State Flexibility in a New Era What are the research priorities for Section 1332 waivers? Section 1332 of the Affordable Care Act (ACA) provides an op- related to the evaluation of Section 1332 waivers. First, the brief portunity for states to waive key provisions of the federal health provides an overview of the state innovation waiver program. It law in service of state-specific strategies to improve coverage. describes substantive requirements and limitations applicable to These “state innovation” waivers must adhere to the overarching waivers; state and federal responsibilities during the waiver ap- goals and objectives of the ACA itself: waivers that are likely to plication process and after application approval; and similarities undermine the comprehensiveness, accessibility, or affordability and differences between the new Section 1332 waiver program of coverage, or that will impose additional costs on the federal under the ACA and Section 1115 of the Social Security Act, which government, are prohibited by statute. Yet within these limits, Sec- authorizes the Secretary of Health and Human Services (HHS) tion 1332 waivers may offer significant policymaking flexibility, to undertake Medicaid demonstrations that promote Medicaid potentially enabling states to make broad changes to their cover- objectives. Second, the brief explores critical issues related to the age systems as well as more targeted modifications to the federal evaluation of Section 1332 waivers. It identifies research priori- regulatory framework.1 ties, methodological challenges, data requirements, and key considerations related to the timing of evaluations. It concludes Though Section 1332 waivers are a potentially powerful tool, with recommendations by participants to support robust waiver experience with the program is quite limited. The ACA’s draft- evaluations and improved state and federal policymaking under ers designed the waiver program with a lagged start, so that 2017 Section 1332. was the first year in which a waiver could take effect. To date, the federal government has approved just four waiver applications, The Innovation Waiver Framework: Substantive and and only one state has had a waiver in place for a full year.2 Efforts Procedural Considerations to understand how Section 1332 waivers are working in practice, What States Can Waive, and What They Can’t and to assess their effects in relation to federal requirements and Section 1332 of the ACA authorizes states to request waivers of state expectations, have just begun. many of the health law’s core requirements, to allow states to pursue alternative strategies for achieving coverage outcomes In October 2017, AcademyHealth hosted an expert meeting to ex- comparable to or better than those produced without a waiver. plore research priorities related to the evaluation of Section 1115 Waivers can last for up to five years at a time and may be renewed. Medicaid demonstration waivers and Section 1332 state innova- States may seek to modify or eliminate rules concerning: tion waivers. This brief summarizes key findings from the meeting Genesis of this Brief: AcademyHealth hosted an expert meeting to explore research priorities related to the evaluation of Section 1115 Medicaid demonstrations and Section 1332 state innovation waivers on October 17, 2017 in Washington, D.C. This brief includes highlights from the Section 1332 discussions between meeting participants including state and federal officials and national and state policy experts from the nonprofit and health services research communities. Funding for the meeting and this brief was provided by the Robert Wood Johnson Foundation. The views expressed here are those of the author and do not necessarily reflect the views of the Robert Wood Johnson Foundation, the meeting participants, or the reviewers. For more information on this topic, visit academyhealth.org/stateflexibility. State Flexibility in a New Era: What are the research priorities for Section 1332 waivers? • Benefits and coverage levels, including the ACA’s definition of • Deficit neutrality. The state waiver program must not increase essential health benefits (EHB), annual limitations on cost-shar- the federal deficit.5 ing, and metal tiers; The Application Process: State Responsibilities • Marketplaces and rules for marketplace health plans, including To gain approval for a Section 1332 waiver, a state must submit in- requirements related to marketplace establishment and operations formation and data sufficient for federal officials to determine that as well as certification requirements for the qualified health plans the state’s proposal meets the guardrails.6 Federal rules and guid- (QHPs) sold through the marketplaces; ance specify that states must include in their applications: • Actuarial analyses and certifications to support the state’s • Subsidies (premium tax credits and cost-sharing reductions), forecasts with respect to the comprehensiveness, affordability, and including subsidy amounts and eligibility requirements; and coverage guardrails; • Requirements to maintain coverage (the individual and • Economic analyses with respect to all four guardrails, including employer mandates), including whether and how to apply the a detailed 10-year budget plan demonstrating federal deficit neu- requirements and the penalty amounts.3 trality and a detailed analysis of the expected impact of the waiver Critically, the waiver framework also makes available federal funding on health insurance coverage