Association for Public Policy Analysis & Management

A key policy objective of the Patient Protection and Affordable Care Act (ACA) was to improve the functioning of the individual market for health insurance. The goal was to better organize the market, facilitate access to information about plan options, strengthen competition, and increase coverage affordability. The ACA introduced significant regulatory changes to the individual market. In combination with legal challenges and uncertain political support regarding the ACA’s future more broadly, insurers’ concerns have escalated. These factors have led insurers to exit local markets or entire states, fewer plan choices, and exceedingly high premium increases. Lack of robust competition in the individual market has important implications for consumers. Regardless of what federal policy changes occur, states will need to understand the impact and implications of insurer exit on their individual markets locally as they consider strategies to support market stabilization.

This study has three research objectives:

1) To understand the differences in population attributes of local markets (counties) experiencing insurer exit versus those without exit by examining demographic, economic, health status, provider market, and policy environment factors between 2015-2018;

2) To investigate the set of insurer attributes (ownership status, size, multi-state presence, and past financial performance) associated with a firm’s local market exit;

3) To assess the impact exiting insurers on local market premiums and plan choices.

Our primary data source is the 2015-2018 HIX Compare Individual Market and County Reports data. We augment these data with information from: the QHP Landscape File, the National Association of Insurance Commissioners Supplemental Health Care Exhibit, the American Community Survey, the Bureau of Labor Statistics, County Health Rankings, the Area Health Resource File, the AAP State Practice Environment, the National Conference of State Legislatures, and the Kaiser Family Foundation.

For the first two objectives, we estimate binary logistic models to examine the set of factors influencing exit at the county-year level and the insurer-county-year level between time t and t+1, respectively. To allow patterns to differ over time as insurers acquire information about local markets, we estimate separate models by year. For the third objective, we examine three outcomes: log premium of the lowest-priced silver plan in the local market, log deductible of the lowest-priced silver plan, and the mean number of unique plans offered per insurer in a geographic rating area r in state s and time t. We estimate multivariate regression models for the subset of markets that have four or fewer insurers as of 2015. From our model estimates, we test for whether and to what extent exit in time t is associated with higher premiums and deductibles as well as fewer plan options per insurer in time t+1.

Preliminary results for the first objective find that the pattern of associations with exit varies over time. However, larger population size, the number of insurers in operation, and whether the local market is in a state that expanded Medicaid are consistently negatively associated with the probability of exit in a given year. Analyses for the other objectives are in progress.