To Reduce Coverage Losses, CMS Should Revise Limits On How States Verify Medicaid Eligibility

On September 7, 2022, the Centers for Medicare and Medicaid Services (CMS) published proposed rules that would dramatically streamline eligibility determination, enrollment, and retention in Medicaid and the Children’s Health Insurance Program (CHIP).The proposed regulations reflect an effort by the Biden administration to ensure that people who are eligible can obtain and retain health coverage, free of needless paperwork burdens.

The proposed rule would correct many shortcomings and oversights in existing regulations. However, it leaves intact a policy that limits verification methods that may be used to confirm beneficiaries’ continued eligibility. This policy imposes constraints that could needlessly increase the number of people who lose Medicaid when the public health emergency (PHE) ends. Despite the administration’s considerable efforts to prevent unnecessary coverage loss, its work will not come close to achieving its potential unless CMS revises its old verification restrictions. Fortunately, CMS can make the necessary adjustments by issuing guidance, without any need for further regulatory change.

When The Public Health Emergency Ends, Huge And Inequitable Medicaid Losses Are Likely

A recent report by the Department of Health and Human Services (HHS) Assistant Secretary of Planning and Evaluation (ASPE) finds that, when the PHE ends, 5.5 million children will lose Medicaid if terminations are no more likely than they were before the pandemic, as will 4.6 million Latinos, 2.2 million African Americans, and nearly a million Asian Americans and Pacific Islanders.

Altogether, 15 million people are projected to lose Medicaid. This would be more than seven times the size of the largest annual reduction in Medicaid enrollment recorded in Census Bureau data, when the number of Medicaid beneficiaries fell by 2 million people in 2018 and again in 2019.

A particularly shocking finding is that more than half of the children and people of color who are expected to lose Medicaid will remain eligible but be terminated for administrative reasons. Nearly three-fourths of children terminated from Medicaid (74 percent) will lose coverage despite remaining eligible, according to ASPE, as will almost two-thirds (64 percent) of Latinos. By contrast, just 30 percent of terminated adults and 17 percent of terminated non-Hispanic Whites are projected to remain eligible but lose Medicaid for procedural reasons, under ASPE’s analysis.

Even these numbers may be low: During a mass redetermination, most beneficiaries are at risk of stumbling into procedural termination. The ASPE’s estimates also do not account for the people who will experience a significant delay in securing initial application review because eligibility case workers are so overwhelmed by the redetermination process that new enrollments get put on hold.

All of these complications would be a problem for any insurance program. For Medicaid, they are especially troubling because, unlike other forms of health insurance, Medicaid has a dual role as both insurer and health care safety net. Enrollment is not restricted to annual open enrollment periods and a limited number of special enrollment periods. Instead, people can join Medicaid any time, as part of the process of seeking health care, with providers attempting enrollment that covers prior costs. As a result, enrollment delays and unjustified coverage interruptions could carry especially serious health implications, especially in communities of color, which already suffer deep disparities in health care access and health status.

According to leading national experts, “increasing the proportion of renewals that a state conducts electronically—without sending paperwork to beneficiaries—will likely be the most important single step states and CMS can take to avoid coverage losses.” As we explain next, current federal legal restrictions hamstring efforts to take this vital step.

By Limiting The Range Of Permitted Verification, CMS Needlessly Obstructs State Efforts To Safeguard Eligible Families’ Medicaid Coverage

When the Affordable Care Act (ACA) was enacted in 2010, it required Medicaid eligibility for most people to be based on modified adjusted gross income (MAGI), a federal tax law concept. CMS does not view MAGI as merely defining financial eligibility, however. The agency also sees MAGI as a constraint on permitted verification. The only evidence that states may consider, according to CMS, specifically shows countable income and household size as defined by MAGI, such as data about earnings or interest that is MAGI-countable income. Information about other household characteristics cannot constitute financial verification, no matter how great a likelihood of eligibility such information provides.

For example, Louisiana’s highly successful pre-ACA renewal program, detailed below, verified a child’s continuing eligibility and renewed their coverage administratively if the child lived with a grandparent because the state had found, as a practical matter, that such children almost never became ineligible for Medicaid. Such verification is not permitted by the current CMS approach because having a grandparent as a caretaker is not a personal characteristic defined in terms of countable income.

Using SNAP Receipt To Verify Medicaid Eligibility

The agency’s treatment of Supplemental Nutrition Assistance Program (SNAP) records illustrates CMS’s approach and its consequences. The year before MAGI requirements took effect, CMS acknowledged that 97 percent of people who receive SNAP benefits qualified for MAGI-based Medicaid as expansion adults or children eligible for Medicaid or CHIP. Other research sponsored by ASPE found that, out of 20.8 million children who received SNAP, only 115,000, or 0.6 percent, were ineligible for Medicaid and CHIP.

