When Should An 1115 Medicaid Expansion Experiment be Approved? Considering The Case Of Georgia Pathways

Historically, demonstration waivers have been used to expand eligibility, improve delivery, increase benefits, and otherwise improve the Medicaid program. The waivers granted during the Trump administration were a departure from both Republican and Democratic administrations’ determinations as to the merits of state waiver proposals. Georgia’s proposed demonstration, known as Georgia Pathways, merited especially close scrutiny: a Medicaid “expansion” for low-income adults in name only, with exclusionary conditions on enrollment and coverage already proven to cause major harm when attempted in other states. In addition to the legal issues raised, the Georgia case raises deeper policy questions about when the Department of Health and Human Services (HHS) should use its considerable section 1115 powers. 

In her discussion of Georgia’s lawsuit to stop the Biden administration from revoking its predecessor’s approval of the Georgia Pathways experiment, Nicole Huberfeld flags the basic legal flaw in the state’s argument: namely, in claiming that section 1115 permits states to depart from federal Medicaid conditions they don’t like, the plaintiff stands federal law on its head. A longstanding, special grant of federal demonstration authority, section 1115 of the Social Security Act gives the HHS Secretary alone the power to allow a state to deviate from provisions of federal Medicaid law when, in the Secretary’s judgment, doing so “is likely to assist in promoting the objectives” of the Medicaid program.

The standard terms and conditions that accompany all section 1115 demonstration project waiver approvals empowers the Centers for Medicare and Medicaid Services (CMS) (which oversees Medicaid waivers) to withdraw waiver approval at any time. Georgia’s waiver was no exception. Furthermore, HHS’s December 23, 2021, withdrawal decision—sent only after a nearly year-long effort to reach an accommodation with the state—was exceptionally well-reasoned. Reversal was based on the impossibility of undertaking a compelled work demonstration during the pandemic, the fact that the federal public health emergency bars states that take extra Medicaid funding (which Georgia has done) from disenrolling beneficiaries, and the fact that recent section 1115 Medicaid expansion demonstrations had tested virtually identical eligibility restrictions (compelled work, premiums, additional reporting) and revealed their adverse impact. In the Secretary’s judgment, the legal justification that authorizes him to approve a demonstration and expend federal funding—the “likely to promote the objectives” test—was simply lacking.

Georgia also argues that it should be permitted to proceed because of its “reliance interest” on the Trump administration’s October 2020 approval. This argument also holds no water since it is impossible to have a reliance interest in a situation in which HHS has explicitly reserved the right to cancel approval when the demonstration no longer fulfills section 1115 requirements. States are never released from HHS oversight in Medicaid. Recognizing that such a clause prevented states from claiming reliance, the Trump administration sent states a notice on January 4, 2021, that sought to summarily amend all outstanding section 1115 approvals to effectively guarantee a demonstration operational period without threat of cancellation. The Biden administration quickly set aside this unlawful attempt to change the terms of the approvals without notice and comment. These events raise broader questions in Medicaid’s reliance on waivers to set policy.

A Broader Set Of Policy Concerns

Beyond its legal failings, Georgia’s lawsuit raises a deeper policy question: When should a section 1115 demonstration project be considered one likely to promote Medicaid’s principle objective of helping eligible families and individuals “whose incomes and resources are insufficient to meet the cost of necessary medical services”? HHS’s focus on health equity is a reflection of underlying health and social conditions that have exacted a significantly greater toll on the most vulnerable populations during an unprecedented pandemic. Waiver provisions granted during the Trump administration including enforceable premiums and work requirements have been shown to deter enrollment with no offsetting gains in access to other forms of insurance coverage and no increase in employment. Yet, Georgia persisted in requesting authority to impose work requirements. Should section 1115 ever be used to allow states to replicate harmful changes in federal Medicaid law on the theory that some people nonetheless might be helped?

In our view, HHS should authorize state experiments on low-income, medically vulnerable populations only when principles based in evidence, equity, and inclusiveness are met. Section 1115 demonstrations are meant to be evidence-based initiatives that reflect the norms of scientific research and evaluation. A waiver is not license on the part of federal officials to change the terms of a law they object to politically. Indeed, section 1115 is not simply a legislative workaround for laws the HHS Secretary administers; it is a special grant of experimental authority.

These questions grow even more urgent in situations such as the incomplete Medicaid expansion, which continues to leave more than two million low-income people out of the Affordable Care Act’s (ACA’s) near-universal coverage plan. The terms under which states can receive hundreds of billions of dollars in federal funding in exchange for adopting more expansive Medicaid eligibility policies are tied to an eligibility category that is precise in its definition (42 U.S.C. § 1396a(a)(10)(A)(i)(VIII)).

Georgia was willing to forgo enhanced ACA funding for the expansion population to gain flexibility imposing eligibility rules of unprecedented strictness. But states should not be permitted to devise an eligibility group of its own preference and still qualify for even the basic federal medical assistance payments (for Georgia, more than 73 cents on the dollar). Just because a state characterizes a demonstration project application as an “expansion” does not mean that the Medicaid program’s goals are served when that expansion comes at a high cost to the ACA expansion population.

