What We Do and Don’t Know from the Medicaid Unwinding Data 

By Suzanne Wikle

As part of the Medicaid “unwinding” from COVID-era protections that kept people enrolled, states are required to submit data to the Centers for Medicare and Medicaid Services (CMS) every month. 

CMS is expected to release early data from approximately 18 states later this week, but many more states are voluntarily posting several months of data. State timelines for disenrolling people vary. Some started in April, and by now almost all states have begun disenrolling people as part of the unwinding process. As more data rolls out from states, troubling trends have started to emerge, along with some unanswered questions. 

What We Know from the Data 

Significant coverage loss is happening. Data available from 33 states and the District of Columbia show that at least three million people have lost coverage since April 1, 2023. Estimates produced before unwinding began suggested that 15 million people may lose coverage during the year of unwinding.  

Procedural disenrollment rates are high. A procedural disenrollment happens when someone is disenrolled for a reason other than proof of ineligibility. This may occur when paperwork isn’t returned or was never received. It is likely that many people who lose Medicaid coverage because of procedural reasons are actually still eligible.  

According to KFF reports, for the data available so far, 74 percent of disenrollments are procedural. But we should remember two caveats when looking at the data from states. First, some states choose to prioritize people who they have reason to believe are no longer eligible for the first months of unwinding. We expect to see a higher overall percentage of disenrollments and higher procedural disenrollments in those states, as residents who know they are no longer eligible may not respond to renewal paperwork. Secondly, if a state has a higher ex parte renewal rate, it may also have higher procedural disenrollment. If the state can complete more renewals with available data, it’s likely that a greater share of remaining renewals will be ineligible; and again, those who know they are ineligible may be less likely to respond to renewal paperwork.  

Even with those caveats, procedural termination rates are disturbingly high and vary greatly across states. States with the lowest procedural denial rates still have at least 30 percent, meaning that over 30 percent of all disenrollments were for a reason other than an ineligibility determination. In some states, close to or over 90 percent of all disenrollments are procedural.  

Procedural disenrollments are worrisome because we ultimately do not know how many people disenrolled this way are still eligible. Additionally, lapses in coverage can result in missed medications, confusion about eligibility, and a lack of trust in Medicaid and other benefit programs. Lowering procedural disenrollment rates will be critical to ensure that people who are eligible for Medicaid stay enrolled when they renew. 

Ex parte renewal rates are disappointingly low. An ex parte renewal is when someone’s Medicaid eligibility is renewed based on available electronic data without the Medicaid enrollee taking any action. Medicaid regulations require that states attempt an ex parte renewal for all Medicaid cases, but we’ve learned through our work with state advocates that those regulations are not always followed. For example, advocates in Pennsylvania realized that Medicaid cases where someone was also enrolled in SNAP were automatically excluded from the ex parte process. We also know that many states are not yet attempting ex parte renewals for the non-MAGI Medicaid cases (those that are aged, blind, and disabled). 

We’ll know more when CMS releases data from all states, but available data show disappointing ex parte rates. Some states’ rates are in the single digits, and most that we’ve seen are well below 50 percent. CMS recently provided states with additional tools they can use to improve their ex parte rates; ideally, these rates will improve in the states that choose to implement the tools.  

In the long term, ensuring states are maximizing the use of ex parte renewals will help reduce disenrollments at renewal, including procedural ones; reduce churn; and make state operations more efficient by reducing the amount of paperwork handled by state caseworkers. 

State decisions matter. Medicaid is jointly financed by the federal and state governments, but it is solely administered by states. The complexity of application, enrollment, and renewal processes differs across states, and the effect of some of those differences is now apparent in the unwinding data. Further analyses will be needed, but it’s safe to assume that states with the most consumer-friendly online application systems, those that have invested in adequate call center capacity, and those that chose to adopt flexibilities by CMS will outperform states that didn’t take these steps. Unfortunately, the ability to access Medicaid by those who are eligible continues to vary across states. CMS has been working closely with states to, at a minimum, ensure states are following all renewal requirements and recently shared information about where states are falling short and what mitigation measures are in place. 

What We Don’t Know from the Data 

Who is losing Medicaid coverage. The data doesn’t give us a breakdown by race and ethnicity, by eligibility category, or by age. A few states share some of this data, but as it’s not required to be sent to CMS, not all states will make this information available. We know that administrative burdens, such as the renewal process, exacerbate inequities. Prior to unwinding, the U.S. Department of Health and Human Services (HHS) projected that Latino, Black, Asian American and Pacific Islander, and multi-racial Medicaid enrollees are more likely to be disenrolled during unwinding despite being eligible for Medicaid (see Figure 4). Because of this, including data by race and ethnicity is necessary to fully understand the picture of who is most and least likely to successfully complete the renewal process. 

The same HHS projections showed that children would be disproportionately affected, with more than 5 million losing coverage and 73 percent of those being procedural disenrollments (see Figure 3). Because states are not required to report disenrollments by age to CMS, we must monitor enrollment numbers by state. The Georgetown Center for Children and Families has started tracking this data. 

How many people will re-enroll? Many people who lose coverage are likely to re-apply. Many will be eligible and re-enroll in Medicaid. This cycle of losing coverage and then re-enrolling is known as churn. Monitoring application numbers in the coming months will help paint a picture of how many people who are losing coverage believe they are still eligible and ultimately churn back onto Medicaid.    

Do these data reflect normal Medicaid operations or are the results unique to unwinding? During the unwinding process we are seeing new data from states. Greater transparency around renewal outcomes, such as ex parte renewal rates and procedural denial rates, sheds important light on how well state Medicaid programs are retaining eligible people. One unanswered question is whether the unwinding data is reflective of normal operations in Medicaid, or whether the troubling data we are seeing now is unique to the unwinding period.  

CMS is expected to release its first set of data that will allow better comparisons across states. This data, along with continued data transparency from states, will continue to show where state administrators and advocates should focus in order to ensure eligible people are able to successfully renew their Medicaid coverage.