in the state; in certain circumstances. If a state waiver program is forecast to reduce • Data and assumptions used to demonstrate compliance with the federal spending on subsidies, the state is entitled to have these savings guardrails, including a description of the model used to produce passed through to it for purposes of implementing the waiver. It is this coverage and deficit estimates; feature of the program that has driven most of the waiver applications so far submitted to and approved by the federal government.4 • An implementation timeline; Provisions of the ACA not specifically designated as waivable in • An explanation of the waiver’s expected effects on provisions of Section 1332 cannot be modified. Non-waivable provisions include the ACA that are not being waived; rules prohibiting insurers from denying or limiting coverage • Periodic data reporting targets to demonstrate ongoing compli- because of a pre-existing condition, or charging a higher premium ance with the guardrails; and based on an individual’s health status or gender; bans on annual and lifetime coverage limits; the requirement to cover certain pre- • Additional supporting information, as requested by federal of- ventive services without cost-sharing; and other protections barring ficials during application review. discrimination based on health status or other factors. The state’s analyses must include certain specific types of information “Guardrail” Provisions Set Limits on Flexibility that would support a determination that each guardrail is satisfied: The health law conditions approval of any state waiver on the satis- • Coverage take-up: the state must supply information on the faction of four requirements, generally designed to ensure that state number of individuals covered, by income, health status, and age residents are not made worse off by their state’s alternative coverage groups, under current law and under the waiver, including year- approach than they would have been under the ACA’s standard by-year estimates; the application should also identify any types of framework, and to limit the federal fiscal commitment to the state individuals less likely to be covered under the waiver, compared program. These conditions, sometimes called the waiver “guard- to current law; rails,” address: • Affordability: the state must supply information on estimated • Coverage take-up. The state waiver program must “provide cov- individual out-of-pocket costs by income, health status, and erage to at least a comparable number of its residents” as would age groups, absent the waiver and with the waiver; the applica- receive coverage without the waiver. tion should also describe any changes in employer contributions • Affordability of coverage. The state waiver program must to health coverage or in wages expected under the waiver, and “provide coverage and cost-sharing protections against excessive should identify any types of individuals for whom affordability of out-of-pocket spending that are at least as affordable” as would be coverage would be reduced by the waiver; provided without the waiver. • Comprehensiveness of benefits: the state must supply informa- • Comprehensiveness of benefits. The state waiver program must tion explaining how the benefits offered under the waiver differ “provide coverage that is at least as comprehensive as the coverage from the benefits provided absent the waiver, and how the state defined in” the ACA’s EHB requirement. determined the benefits to be comprehensive; and 2 State Flexibility in a New Era: What are the research priorities for Section 1332 waivers? • Deficit neutrality: the state must supply information showing yearly In addition, states are bound to “fully cooperate” with mandatory changes in the federal deficit due to the waiver and a description of federal evaluations of the waiver program.12 This cooperation in- all costs associated with the program, including but not limited to cludes the requirement that states submit “all” data and information federal administrative costs and foregone tax collections. requested by federal officials to support the evaluations.13 The Application Process: Federal Review Federal Responsibilities, Post-Approval: Oversight and The Secretaries of HHS and of the Treasury (together, the Depart- Evaluation ments) share authority for reviewing and approving state Section Federal officials retain significant responsibility with respect to a 1332 waiver applications.7 “[O]nly if ” the Departments determine state’s Section 1332 waiver program after its approval and imple- that the guardrails have been met “may” an application be granted.8 mentation. Officials must re-determine a state’s pass-through funding on an annual basis; review and, as appropriate, investigate In 2015, the Obama administration issued guidance describing documented complaints that a state is failing to comply with the how the Departments would assess whether a waiver proposal terms and conditions of the waiver; and provide comments to the meets the statutory guardrails (Exhibit 1). As of January 2018, this state in response to its annual waiver report. guidance remained in force, though participants observed that it might be modified—perhaps significantly—by later actions of the Trump administration.9 The guidance specifies that officials Federal officials are also charged with conducting periodic evalua- will examine both the overall effect of a waiver and its particular tions of the state’s Section 1332 waiver program. This