Continuous Eligibility Policies And CHIP Structure Affected Children’s Coverage Loss During Medicaid Unwinding

Abstract

In April 2023, with the ùnwinding’ of the Families First Coronavirus Response Act (FFCRA) continuous enrollment provisions in Medicaid, states were permitted to commence redetermination and disenrollment procedures for Medicaid beneficiaries. Using January 2021-December 2023 Centers for Medicare and Medicaid Services monthly state enrollment data for 49 states and the District of Columbia, we examined changes in children’s Medicaid and Children’s Health Insurance Program (CHIP) coverage during the Medicaid unwinding overall and by 1) whether states had previous 12-month continuous eligibility policies for children and 2) structure of the state program for CHIP. We found substantially lower Medicaid and CHIP enrollment among children during the unwinding compared to the FFCRA period, with lower levels of coverage declines among children in states that had previous 12-month continuous eligibility policies and states with a program structure of separate CHIP or Medicaid expansion CHIP, rather than combination CHIP. This study highlights the consequences of the FFCRA unwinding for children’s Medicaid and CHIP enrollment, as well as potential state health policies that can promote coverage continuity and prevent further coverage loss for children moving forward.

Introduction

Providing 12 months of continuous enrollment for children insured through Medicaid or Children’s Health Insurance Program (CHIP) has been a long-standing policy option to improve children’s coverage stability and reduce periods of uninsurance.¹ During the COVID-19 public health emergency (PHE), the Families First Coronavirus Response Act (FFCRA) included a provision that offered states enhanced federal matching funds for the continuous enrollment of Medicaid beneficiaries.² This provision, which all states opted in to, was in place from March 2020 until April 2023, when states could commence Medicaid redetermination and disenrollment procedures.²

Under the FFCRA, children had continuous enrollment in Medicaid coverage throughout the PHE, but the magnitude of coverage benefits for children resulting from this policy varied based on state health policy characteristics. States operate CHIP in one of three ways: either via a separate CHIP structure, an expansion of Medicaid eligibility, or a combination of these approaches. Importantly, the FFCRA provision did not apply to children enrolled in separate CHIP coverage, but did apply to children enrolled in expanded Medicaid eligibility programs.³ Additionally, states without pre-FFCRA 12-month continuous eligibility for children—approximately half of all states—experienced larger coverage gains for children.^4,5 In contrast, states with existing 12-month continuous eligibility policies saw smaller increases in enrollment, as fewer children were impacted by the FFCRA provision.^4,5

These policy characteristics are likely responsible for some of the state-level variation in Medicaid and CHIP coverage loss among children as states rolled back the continuous enrollment provision during the FFCRA ‘unwinding’.⁶ Although there was also state variation in unwinding approaches, including initiation timing, pace, and prioritization for renewals,² this study focuses on state Medicaid and CHIP policies in place prior to the PHE that states can continue to leverage to influence children’s coverage patterns beyond the FFCRA period. Thus, the objective of this study was to examine the association between state 12-month continuous eligibility policies and CHIP structure type and changes in children’s Medicaid and CHIP enrollment during the FFCRA unwinding. This work could provide additional evidence of the protective effects of continuous eligibility policies for children at an important juncture in Medicaid and CHIP policy, as the Consolidated Appropriations Act (CAA) of 2023 made 12-month child continuous eligibility policies mandatory and several states are considering or have implemented multi-year Medicaid continuous eligibility policies for children.^2,7 In addition, these findings could also influence state health policy decisions regarding Medicaid-CHIP integration and inform other policy strategies to improve children’s access to health insurance and ensure stability of Medicaid and CHIP coverage for children in the future.

Study Data and Methods

Study Data and Population

This study used January 2021-December 2023 monthly data on children’s Medicaid and CHIP enrollment by state, available from the Centers for Medicare and Medicaid Services (CMS).⁸ Our analysis included 49 states and the District of Columbia, excluding Arizona, which does not report children’s enrollment data. Our outcome of interest was children’s enrollment in Medicaid and CHIP coverage, measured as a proportion of children in the state using population data available from the U.S. Census Bureau. This study was considered not human subjects research by the Rutgers University institutional review board.

