Three Questions About Medicaid Unwinding: What We Know and What to Expect

During the COVID public health emergency, states were prohibited from disenrolling people from Medicaid in exchange for a substantial increase in federal funding. When continuous enrollment ended in March, states began the process of reviewing eligibility for people enrolled in the program and disenrolling those who were no longer eligible or who did not complete the renewal process. Ten months into the unwinding of the Medicaid continuous enrollment provision, states have conducted renewals for roughly half of all enrollees in the program. This policy watch examines three key questions to monitor as the unwinding continues.

1. What do we know about changes in Medicaid enrollment so far during unwinding?

Overall, Medicaid enrollment has declined by nearly 10% across states since the start of unwinding, a decline of almost 10 million people; however, the national decline in Medicaid enrollment masks significant variation across states. Changes in net enrollment reflect the people who are dropped from Medicaid as well as those who newly enroll, and those who re-enroll within a short timeframe following disenrollment, also known as “churn.” Because of churn and new enrollees — and lags in reporting — the change in net Medicaid enrollment is lower than the total number of people who have been dropped through the unwinding (which is currently more than 17 million). Unlike other types of coverage, there is no open enrollment period in Medicaid, so individuals can apply for coverage at any time. KFF research shows that pre-pandemic, one in ten enrollees disenrolled and re-enrolled in less than 12 months and children and adults had higher rates of churn compared to people who qualify for Medicaid based on age or disability. Enrollment declines vary tremendously by state. Since the start of unwinding, the decline in Medicaid enrollment across states ranges from 31% in Idaho to 0.5% in Hawaii. Net enrollment declines are measured against each state’s baseline enrollment, which is enrollment in the month prior to when the state resumed disenrollments.

State policy choices may be a better predictor of variation in state enrollment changes than how far along states are in processing renewals. While some variation in enrollment declines may reflect when states resumed disenrolling people as well as differences in the pace of processing renewals, this does not go far in explaining the variation in enrollment declines across states.  For example, some states — most notably, Texas — that still have about half of renewals yet to be completed have net enrollment declines that exceed 20% or double the national average of 10% (Figure 1).  Conversely, some states, such as Oregon, are quite far along in completing renewals, yet have seen a very small drop in enrollment. The Centers for Medicare and Medicaid Services (CMS) has released guidance and identified five strategies as most impactful in protecting coverage for children: improving auto-renewal or ex parte rates, adopting unwinding waivers, partnering with managed care plans to help people eligible for Medicaid use and keep their coverage, reducing call center wait times, and supporting coverage transitions for people no longer eligible for Medicaid. States are implementing many of these strategies, often multiple strategies at the same time, as well as others, such as enhanced outreach efforts to improve renewal processes. Ultimately, how states are proceeding with renewals will likely have a substantial effect on the number of people who are dropped from Medicaid, particularly for “procedural” disenrollments, where the renewal process is not completed and there is no way to tell if the person is still eligible for Medicaid or not. Overall, about 70% of disenrollments so far have been “procedural,” but the rate varies substantially across states.

2. Where will Medicaid enrollment wind up at the end of unwinding?

It is highly uncertain what national Medicaid enrollment will be at the end of unwinding. With about half of renewals left to complete, Medicaid enrollment will likely continue to decline. However, it is unknown how many of those who are procedurally disenrolled will regain Medicaid coverage and how many new enrollees may come on to the program. During the pandemic, Medicaid enrollment increased by 23 million or 32% from pre-pandemic levels. Given that unwinding is about halfway over – judged by the share of renewals that have been conducted so far – it’s possible that Medicaid enrollment could end up at roughly the same level as before the pandemic. In some respects, this could be considered a success. The continuous enrollment provision was intended as a temporary measure to ensure people didn’t lose Medicaid coverage during a public health crisis. A return to pre-pandemic enrollment levels is not an unreasonable outcome, particularly in an environment of low unemployment. However, those pre-pandemic Medicaid enrollment levels are not necessarily a good baseline for measuring success. Roughly one-quarter of people who are uninsured are eligible for Medicaid and not enrolled, suggesting that Medicaid was not reaching everyone who could be helped by the program before the pandemic. And, while there was churn prior to the pandemic and some churn is to be expected, rates may have been too high, resulting in eligible people losing coverage for short periods of time.

There is likely to be a lot of variation in where individual state enrollment lands relative to pre-pandemic due to unwinding and other policy changes. Some states are taking advantage of the unwinding to put in place policies that stabilize or increase Medicaid enrollment, and there is reason to believe that those states will end up with higher enrollment than pre-pandemic. However, other states are likely to see enrollment fall below pre-pandemic levels, resulting in more people uninsured than before the public health emergency. In addition, two states, South Dakota and North Carolina, implemented the Medicaid expansion during the unwinding (in July and December 2023, respectively), which should mitigate enrollment declines in these states. Other policies, such as mandatory 12-month continuous eligibility for children, optional 12-month postpartum coverage and broad state interest in pursuing multi-year continuous eligibility for children or other continuous coverage for adults, may help to reduce churn.

3. What will happen to coverage more broadly?

While changes in Medicaid enrollment are important, those numbers will matter less than what happens with the number of people who are uninsured.  The number of people without insurance and the uninsured rate dropped to record lows because of pandemic-era coverage policies, including the continuous enrollment provision in Medicaid and enhanced premium subsidies in the Affordable Care Act (ACA) Marketplaces. Unwinding will likely contribute to increases in the number of people who are uninsured and in the uninsured rate. Changes in the uninsured will depend on whether individuals who are no longer eligible and are disenrolled from Medicaid transition to other coverage, including employer plans and the Marketplace. Recent data show that Marketplace signups have reached 21.3 million people, exceeding last year’s record high by another 5 million people. Medicaid unwinding is only one factor contributing to that growth, and a relatively small share of people disenrolled from Medicaid are transitioning to Marketplace or Basic Health Plan coverage. Federal survey data that can show changes in employer coverage is lagged, so it will be quite some time until we understand the full picture of coverage trends. In addition, federal surveys are based on self-reported health coverage, which is subject to error, particularly in an environment like now, when so many transitions are occurring. As data become available, it will be important to assess how coverage changes vary across states and populations and how those changes affect children, low-income people, people of color, people with disabilities, and other groups that disproportionately rely on Medicaid.