requirement effects on different groups—especially vulnerable populations. is not yet well defined: no evaluation of an approved waiver has Meeting participants observed that while this framework does not yet occurred and the statute and its implementing regulations and prohibit tradeoffs that create policy winners and losers—for ex- guidance describe only the general parameters of the evaluation ample, the fact that a waiver is likely to cause coverage to be more process. Evaluations must involve a review of the state’s annual affordable for some residents, and less affordable for others, does reports, and officials have suggested that the “primary focus” of any not, in and of itself, require denial—proposals likely to negatively evaluation will be on the waiver’s compliance with the guardrails.14 affect already vulnerable residents are impermissible. In the preamble to a 2012 regulation specifying certain process requirements related to Section 1332 waivers, the Departments In addition to determining whether a waiver request complies with listed a number of broad potential evaluation criteria to be used as the guardrails, federal officials must also calculate its funding rami- a “starting point” for the development of more detailed guidance fications. For applications that seek to capture excess federal funds setting forth evaluation standards that the Departments promised that would have been spent on marketplace subsidies absent the to issue in the future.15 This list suggested that evaluations should waiver (“pass-through funds”), officials must determine the amount consider the impact of a waiver on: of funding due to the state “tak[ing] into account experience in the relevant state and similar states.”10 • “Choice of health plans for individuals and employers; • Stability of coverage for individuals and employers; State Responsibilities, Post-Approval: Reporting and Cooperation States that are given approval for a Section 1332 waiver must comply • Small businesses, individuals with pre-existing conditions, and with additional requirements designed to facilitate monitoring of the the low-income population; waiver and its ongoing compliance with the guardrails.11 Within six months of a waiver’s implementation and annually thereafter, states • The overall health care system in the state; and must hold a public forum to solicit comments on the waiver’s prog- • Other states and the federal government.”16 ress. States must also submit quarterly and annual reports to federal officials that include, among other things, information regarding To date, the Departments have not issued any additional guidance operational challenges, data relevant to an assessment of guardrail further defining these potential criteria or otherwise establishing compliance, and a summary of the annual public forum, including formal evaluation standards for the waiver program. all public comments received and the state’s responses to them. States must also perform periodic reviews related to the implementation of Consistent with the federal government’s approach to other waiver their waivers. programs, the Departments reserve the right to suspend or termi- nate a waiver at any time prior to its scheduled expiration, if they determine that the state materially failed to comply with the terms and conditions of the waiver, including the guardrails. 3 State Flexibility in a New Era: What are the research priorities for Section 1332 waivers? Exhibit 1. Federal Requirements for Satisfying the Section 1332 Waiver Guardrails Statutory Guardrail Guidance Coverage take-up: Waiver program must provide coverage to a At least as many individuals who had minimum essential coverage (MEC) absent comparable number of state residents as would receive coverage the waiver must be forecast to have MEC under the waiver program, in each year without it. of the program. This assessment requires consideration of the overall impact of the waiver program on all state residents, regardless of the type of coverage they would have absent the waiver, as well as its effects across different groups of residents. In particular, this review must account for effects on vulnerable populations, including those with low incomes, elderly individuals, and those with or at greater risk of developing serious health issues. The assessment must also account for whether the waiver program sufficiently prevents gaps in or discontinuations of coverage. Affordability: Waiver program must provide coverage and cost-sharing Coverage under the waiver program must be forecast to be as affordable overall protections against excessive out-of-pocket spending that are as for state residents as coverage absent the waiver, in each year of the program. affordable as would be provided without it. Affordability is measured by comparing residents’ net out-of-pocket spending, including premium contributions, cost-sharing, and spending on non-covered services (to the extent affected by the waiver proposal), to their income. This assessment requires consideration of the average impact of the program on all state residents, regardless of coverage type, as well as its effects on individuals with large health care spending burdens. This review must also account for effects across different groups of residents, particularly vulnerable populations, including those with low incomes, elderly individuals and those with or at greater risk of developing serious health issues. Comprehensiveness of benefits: Waiver program must provide Coverage under the