Study Design

We used interrupted time series analysis to examine immediate and ongoing changes in children’s Medicaid and CHIP enrollment during the FFCRA unwinding relative to enrollment trends during the FFCRA coverage provision. We considered the start of the unwinding to be the month when each state began their disenrollment procedures, and the ongoing period to be until December 2023. Our primary interest was in the enrollment changes occurring after states commenced disenrollment. We did not examine coverage changes beyond December 2023 as we aimed to focus our analysis on the period before 12-month continuous eligibility for children became mandatory in all states starting January 2024.²

Statistical Analysis

The interrupted time series models included a monthly linear time trend, a binary indicator for the period before or during the FFCRA unwinding, and an interaction between the monthly linear trend and the FFCRA unwinding period indicator (described further in online Appendix A1).⁹ We controlled for the state unemployment rate from the U.S Census Bureau, the state upper income eligibility limits for children for Medicaid and CHIP as a % of the federal poverty level, state Medicaid expansion status under the Affordable Care Act (ACA), state 12-month continuous eligibility policies for children, state CHIP structure (separate, combination, or Medicaid expansion CHIP), and state, year, and month fixed-effects. We included these characteristics as they have been historically linked to Medicaid/CHIP enrollment. Prior studies have found increased Medicaid enrollment following rises in unemployment rates,¹⁰ increased public coverage enrollment in states with higher income eligibility limits for Medicaid/CHIP and expanded Medicaid under the ACA,^11,12 higher Medicaid coverage rates among states with 12-month continuous eligibility policies for children,¹³ and higher uninsurance rates among children in states with separate or combined CHIP compared to Medicaid expansion CHIP.¹⁴

As states took different approaches to the FFCRA unwinding process, we additionally controlled for several unwinding-specific policies including strategy for prioritizing renewals (time-based approach, population-based approach, hybrid population and time-based approach, or state-determined approach), estimated time to complete all renewals (<9 months, 9–12 months, or 12–14 months), whether states flagged enrollees who may no longer be eligible, and number of waivers granted under section 1902(e)(14)(A) of the Social Security Act. The waivers were granted to address challenges faced by states in the unwinding.¹⁵ The waivers include strategies to increase ex parte renewal rates, support enrollees with renewal form submission or completion to reduce procedural terminations, update contact information, facilitate reinstatement of eligible individuals for procedural reasons, and other strategies to protect beneficiaries during the unwinding period) (presented in online Appendix A2).⁹ Analyses were weighted by child population size. We adjusted for serial correlation in the models and used Newey-West standard errors to address autocorrelation and heteroskedasticity.

To examine differences by state health policy characteristics, we stratified models by 1) whether a state had a 12-month continuous eligibility policy for children prior to the FFCRA and 2) state program structure for CHIP as of the start of the FFCRA unwinding. We did not consider states to offer 12-month continuous eligibility for children if they only provided continuous eligibility for sub-populations of children, such as children under a certain age.¹⁶ We considered state program structure for CHIP based on CHIP structure type for children’s coverage. As a result, states that operated separate CHIP coverage only for pregnancy coverage (Illinois, Michigan, Minnesota, Nebraska, Oklahoma, and Rhode Island) were considered to have Medicaid expansion CHIP structures rather than a combination of separate and Medicaid expansion CHIP.¹⁷ California was considered to have Medicaid expansion CHIP as California’s separate CHIP coverage is available in only three counties in the state (San Francisco, Santa Clara, and San Mateo counties).¹⁷

We used multiple-group interrupted time series analysis to test for significant differences among states without 12-month continuous eligibility relative to states with 12-month continuous eligibility, and among states with separate CHIP and combination CHIP structures relative to states with Medicaid expansion CHIP. We plotted data at the state level to examine differences in the unadjusted changes in the proportion of children enrolled in Medicaid or CHIP from March 2023, the last month before the FFCRA unwinding, to December 2023 (presented in online Appendix A3).⁹ As previous research found higher uninsurance among children in states that run any or all of their CHIP coverage as a separate program (states with combination CHIP or separate CHIP) compared to states with no separate CHIP (Medicaid expansion CHIP only),¹⁴ we conducted a sensitivity analysis collapsing CHIP structure type by whether the state had any separate CHIP coverage in the state (available in online Appendix A4).⁹

