Essays on Disruptions in Medicaid Coverage

Abstract

This dissertation presents three empirical papers investigating how disruptions in Medicaid coverage affect enrollees. Medicaid is unique among public insurance programs in the United States because eligibility is means tested and, in most states, enrollment in private managed care plans is mandatory. Medicaid enrollees are low-income adults, children, and seniors, as well as people with disabilities, and these populations often face barriers to navigating the complexities of Medicaid enrollment, eligibility, and managed care. Changes in personal circumstance or state and federal policies can lead Medicaid enrollees to experience unanticipated changes to their health insurance coverage. Most Medicaid enrollees receive coverage through publicly financed, privately administered Medicaid managed care plans. States contract with Medicaid managed care organizations and offer most enrollees a choice of several plans. The high dollar amount of managed care contracts has led most states to select Medicaid plans using competitive bidding. When states conduct competitive bidding to select which plans to offer, enrollees may be forced to change managed care organizations if their plan's contract is not renewed. There is a sizable literature about plan switching in private insurance, but this type of disruption has not been extensively studied in Medicaid. Chapter 1 investigates how being forced to switch Medicaid managed care plans affects health care use and continuity of care for Medicaid enrollees. In 2016, Minnesota's state government used competitive bidding to contract with a new set of Medicaid managed care organizations. More than half of enrollees in the state were forced to change plans as a result of the bidding. I use data from the Minnesota All Payer Claims Database to show that enrollees who were forced to switch plans used fewer health care services after enrolling in their new plan. Plan switching also led to increased new provider visits, which is a sign of disrupted continuity of care. The effects on health care use were large, representing 30 percent reductions across a wide range of health care services, but were concentrated among enrollees who joined a specific managed care organization. These findings suggest that while states may be able to leverage competition between managed care plans to generate financial savings, being forced to switch insurers can be disruptive for Medicaid enrollees. Disenrollment from Medicaid is a second type of disruption. Some people lose Medicaid benefits because changes in their income or circumstances make them ineligible for the program, but others are disenrolled despite remaining eligible. This can occur when enrollees do not complete the necessary steps to renew or re-certify their Medicaid eligibility. Many people who are disenrolled later return to Medicaid. Losing Medicaid may result in uninsurance, even if benefits are restored in subsequent months. Transitioning in and out of Medicaid coverage may make it difficult for some people to receive regular medical care. Chapter 2 uses ten years of Medicaid enrollment data to measure the frequency of disenrollment and coverage disruption in Minnesota. I estimate the rate at which individuals disenroll from Medicaid, the share of disenrollments that result in uninsurance, and the share of enrollees who disenroll but return to the program within twelve months, which is called churn. I use medical and pharmacy claims data to show that the adults and children who experience disruptions in coverage are a lower spending population and leverage a unique feature of the Minnesota All Payer Claims Database to show that most enrollees do not have private insurance coverage during periods outside of Medicaid. This chapter demonstrates the role that All Payer Claims Databases can play in understanding coverage transitions in the fractured American health insurance system. I discuss state and federal policies that can help streamline Medicaid renewal and enrollment, with the goal of improving retention in the program and reducing the frequency of churn. Chapter 3 expands on Chapter 2 by examining how health care spending and use patterns differ between Medicaid enrollees who churn. I show that re-enrollment is highly correlated with short-term increases in medical spending and health care use and that enrollees who churn back into Medicaid coverage have higher spending throughout the re-enrollment period. This is the first paper to use administrative data to estimate the association between Medicaid churn and spending among non-elderly adults and shows how Medicaid enrollment policies like retroactive eligibility interact with disenrollment and churn.