waiver program must be forecast to be at least as coverage that is as comprehensive as would be provided without it, comprehensive overall for state residents as coverage absent the waiver, in each as certified by the Office of the Actuary of the Centers for Medicare & year of the program. Medicaid Services based on sufficient data from the state and from comparable states about their experience with programs created by the Comprehensive coverage is coverage that meets the ACA’s EHB requirements or, ACA and the provisions of the ACA that would be waived. as appropriate, standards under the state’s Medicaid and CHIP programs. This assessment requires consideration of the impact of the waiver program on all state residents, regardless of coverage type, as well as its effects across different groups of residents. In particular, this review must account for effects on vulnerable populations, including those with low incomes, elderly individuals, and those with or at greater risk of developing serious health issues. Deficit Neutrality: Waiver program must not increase the federal deficit. The waiver program must be federal deficit neutral over the period of the waiver and over a ten-year budget period. A waiver that increases the deficit in any given year is less likely to meet this requirement. This analysis must account for the effect of all changes in federal revenue and spending resulting from the waiver program while holding the state’s Medicaid policies constant. That is, any spending effects produced by changes to the state’s Medicaid program under a Section 1115 demonstration will not be considered when evaluating the Section 1332 waiver. Source: Department of the Treasury and Department of Health & Human Services, “Waivers for State Innovation, Guidance,” 80 Fed. Reg. 78131, Dec. 16, 2015. 4 State Flexibility in a New Era: What are the research priorities for Section 1332 waivers? Section 1332 Waivers and Section 1115 Waivers: Similarities, But demonstration evaluation. By contrast, Section 1332 waivers are, as Key Differences one participant put it, “requests to do business differently.” Though Section 1115 of the Social Security Act authorizes states to seek participants argued that Section 1332 waivers ought to be premised federal approval to implement an “experimental, pilot, or demon- on clearly defined policy objectives, a waiver’s approval and renewal stration project” that is “likely to assist in promoting the objectives” depends primarily on the satisfaction of the statutory guardrails, of the Medicaid program.17 Demonstration waiver authority under whatever its theory of change. Section 1115 long predates the ACA’s Section 1332 waiver program and has been used frequently, particularly since the mid-1990s, to Finally, on a more practical level, participants noted that federal make significant changes to state Medicaid programs.18 capacity to review and assess waivers under the two programs was highly likely to be different. Participants suggested that institutional Meeting participants suggested that the federal framework for Sec- resources and—given the newness of the program—experience tion 1115 waivers informed the development of the Section 1332 supporting the Section 1332 waiver program is more limited than waiver program. Rules requiring public input on the waiver process what is available for Section 1115 Medicaid demonstrations and and governing state reporting on waiver progress are similar, for there is a risk that this could affect both the speed and thorough- example, and the ACA specifically provides for the coordination ness of federal review. and consolidation of waivers under Section 1115 and Section 1332. However, participants cautioned that there are important differ- Evaluating Innovation Waivers: Research Priorities And ences between the two authorities that limit the degree to which Challenges experiences with Section 1115 demonstrations should inform Who Must (And Should) Evaluate? expectations under Section 1332. Meeting participants noted that evaluation requirements have been important to understanding the impacts of state Section 1115 The core difference between waivers under the two sections lies in demonstrations and anticipated that rigorous evaluations would be the degree of flexibility each provides—to states, to modify federal especially valuable in the Section 1332 waiver context, as well. Some requirements, and to the federal government, to grant a waiver. In participants suggested that the Section 1332 evaluation framework each case, authority under Section 1115 is significantly greater than contains an improvement over its Medicaid counterpart in that under Section 1332. Section 1115 is, in the view of the Medicaid it requires that periodic evaluations be conducted by the federal and CHIP Payment and Access Commission (MACPAC), perhaps government or an “independent evaluator” selected by the Depart- “unique among waiver authorities . . . combin[ing] extensive waiver ments.21 Whereas, for Section 1115 demonstrations, it is the states authority with a broadly defined purpose for which waivers may be that are charged with evaluating their own policy changes and a granted.”19 Waiver authority under Section 1115 is not unlimited: in federal assessment is optional, these participants saw value in the addition to important process and transparency requirements and fact that the current Section 1332 framework relies on evaluations limitations on the federal Medicaid provisions subject to waiver, conducted