Although the US Supreme Court’s decision in National Federation of Independent Business v. Sebelius (NFIB) may have limited the Secretary’s power to enforce the ACA expansion as a mandate, nothing in the decision suggests that federalism concerns bar Congress from setting the terms of federal investment in a Medicaid expansion undertaken by state choice. In other words, Congress fixed the terms of the Medicaid program, and the Secretary must decide whether a state’s proposal offers terms that are reasonable, with real benefits to be gained by the affected population.

When Do Possible Harms Outweigh The Benefits To Be Gained For People Who Need Medical Assistance?

Georgia claims, as the basis for its proposal, the cost savings it will achieve. But reducing program costs is not an independent basis for approving section 1115 proposals. Innovations might carry some risks, but past demonstrations have been based in evidence showing that benefits outweigh risks. For example, section 1115 waivers spurred a new generation of home- and community-based care experiments at a time when institutional care was the norm, and these demonstrations led to major changes in Medicaid’s approach to long-term care. Similarly, prior to the ACA, when Congress had not yet articulated standards regarding coverage of low-income adults, HHS used section 1115 to permit states to test coverage of subcategories of low-income adults. Post-NFIB, HHS used section 1115 waiver negotiation with states to enable Medicaid expansion, although at times with some questionable limitations such as the use of premiums with disqualification of certain members of the newly eligible population for nonpayment. Most prominently, perhaps, HHS allowed Arkansas, as well as New Hampshire and Iowa, to implement the ACA expansion as a “private option” Medicaid demonstration, which used Marketplace plans in lieu of standard Medicaid plans.

HHS has also tested policies that move Medicaid away from its traditional definition of “medical assistance” and in the direction of greater funding of certain services that address the underlying determinants of health. For example, HHS approved North Carolina’s Medicaid managed care experiment, which effectively expands the definition of “medical assistance” for beneficiaries enrolled in managed care plans to include services aimed at improving health broadly. North Carolina’s demonstration, which rests on evidence of the value of investing in services that affect health and not just health care, poses risks because section 1115 budget neutrality rules could lead to artificial limits that force states to limit funds for health care to make room for social investments. The approval process involves negotiations that account for these kinds of concerns and that must be transparent to some degree under the ACA’s section 1115 procedural rules.

Like all experiments involving human subjects, section 1115 waivers as demonstrations carry risks. Poor and medically vulnerable people depend on the government to use reasoned analysis, make an evidence-based determination of whether benefits outweigh risks, design an evaluable experiment, and engage in efforts to understand and mitigate risks before the experiment is launched. What stopped experimentation on the poor with work requirements was how thoroughly ill-considered HHS’s approval was because HHS failed to consider evidence material to approval of any Medicaid demonstration project—the experiment’s impact on access to health insurance.

Georgia’s experiment does not apply to existing beneficiaries, so in a narrow sense, no one loses benefits. But that hardly justifies pouring millions of dollars of federal funding into a demonstration where so few gain anything either. Both HHS and Georgia ignored evidence of predictable harms. The state estimated that over the life of the experiment, it would insure slightly more than 64,000 people—about 1 in 10 of the Georgia population estimated to newly enroll in Medicaid under the ACA expansion. Moreover, ample evidence showed by the time of HHS’s approval that those gaining coverage would face a much higher risk of losing it, particularly in times of economic hardship, due to job loss or illness—when they need benefits the most and are least able to navigate barriers embedded in the experiment. In its October 2020 approval of Georgia Pathways, HHS ignored multiple federal court decisions decrying the Secretary’s pattern of ignoring evidence of harm while authorizing waivers with work requirements. HHS also dismissed extensive research documenting adverse impacts of the Arkansas work experiment on the grounds that Georgia was a “completely different model.” Neither state nor federal officials weighed the waiver’s potential indirect effects on traditional beneficiaries who might erroneously believe that the new rules applied to them.

Section 1115 Medicaid Demonstration Precedents And The ACA Expansion Demonstrations

Over decades section 1115 experiments have also played a central role in advancing health equity, in particular enabling the HHS Secretary to expand access to medical assistance rapidly in times of grave public health threats such as hurricanes and other disasters. As we have noted, section 1115 experiments have enabled states to test more generous eligibility rules, innovations in coverage, and new ways of organizing and delivering health care. 

In an effort to overcome the impact of the US Supreme Court’s 2012 decision making Medicaid expansion optional for states, the Obama administration opened a new era in section 1115 waiver approvals by allowing states to start tailoring Medicaid expansion to impose eligibility conditions never before permitted. The universal entitlement to coverage was thwarted by an unprecedented US Supreme Court decision regarding the constitutional limits of federal spending powers, and the task became getting states to voluntarily embrace expansion. Viewed in this light, it is perhaps understandable that in the scramble to salvage the coverage that was intended, the administration made concessions regarding eligibility, enrollment, and coverage restrictions, which in turn were shown to have an adverse impact on eligibility and enrollment. But the Trump administration escalated waiver approvals in an unprecedented fashion. Georgia was a continuation of these concessions, a proposal to freeze out 90 percent of the expansion population, leaving state residents virtually nothing to show for a very expensive waiver. 