At the time, states faced serious challenges implementing MAGI-based eligibility, which required modernizing often antiquated eligibility systems. Recognizing those exigencies, CMS permitted temporary waivers of MAGI requirements, letting states grant Medicaid eligibility based on SNAP receipt. But once states completed the transition to MAGI, CMS, in 2015, forbade them from using SNAP receipt to verify MAGI-based Medicaid eligibility, since SNAP-recipient children and expansion adults were only “highly likely” to qualify for MAGI-based Medicaid, and “a small percentage of those individuals might be found income-ineligible for Medicaid based on MAGI.”

Under current policy, states may verify MAGI with SNAP data under only two circumstances:

1. The SNAP recipient’s household structure makes them “certain to have MAGI-based household income at or below the applicable modified adjusted gross income standard [emphasis in original].” As a result, confirmation that a beneficiary receives SNAP provides the state with unambiguous information about income as defined by MAGI.
2. SNAP records are being used to “verify the amount of a particular type of income—for example, wages or unemployment benefits,” that count in determining MAGI.

Recognizing that states once again face tremendous administrative challenges, CMS recently authorized temporary MAGI waivers to let SNAP receipt trigger renewal. But those waivers end after the transition from the PHE to normal operations, so state administrative work required to implement such a waiver may have only a temporary payoff. And that administrative work can be considerable; states must sort through SNAP records and limit renewal to Medicaid beneficiaries with gross income, as determined by SNAP, at or below Medicaid standards. As of September 6, only eight states had sought these waivers.

An alternative approach would let states renew Medicaid eligibility based on SNAP receipt, without requiring states to shoulder the administrative burden of sorting through the details of SNAP case records. Much like having a grandfather as a caretaker relative under Louisiana’s pre-ACA policy, a state could use the simple fact of SNAP receipt to verify MAGI at Medicaid-qualifying levels. The basis for such a verification rule would be that SNAP receipt establishes a very high likelihood of Medicaid-qualifying MAGI, as noted earlier.

Almost every state Medicaid program already uses SNAP data, in some form, for eligibility verification. For many hard-pressed Medicaid programs, it could be substantially easier to use the simple fact of SNAP receipt as the basis for administrative renewal, without being forced to analyze the characteristics of each individual SNAP case. By using SNAP-based renewal to verify MAGI rather than to replace MAGI, this approach can continue long after the PHE ends, so the initial work of implementation would yield benefits over a much longer time horizon.

This approach would shield millions of families from procedural termination. According to raw Current Population Survey data, 42 percent of all Medicaid beneficiaries younger than age 65, including 46 percent of children, received SNAP in 2020. When you consider that CPS data notoriously understate SNAP participation, the actual proportion of Medicaid beneficiaries who receive SNAP now is doubtless much higher.

More Effective Use Of Beneficiary Case Records And Other Data To Develop And Validate Verification Rules

SNAP-based verification is just one example of how CMS could empower states to more effectively protect beneficiaries by conducting verification based on any personal characteristics that establish a very high probability of MAGI-based eligibility.

States could mine both case records and longitudinal survey data to develop business rules that renew beneficiaries administratively based on any characteristics that make them reasonably certain to be eligible, protecting numerous eligible beneficiaries from procedural termination. Fifteen years ago, the state of Louisiana took this more empirically based approach before the ACA and achieved high levels of renewal with very low error rates. Administrative renewal in Louisiana took place, not just when grandparents served as caretaker relatives, but also when the child’s household received only Social Security income; the child had a single parent whose income consisted entirely of child support; or the child had qualified for Medicaid or CHIP for at least three years and had household income below $500 a month.

Administrative renewal based on household characteristics was the most common method of renewal, accounting for 36 percent of all children who retained coverage under Louisiana’s remarkably successful program. In 2008 only 1 percent of Louisiana’s Medicaid and CHIP children lost coverage for procedural reasons, while 90 percent of Medicaid children and 84 percent of those receiving CHIP were renewed without any need for the families to complete paperwork. The state achieved these results without endangering program integrity, lowering eligibility error rates to just 0.3 percent, less than a tenth the national average at the time.

These impressive efforts—and the coverage rates they achieved—illustrate the breadth and potential of verification strategies that would again become available to states if CMS were to revisit its view of MAGI.

As A Matter Of Law, CMS Should Not View MAGI As A Constraint On Permitted Verification

CMS’s view that MAGI is a constraint on verification, not just a definition of eligibility, has created unnecessary restrictions on important tools available to states to renew coverage based on persuasive electronic evidence of continued eligibility.