independent from the state —provided the Departments Section 1115 proposals must be budget neutral for the federal gov- maintain sufficient resources to perform them. ernment and are evaluated for approval under a set of non-statutory criteria designed to reflect the current objectives of the Medicaid At the same time, participants cautioned that the evaluation program.20 Nevertheless, because Section 1115 waivers are not sub- framework for Section 1332 waivers has not been fleshed out—as ject to any limitation equivalent to the Section 1332 statutory cover- noted above, guidance describing evaluation criteria and standards age guardrails, requiring assessment and oversight of a waiver’s has been promised but not yet been issued—and it is not fully clear impact on coverage take-up, affordability, and comprehensiveness what a federal evaluation will look like in practice. They also point- of benefits, the flexibility offered in the Medicaid context is notably ed out that there remains significant uncertainty about possible broader than what is available under Section 1332. legislative or regulatory changes to federal health policy under the Trump administration, including modifications to the Section 1332 Participants also suggested that the two programs are intended waiver program and its key guardrail provisions. Because of these for distinct purposes. Waivers under Section 1115 are, by defini- unsettled issues, and because of the potential value in exploring tion, experiments, designed to test the effectiveness of different waiver impacts in addition to those focused on by federal regula- approaches to the provision or financing of Medicaid services tors, participants stressed that there is a critical role to be played by in the state. Waiver applications must describe a hypothesis that entities outside of government capable of performing rigorous and the demonstration will test and that will be incorporated into the independent evaluations. 5 State Flexibility in a New Era: What are the research priorities for Section 1332 waivers? Research Approaches for Federal and Third-Party Evaluators the extent to which the waiver has met its own objectives may facili- In approaching the question of how to assess the impacts of Section tate a better understanding of the relative effectiveness of the state’s 1332 waivers, meeting participants sought to distinguish between policies and its underlying theory of change. Such an assessment efforts to report on a waiver’s progress and the task of formally would likely have broad value: to state actors, including policymak- evaluating its effects. Mandatory reports compiled by states with ers, stakeholders, and the public, considering whether to support a active waivers are important descriptive documents that, by sup- continuation of the waiver’s policies, as well as to those outside of plying extensive data and other information regarding the waiver’s the state, weighing the merits of replicating them. implementation, can provide key inputs informing an evaluation. But participants were quick to note that these materials are not a Third, participants suggested that evaluations include a qualitative substitute for a robust empirical investigation of the effects of key component assessing a waiver’s implementation. This element would waiver features on a policy baseline, utilizing appropriate controls. likely be of particular value in the case of more complex proposals to develop and administer new coverage or subsidy systems, to aid in un- To this end, participants emphasized that any waiver evaluation derstanding not only the ultimate outcomes of the program, but also design must be grounded by clear, testable research questions. Rec- the operational challenges and burdens—on state government, resi- ognizing that demonstration hypotheses are not currently part of dents, and stakeholders—of achieving them. One participant pointed the Section 1332 regulatory framework, as they are under Section to a recent Section 1332 application submitted, and later withdrawn, 1115, participants suggested that defining a clear set of evaluation by the state of Iowa, to highlight this need. The Iowa proposal would criteria was, nevertheless, essential. have required, among other things, that the state develop and admin- ister new tax credit eligibility and coverage enrollment processes in Participants identified three overlapping areas of research. First, place of existing federal systems and conduct extensive outreach and evaluations must carefully operationalize the statutory guard- education to state residents, all within a span of a few weeks before the rails—the critical benchmarks for any waiver program—and assess start of the open enrollment period. The participant suggested that, the waiver’s compliance with them. The fundamental promise and had the Iowa application been approved, it would have been especially limitation of Section 1332 is that “does not exempt states from ac- important for any subsequent evaluation to analyze the state’s efforts to complishing the aims of the ACA,” but rather “gives them the ability stand up its new coverage system and document residents’ experiences (and responsibility) to fulfill [these goals] in a different manner.”22 with the program. Another participant suggested that timely monitor- Participants repeatedly observed that the guardrails are a statutory ing and analysis of implementation would be of sufficient value that safeguard to ensure that a waiver’s “request to do business