Limitations

Our study had several limitations. First, this CMS enrollment data is aggregated at the state and month level, limiting our ability to examine coverage changes across demographic characteristics such as race, ethnicity, age, or family income. Second, we could not track coverage trajectories in this data, such as whether children transitioned to alternative coverage types or became uninsured after the unwinding. Third, although we control for several unwinding-specific policies, we could not account for all state-specific unwinding characteristics that could have affected changes in children’s coverage. Finally, we focused on the 2023 unwinding period prior to the January 2024 CAA implementation to isolate differences by whether a state had an existing 12-month continuous eligibility policy. Future studies should consider the effects of the FFCRA unwinding on children’s coverage inequities, transitions to alternative coverage sources, the role of unwinding-specific policies, and children’s enrollment beyond this period.

Study Results

Most states had adopted Medicaid expansion under the ACA (74.2%) and had combination CHIP structures (61.3%) during the FFCRA (Exhibit 1). Only 4.0% of states had separate CHIP structures, with 34.7% covering children through Medicaid expansion CHIP. States were split with 47.8% having 12-month continuous eligibility and 52.2% not having continuous eligibility policies.

Exhibit 1.

State Characteristics of Children’s Medicaid and CHIP programs During the FFCRA, 1/2021–3/2023.

Characteristics:	Pre-Unwinding Baseline, 1/21–3/23
Unemployment Rate	4.4%
Median Medicaid or CHIP Upper Income Eligibility Limit for Children	231% of the FPL
Medicaid Expansion State	74.2%
No Previous 12-Month Continuous Eligibility Policies	52.2%
With Previous 12-Month Continuous Eligibility Policies	47.8%
Separate CHIP Structure	4.0%
Combination CHIP Structure	61.3%
Medicaid Expansion CHIP Structure	34.7%

Source: Author’s analysis of 1/2021–3/2023 data on state characteristics from Kaiser Family Foundation and the U.S Census Bureau.

Notes: FFCRA is the Families First Coronavirus Response Act. FPL is the federal poverty level.

Exhibit 2 presents unadjusted monthly trends in children’s enrollment in Medicaid or CHIP by whether a state had a previous 12-month continuous eligibility policy for children. During the FFCRA, both sets of states experienced increases in the proportion of children enrolled in Medicaid or CHIP. There is a clear change in the enrollment trend starting at April 2023, the beginning of the FFCRA unwinding, with enrollment declines that were steeper among states without 12-month continuous eligibility policies for children relative to states with 12-month continuous eligibility policy during the unwinding period.

Exhibit 2.

Monthly Trends in Children’s Enrollment in Medicaid or CHIP During the FFCRA and Unwinding by State 12-Month Continuous Eligibility Policies for Children, 1/2021–12/2023.

Source: Author’s analysis of Centers for Medicare and Medicaid Services (CMS) monthly Medicaid and Children’s Health Insurance Program enrollment data among 49 states and the District of Columbia, 2021–2023.

Notes: FFCRA is the Families First Coronavirus Response Act.

Exhibit 3 shows the unadjusted monthly enrollment trends by state program structure for CHIP. There were similar increases in children’s Medicaid or CHIP enrollment across CHIP structure type during the FFCRA. During the FFCRA unwinding there were declines across CHIP structure types, with the steepest declines among states with a combination CHIP structure relative to more similar declines among states with separate CHIP or Medicaid expansion CHIP.

Exhibit 3.

Monthly Trends in Children’s Enrollment in Medicaid or CHIP During the FFCRA and Unwinding by State Program Structure for CHIP, 1/2021–12/2023.

Source: Author’s analysis of Centers for Medicare and Medicaid Services (CMS) monthly Medicaid and Children’s Health Insurance Program enrollment data among 49 states and the District of Columbia, 2021–2023.

Notes: FFCRA is the Families First Coronavirus Response Act.