We urge the agency to issue subregulatory guidance authorizing states to supplement current verification methods with a new set of permitted methods. This new set would be modeled on the federal evidentiary rule that “evidence is relevant” to a fact if the evidence “has any tendency to make [that] fact more or less probable than it would be without the evidence.” Proof of any personal characteristic that makes a beneficiary more likely to have MAGI at Medicaid-qualifying levels would be allowable verification, under this approach. States would have flexibility to define the probability of eligibility above which administrative renewal, rather than procedural termination, takes place.

It should not matter whether such characteristics specifically involve income or household size, as defined by MAGI. The MAGI statutory requirement addresses who is eligible, not the kinds of verification that can be used to confirm eligibility. Section 1902(e)(14)(A) of the Social Security Act, added by ACA Section 2001, provides that, “for purposes of determining income eligibility for medical assistance … a State shall use the modified adjusted gross income of an individual.” The next and final sentence of the subparagraph requires states to set MAGI standards that incorporate an “equivalent income test” that prevent people subject to maintenance of effort requirements from losing eligibility as they transition to MAGI. The next two subparagraphs forbid states from adding income deductions or considering assets in determining eligibility. Put simply, the provisions immediately surrounding the MAGI requirement all define who qualifies for Medicaid, not how those qualifications may be proven.

Any doubts that the MAGI requirement is a definition of eligibility, not a constraint on verification, are erased by another ACA provision, Section 1413(c)(3)(A). This provision requires a state “to the maximum extent practicable” to “establish, verify, and update eligibility” using “reliable, third-party data,” specifically including information from SNAP records and other public benefit programs. Artificially limiting verification to personal characteristics that explicitly involve MAGI contravenes this statutory command because those limits prevent reliable, third-party data from verifying and updating eligibility “to the maximum extent practicable.”

The proposed rule noted at the start of this post highlights other sources of legal authority that justify a broadened approach to verification. In particular, it aims to address the serious problem that only about half of eligible people receive Medicare Savings Program (MSP) benefits, through which Medicaid helps pay Medicare premiums and out-of-pocket costs for poor and near-poor Medicare beneficiaries. The proposed rule would:

• Require states, to the maximum extent possible, to determine MSP eligibility based on data showing participation in a different, federally administered program in which “most” beneficiaries are “likely eligible” for MSP.
• Prohibit states from asking applicants to verify attestations about income and assets by proving the amount of their dividend and interest income, the value of their non-liquid assets, their burial fund amounts, or the face value of their life insurance policies.
• Require certain states to deem people eligible for MSP, based on such people’s prior determination of eligibility for various other Medicaid categories, while giving all states the option to deem additional groups eligible for MSP.

CMS cited multiple sources of legal authority to support these changes:

Those same sources of legal authority support the change in subregulatory policy urged here.

If CMS Allowed Medicaid Verification Based On Federal Evidentiary Rules, States Could Cut Administrative Costs And Strengthen Program Integrity

As noted earlier, the approach to verification urged here is modeled on the Federal Rules of Evidence. Renewal based on data matches, rather than manual paperwork review, would become more common, thereby lowering renewal’s administrative cost. With states facing a mountain of Medicaid redeterminations following the PHE, the verification options proposed here would offer states administrative efficiencies that could prove quite useful.

Both now and under our proposed new verification option, data-based renewal incorporates important program integrity safeguards. Under both Louisiana’s pre-ACA methods and current Medicaid regulations, the state sends the family a notice that describes the basis for the state’s finding of continued eligibility. The notice also explains the family’s legal obligation to correct any errors in the notice. But if the family does not respond, the state continues coverage, instead of terminating it.

Notably, the approach suggested here strengthens program integrity by improving the overall accuracy of redetermination outcomes. As current regulations make clear, program integrity is undermined by the potential both “for ineligible individuals to be approved” and “for eligible individuals to be denied coverage.” Research suggests that fewer overall eligibility errors occur when administrative renewal applies to beneficiaries who are known to have at least an 80 percent likelihood of eligibility. In such cases, administrative renewal prevents numerous procedural terminations of eligible people, while extending coverage to a substantially smaller number of ineligible people. With a population known to have at least a 90 percent likelihood of Medicaid eligibility, such as SNAP recipients, administrative renewal is estimated to lower the number of erroneous redetermination outcomes by more than 50 percent.

Shielding Families From Harm

A tsunami of Medicaid loss is headed our way. We do not know when it will hit, but both federal and state officials are already doing their utmost to trim the tsunami’s eventual size and shield families from harm. Those efforts are being hamstrung by unnecessary and legally questionable limitations on the kinds of information that states may use to verify financial eligibility. Changing old legal views can be uncomfortable, but precisely such changes are needed to protect health care for millions of children and families, especially in communities of color.

Authors’ Note

When Stan Dorn worked at Families USA, the organization urged the Centers for Medicaid and Medicaid Services to take an approach similar to that recommended in this article.