differ- they should be undertaken, perhaps by third-party researchers, on a ently” does not undermine federal objectives. The federal govern- stand-alone basis prior to a more formal and complete waiver evalua- ment thus has a core interest in determining whether a waiver has tion. Because such assessments might be produced relatively quickly, produced an environment in which coverage take-up, affordability, they could inform state efforts and federal oversight comparatively benefit comprehensiveness, and net federal costs are comparable to early in the life of the waiver, potentially facilitating prompt corrective what they would have been had the state implemented federal law actions to improve the program and safeguard residents. as written. And participants suggested that this interest remains during the life of the waiver. Just as states must demonstrate that a Methodological Challenges proposed waiver will provide coverage that meets the guardrails be- To assess a waiver’s impacts, evaluators must compare the coverage fore such waiver may be first approved, so too must an active waiver landscape with the waiver in effect to a baseline scenario describing be judged against these provisions and shown to comply with them, the expected coverage landscape in the state in the absence of the before it may be continued. waiver. Guidelines for specifying the waiver and baseline scenarios are relatively straightforward: federal regulators have stated that the Yet, if assessing a waiver’s adherence to the ACA’s objectives, as waiver scenario must take into account only the projected effects reflected in the guardrails, should be the primary focus of federal of the waiver itself and other related changes to the state’s health evaluators, participants expressed that evaluations—particularly care system that are contingent only on the approval of the waiver. those undertaken by third parties—should investigate, second, how The baseline scenario, by contrast, is a projection of the state of the a waiver has performed against its own stated goals. Participants world under current law, were the waiver not approved. argued that Section 1332 waiver applications should be driven by and include explicit and concrete policy objectives that, though Participants noted that, in practice, specifying a rigorous baseline at likely to overlap with the guardrails, should be separately treated. the outset of the waiver would require use of significant state-specif- Whereas determining compliance with the guardrails may reveal ic data and warned that modeling the counterfactual in subsequent whether a waiver has done harm, a broader evaluation that gauges years of the waiver program was likely to be difficult. This might be 6 State Flexibility in a New Era: What are the research priorities for Section 1332 waivers? especially true if there were other significant legal or policy changes enrollment outside the marketplace. One participant noted that, in to health insurance regulation at the state or federal levels during evaluating waiver effects on a state’s individual market, it would be the course of the waiver for which evaluators would need to con- important to understand enrollment in alternative coverage prod- trol—not an unlikely scenario in the current environment. These ucts, such as short-term, limited duration coverage and healthcare types of complexities, common to policy and program evaluation sharing ministries, about which data are often lacking. The partici- more generally, including evaluations of demonstrations under Sec- pant suggested that it should be obligatory for a waiving state to col- tion 1115, will require evaluators to devote significant resources and lect data, such as these, necessary to evaluate the program. Depend- employ sophisticated research methods. ing on the waiver’s focus, evaluators might also require enrollment data for other market segments. For example, to analyze Hawaii’s Participants also described the challenge posed by attempting to approved waiver of certain ACA rules related to the Small Business unpack the effects of a complex waiver that carries out multiple Health Options Program (the SHOP), evaluators may need data policy changes. While untangling the effects of constituent parts showing enrollment in the state’s small group market. of a waiver may not be necessary for an analysis of the program’s compliance with the guardrails, participants expressed that it would Evaluation of residents’ access to health care services, utiliza- be valuable to do so, as noted above, in the context of a broader tion, and out-of-pocket spending may be informed by federal investigation of the effectiveness of specific policy choices. Partici- data sources, claims and encounter data, and, depending on the pants noted that certain types of implementation approaches could nature of the waiver, evaluation-specific surveys of state residents. make this difficult task more manageable: for example, a program Claims data might be obtained from all-payer claims databases in that staggered the implementation of key features over time might the states that have established them or, potentially, via the data make it easier to isolate the outcomes associated with each part. submissions made by insurers to the federal government under However, political and potentially financing constraints were likely the ACA’s risk adjustment program. To understand impacts on to make adoption of such an approach unlikely. affordability of premiums, evaluations would