuity of coverage in the marketplaces by reducing administrative barriers to reenrollment.

Although we found losing an automatic reenrollment option was associated with decreases in reenrollment, this association requires further study. The group that lost the automatic reenrollment option was relatively small. Households with different demographics or different experiences with insurers may have behaved differently if they had lost the option to automatically reenroll. Losing automatic reenrollment because of policy changes rather than insurer exit also may be associated with households behaving differently. Given the magnitude of our findings, it is critical that future studies continue investigating the association between automatic reenrollment and continuity of coverage.

Accepted for Publication: July 3, 2019.

Corresponding Author: Coleman Drake, PhD, Graduate School of Public Health, Department of Health Policy and Management, University of Pittsburgh, 130 DeSoto St, Pittsburgh, PA 15261 (cdrake@pitt.edu).

Published Online: September 23, 2019. doi:10.1001/jamainternmed.2019.3717

Author Contributions: Dr Drake had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Both authors.

Acquisition, analysis, or interpretation of data: Drake.

Drafting of the manuscript: Both authors.

Critical revision of the manuscript for important intellectual content: Both authors.

Statistical analysis: Drake.

Conflict of Interest Disclosures: None reported.

Additional Contributions: Eric Roberts, PhD (University of Pittsburgh Graduate School of Public Health), and Adam Biener, PhD (Lafayette College), provided feedback on an earlier draft of this article. Pearl Nielsen, MA (Carnegie Mellon Dietrich College of Humanities and Social Sciences), provided editorial assistance. No compensation was provided for their roles.