In adjusted interrupted time series models, the FFCRA unwinding was associated with an overall average monthly decline in the proportion of children enrolled in Medicaid or CHIP by 0.9 percentage points (95% confidence interval [CI]: −1.2, −0.6) (Exhibit 4). Among states without 12-month continuous eligibility, the FFCRA unwinding was associated with monthly decreases in children’s Medicaid or CHIP enrollment by 1.2 percentage points (95% CI: −1.6, −0.9), resulting in an average of 15,536 fewer children enrolled per state each month. States with 12-month continuous eligibility policies experienced monthly enrollment declines of 0.7 percentage points (95% CI: −0.9, −0.4), or 9,266 fewer children per state each month. The difference in unwinding enrollment trends between these groups of states was statistically significant (−0.5 percentage points, 95% CI: −0.9, −0.1).

Exhibit 4.

Adjusted Changes in Children’s Enrollment in Medicaid or CHIP During the FFCRA and Unwinding Overall and by State Health Policy Characteristics for Children, 1/2021–12/2023.

Outcome: % of Children Enrolled in Medicaid or CHIP	Pre-Unwinding Baseline, 1/21–3/23	First Month Disenrollment Began		Ongoing Disenrollment Period
		Change	95% CI	Change	95% CI
Overall
Overall Among Study States (N=50 States)	56.5	−1.1**	−2.1, −0.1	−0.9****	−1.2, −0.6
By State 12-Month Continuous Eligibility Policies for Children
No Previous 12-Month Continuous Eligibility Policy (N=27 States)	55.0	−1.1	−2.4, 0.1	−1.2****	−1.6, −0.9
With Previous 12-Month Continuous Eligibility Policy (N=23 States)	58.1	−1.0**	−1.9, −0.2	−0.7****	−0.9, −0.4
Difference Among States With No Policy vs States With 12-Month Continuous Eligibility Policy	−3.1	−0.0	−1.1, 1.0	−0.5**	−0.9, −0.1
By State Program Structure for CHIP
Separate CHIP (N=2 States)	52.3	−1.5****	−2.1, −1.0	−0.3**	−0.6, −0.0
Combination CHIP (N=27 States)	56.7	−0.9	−2.1, 0.2	−1.3****	−1.6, −1.0
Medicaid Expansion CHIP (N=21 States)	56.5	−0.9***	−1.5, −0.3	−0.5****	−0.7, −0.4
Difference Among States with Separate CHIP Vs States with Medicaid Expansion CHIP	−4.2	0.2	−0.9, 1.3	0.2	−0.1, 0.6
Difference Among States with Combination CHIP Vs States with Medicaid Expansion CHIP	0.2	0.3	−0.7, 1.3	−0.7****	−1.1, −0.4

Source: Author’s analysis of Centers for Medicare and Medicaid Services (CMS) monthly Medicaid and Children’s Health Insurance Program enrollment data among 49 states and the District of Columbia, 2021–2023.

Notes: FFCRA is the Families First Coronavirus Response Act. CI is confidence interval.

^**p<0.05

^***p<0.01

^****p<0.001

In our analysis of CHIP program type, the largest declines in children’s Medicaid or CHIP enrollment during the FFCRA unwinding were among states with combination CHIP, with decreases of 1.3 percentage points (95% CI: −1.6, −1.0) or 17,687 fewer children per state each month, which were significantly larger than states with Medicaid expansion CHIP. Among states with Medicaid expansion CHIP, the FFCRA unwinding was associated with declines of 0.5 percentage points each month (95% CI: −0.7, −0.4). In states with separate CHIP structures, the FFCRA unwinding was associated with monthly decreases in children’s enrollment by 0.3 percentage points (95% CI: −0.6, −0.0). The difference in children’s coverage changes during the FFCRA unwinding among states with combination CHIP compared to states with Medicaid expansion CHIP was statistically significant (−0.7 percentage points, 95% CI: −1.1, −0.4). We did not find any evidence that unwinding enrollment trends were significantly different among states with separate CHIP compared to states with Medicaid expansion CHIP (0.2 percentage points, 95% CI: −0.1, 0.6).