need to rely on data showing federal payments for premium subsidies, as well Data Needs as premium rates in the relevant markets. Participants suggested Participants noted that data requirements for a given evaluation that insurers in a waiving state could be encouraged or required would vary depending on the scope of the underlying waiver, but to provide their own estimates of the rate impacts of certain anticipated that thorough assessments might draw upon state- policy changes—for example, the establishment of a reinsurance specific population data, health coverage and cost information for program—in their annual rate submissions, which might inform, the state’s private health insurance market, enrollee surveys, and though not control, subsequent assessments. informational interviews with key stakeholders and state officials. To assess a state’s implementation of its waiver, evaluators might Use of economic and demographic data, including data drawn from interview state officials and government staff as well as key stake- federal surveys, was viewed as foundational, permitting analysis holders, including consumer and patient advocates and individuals of a waiver’s effects on different slices of a state’s population—for representing insurers and providers. Evaluators might also conduct example, by age, income, and health status. Understanding whether surveys of state residents affected by the waiver to assess their un- a waiver produced a disproportionate impact on particular groups derstanding of and experiences with the waiver program. of residents, particularly vulnerable subgroups such as those with low-incomes, elderly individuals, and those with or at greater risk of Timing of Evaluations developing serious health issues, is essential to determining ongo- Section 1332 waivers may run for five years before renewal and ing compliance with the guardrails. must be evaluated “periodically.”23 Federal evaluations must include a review of the annual report or reports submitted by the waiving To study a waiver’s impact on coverage take-up, evaluators may state for the relevant time period, suggesting it may be prudent for need to utilize data showing individuals’ coverage status, includ- an evaluation to occur no earlier than the time by which a state ing uninsurance and churn between insurance programs, avail- report encompassing the waiver’s first year of implementation able through federal surveys. In addition, evaluations would likely has been submitted. Yet there is likely to be value in conducting incorporate data reflecting enrollment through the state’s health a federal evaluation relatively early in the life of the waiver. Early insurance marketplace—available from the federal government in feedback would help federal regulators and state policymakers the case of the 39 states using the federally facilitated marketplace determine whether any mid-course corrections in implementation platform, and from the state-run marketplaces in the remaining are warranted. Such assessments also may offer timely assistance to states and the District of Columbia—as well as individual market other states considering whether to pursue a similarly structured 7 State Flexibility in a New Era: What are the research priorities for Section 1332 waivers? waiver. Participants asserted that, in all events, a federal evaluation • The Departments should clarify that an approved waiver may would need to be performed sufficiently in advance of the deci- not be renewed beyond its initial term if the most recent federal sion about whether to renew the program, so that results from the evaluation determines that the waiver does not comply with the assessment could inform those deliberations. guardrails. Evaluations undertaken by independent entities have relatively • The Departments should maintain sufficient staff resources to greater flexibility in terms of process and timing, but likely would ensure timely, thorough, and independent review of all waiver be driven by similar considerations. One participant suggested applications and to conduct multiple periodic evaluations of that it could be especially valuable for a third-party evaluation to all approved waiver programs before they are considered for occur on a timeframe that would permit the results to be pre- renewal. sented during the waiving state’s annual public forum. As noted • Each waiving state should ensure it collects and reports, timely above, a waiving state is required to respond to public comments and publicly, data and information sufficient to permit a rigor- provided during the forum and include a summary of those com- ous evaluation of its waiver. Collection and public reporting ments and state responses in its annual report to federal regula- should include data and information, including enrollment, tors. In turn, federal officials must review these annual reports describing the state’s commercial insurance market and alterna- as part of the periodic federal evaluation. Thus, an independent tive coverage products available within the state. In addition, evaluation disseminated in this way could raise key issues that and consistent with existing federal requirements obligating the would inform and require action by both state policymakers and public disclosure of state draft and final annual waiver reports, federal evaluators. states should promptly publish all federally-required quarterly reports describing waiver implementation. Participants also saw potential value in efforts by independent