Furthermore, the ACA Medicaid expansion has survived untouched by Congress for a decade, its importance reaffirmed in the face of efforts by the Trump administration and congressional supporters in 2017 to repeal the ACA. Indeed, if anything, lawmakers have retrenched and sweetened the financial incentives to motivate continued expansion as part of the American Recovery Act. Yet, the hold-out states have not changed position, although Oklahoma and Missouri, which already had committed to reform, benefited from the special enhancements provided by the law.

From a policy perspective, no justification exists for rewarding a state such as Georgia by giving it funding to cover 1 in 10 ACA expansion-eligible people using exclusionary criteria already proven injurious. The concerns raised by the Biden administration—experimentation under pandemic and aftermath conditions, and evidence of the past harms caused by such restrictions—would be enough; but also, the experiment offers virtually nothing that is likely to promote Medicaid’s core, indispensable objective of furnishing medical assistance to low-income people. Even if Medicaid has other objectives, basic rules of reasoned agency decision making tell us that any approval must be balanced against the program’s fundamental purpose. Compelled work is a restriction that Congress has consistently rejected, for the obvious reason that Medicaid is the program people turn to because they either cannot work due to being sick or because they are working and have no other access to insurance. HHS already knows premiums and work requirements disenroll otherwise eligible beneficiaries. If costs in Medicaid were truly a concern, this is not how Georgia or HHS would address them—with a multimillion dollar, administratively burdensome exercise in ideology.

Guiding Principles

In an era dominated by equity concerns and faced with a federal demonstration project legacy that permits people who need health care to be unwitting participants in poorly conceived, poorly designed, and potentially harmful experiments, should the Biden administration continue expansion demonstration projects? If so, under what principles? We propose three protective principles guided by concerns for human subjects and health equity.

1. Our first principle is that an ACA expansion demonstration must insure the entire expansion population, not limit it as Georgia would have done. The ACA embraced universality. If a state does not want to cover the entire expansion population in the same way, for example limiting traditional Medicaid to people earning up to 100 percent of federal poverty level and subsidizing Marketplace coverage such as Arkansas, then a state must provide supplementary Medicaid coverage to wrap around subsidized Marketplace coverage to provide benefits only available through Medicaid, such as rehabilitation services. In other words, as with Medicare, Medicaid should become a secondary insurer in a model that splits the ACA population between two markets. Also critical in a bifurcated market approach would be safeguards to ensure a seamless transition between the two markets in the event of slight changes in income or family living circumstances, which are common for low-income populations. For states willing to experiment with Medicaid as a secondary payer and improvements in Medicaid/Marketplace transitions, we would recommend funding at the full ACA enhanced rate. Partial expansion is limited by the ACA, as interpreted by a CMS 2012 ruling. CMS might consider a hybrid to be a testable model that qualifies for the full ACA-funding enhancement so long as the entire expansion population is covered completely.
2. Our second principle is a ban on testing eligibility and coverage restrictions that already have been evaluated and found to have an adverse impact on enrollment accessibility and stability. Premiums with lockout periods for failing to pay or renew on time have been shown to increase enrollment denials and interruptions. No experimental justification exists for demonstrating an approach that has already been shown to be contrary to Medicaid’s core objective. CMS recently revoked premiums in Montana and Arkansas, in alignment with this principle.
3. Our third principle is transparency and inclusion of people most impacted by the experiment. The ACA required both state and federal processes for public comments on waiver applications, but the opportunity to comment is not sufficient, in our view, when people are subjected to involuntary experiments that could compromise access to basic health care. Indeed, in the ACA, Congress revised the definition of “medical assistance” to clarify that Medicaid is not just payment but also includes the promise of access to health care. If a state chooses compulsory waiver participation (nothing in law appears to require states to make acceptance of experimental rules a condition of enrollment and coverage), then HHS should expect that states will precede their submissions with extensive listening sessions that are highly accessible. This includes using plain language so that those potentially affected by the experiment have a chance to understand it, voice their concerns, and have their concerns registered as a part of formal rulemaking, with the listening sessions being part of the record.

Our suggestions do not stifle innovation; they will hopefully help to ensure that health equity guides experimentation on vulnerable people and that those on whose behalf such program innovations are undertaken have a true opportunity to be heard.

Authors’ Note

Sara Rosenbaum is funded through the Commonwealth Fund for work on Medicaid. Alex Somodevilla and Morgan Handley both were previously members of Rosenbaum’s research team, and both now work at private law firms. In each case, however, work on GW projects is pursuant to a service agreement with their respective firms.