In sensitivity analyses collapsing CHIP structure by any separate CHIP, we found significant differences in unwinding enrollment trends, with larger declines in children’s coverage among states with any separate CHIP relative to states with no separate CHIP (−0.7 percentage points, 95% CI: −1.0, −0.3) (available in online Appendix A4).⁹

Discussion

In this analysis of 2021–2023 monthly CMS enrollment data among 49 states and the District of Columbia, we found meaningful declines in children’s Medicaid and CHIP coverage during the FFCRA unwinding, with considerable state-level variation. State health policies such as 12-month continuous eligibility for children and having a state program structure of separate CHIP or Medicaid expansion CHIP, rather than combination CHIP, were associated with lower declines in children’s Medicaid or CHIP coverage during the unwinding.

Our results show that although state-level 12-month continuous eligibility policies for children were not sufficient to prevent declines in children’s enrollment during Medicaid unwinding, these policies did dampen coverage loss. States without prior 12-month continuous eligibility policies for children may have experienced greater declines in enrollment during the unwinding both because they saw larger increases in enrollment during the PHE, and because children in these states may have been initially re-enrolled and then experienced churn during the unwinding period.^4,5 In states with 12-month continuous eligibility policies, children gained new 12-month periods of continuous eligibility if they were found to be eligible at the time of renewal during the unwinding.^18,19 As a result, children’s enrollment was not affected if they experienced any changes in circumstances that could impact eligibility (such as fluctuations in family income or household size) during this new 12-month continuous eligibility period.^18,19 In contrast, in states without 12-month continuous eligibility for children, children who were initially found to be eligible at the time of renewal were not subsequently protected by continuous eligibility, leaving them at risk of losing coverage if there were changes in household circumstances that affected their eligibility during the early unwinding period, prior to the January 2024 CAA implementation.^18,19 Our findings suggest that the January 2024 national mandatory 12-month continuous eligibility for children in Medicaid and CHIP under the CAA may help mitigate some coverage loss in the 27 states that newly implemented this policy.

Our results also found that Medicaid expansion CHIP structures may have been protective against some coverage loss during the unwinding compared to states with combined CHIP structures. Several states transitioned to Medicaid expansion CHIP structures just prior to the unwinding, which could have helped streamline Medicaid administrative processes.² States with separate CHIP structures may have similarly seen lower coverage loss during the unwinding as children with separate CHIP coverage did not have continuous enrollment under the FFCRA and transitions from Medicaid to separate CHIP in the state could have offset some enrollment declines.³

We found that children in states with combined CHIP structures experienced greater coverage loss. Combined CHIP states vary considerably in their structure and administration, with some using Medicaid expansion CHIP to cover younger or lower-income children and separate CHIP to cover older or higher-income children. Future studies should examine which subpopulations of children in combined CHIP states experienced the greatest coverage loss as a result of the FFCRA unwinding, as these findings could inform future policies aimed at promoting continuity of children’s coverage in combined CHIP states.

This research highlights the role that state health policies had on children’s Medicaid and CHIP enrollment during the FFCRA unwinding, focusing on policies in place prior to the unwinding that impacted children’s coverage retention. Understanding how state policies influenced children’s Medicaid and CHIP coverage during this period can help identify state environments where children were most vulnerable to coverage loss and inform policies to improve coverage for these groups. Examining state 12-month continuous eligibility during the unwinding can provide insights into whether and to what extent these state-level policies reduced children’s coverage loss and if national 12-month continuous eligibility could offer similar protective effects. Demonstrating that 12-month continuous eligibility protected children’s coverage during the unwinding strengthens the case for maintaining or expanding the policy in the future, including in the event of federal changes to the health policy agenda.²⁰ Our findings also suggest that policies providing multi-year continuous eligibility for young children in Medicaid and CHIP, which have been implemented or are currently under consideration in thirteen states, have the potential to improve children’s coverage stability over longer periods.⁷

Most states have now completed their FFCRA unwinding processes, and as of December 2024, the most recent data shows a total of 5.53 million children have lost Medicaid coverage.²¹ With this substantial coverage loss among children, it is important to identify the role of state health policies, such as continuous eligibility periods and CHIP design and structure, in promoting continuity of Medicaid and CHIP coverage for children and preventing further coverage loss moving forward.