entities to assess a waiver prior to its implementation, to aid • Apart from the formal federal waiver review and evaluation public and stakeholder understanding of a program’s probable process, qualified non-governmental research entities should effects during the application review process. However, time and undertake independent assessments of state Section 1332 waiver resource constraints associated with this sort of front-end analysis applications and periodic evaluations of approved waivers. was expected to be considerable. Though timeframes for federal These entities should commit resources sufficient to model and review of waiver applications have been criticized by some as timely publish the projected effects of pending waiver propos- slowing state efforts to innovate, participants suggested that, at als and the observed impacts of implemented waiver programs. least with respect to more complex waiver proposals, thorough Entities should present key findings from periodic evaluations modeling could be difficult to manage—for both federal evalu- to states during the required annual public forum. ators and third-parties—even under the existing framework. The difficulty was likely to be especially acute to the extent states Conclusion continue to modify their plans throughout the review process, as The AcademyHealth expert meeting focused on research was the case with Iowa’s proposed waiver. priorities and challenges related to the evaluation of Section 1332 state innovation waivers. Participants expressed that the waiver program in its current form offers states important flexibility to Recommendations develop alternative strategies for improving coverage outcomes. The Section 1332 waiver program is, in effect, only a year old; no At the same time, they noted that significant uncertainty around waiver approved by the federal government has yet been evalu- the ACA and the Trump administration’s approach to the waiver ated, and efforts to formalize the process of waiver assessment are program made it difficult to anticipate the frequency with which in early stages. Meeting participants, acknowledging the newness states might pursue waivers or the content of their proposals. of the program and ongoing uncertainty around the ACA, raised Participants agreed that it will be critical for the fledgling program several suggestions to support robust waiver evaluations and to have in place a framework that requires rigorous, timely, improve future state and federal policy decisions under Section and independent assessments of all waivers against federal law 1332. guardrail protections and ensures transparency in implementation • The Departments should publish guidance identifying the spe- and data reporting to enable robust waiver evaluations by cific criteria that will be used for federal evaluations. researchers in the field. 8 State Flexibility in a New Era: What are the research priorities for Section 1332 waivers? About the Author: Justin Giovannelli is an associate research 8. 42 U.S.C. § 18052(b)(1). professor at the Center on Health Insurance Reforms, George- 9. While formal federal review processes and criteria have so far remained unchanged, participants noted that states’ practical experiences with the waiver town University. For more information, contact him at Justin. application process have, in the view of state officials and some outside observ- Giovannelli@georgetown.edu. ers, varied widely. 10. Department of the Treasury and Department of Health & Human Services, Acknowledgments: The author would like to thank the meeting “Waivers for State Innovation, Guidance,” 80 Fed. Reg. 78131, 78134, Dec. 16, 2015; 42 U.S.C. § 18052(a)(3). participants for sharing their time and expertise and the Robert 11. State and federal obligations during the term of the waiver are memorialized in Wood Johnson Foundation for its support of the project. The au- a terms and conditions document executed by the waiving state and the Depart- thor is grateful to Linda Blumberg, Emily Brice, Peter Brickwedde, ments and made publicly available by CMS. See CMS, “Section 1332: State Innovation Waivers,” last visited Jan. 24, 2018. Stuart Butler, Michael Cohen, Sabrina Corlette, Jason Levitis, Enrique Martinez-Vidal, Dan Meuse, Sara Rosenbaum, Mona 12. 45 C.F.R. § 155.1320(f)(1); 31 C.F.R. § 33.120(f)(1). Shah, and Anne Weiss for their thoughtful review of and comments 13. 45 C.F.R. § 155.1320(f)(2); 31 C.F.R. § 33.120(f)(2). on the drafts of this brief. 14. Department of the Treasury and Department of Health & Human Services, “Ap- plication, Review, and Reporting Process for Waivers for State Innovation, Final Rule,” 77 Fed. Reg. 11700, 11711, Feb. 27, 2012 (“Process Requirements for State Innovation Waivers, Final Rule”). Endnotes 1. See, e.g., D. Bachrach, J. Ario, and H.E. Davis, “Innovation Waivers: An Oppor- 15. Process Requirements for State Innovation Waivers, Final Rule, 77 Fed. Reg. at tunity for States to Pursue Their Own Brand of Health Reform,” The Common- 11711. wealth Fund, Apr. 2015. 16. Process Requirements for State Innovation Waivers, Final Rule, 77 Fed. Reg. at 2. The federal government has approved waivers for Hawaii (effective for 2017), 11711. Alaska, Minnesota, and Oregon (all for terms beginning in 2018). Centers for 17. 42 U.S.C. § 1315(a). Section 1115 Medicaid “demonstration” projects are Medicare & Medicaid Services (CMS), “Section 1332: State Innovation Waivers,” frequently described informally as Medicaid “waivers.” These terms are used last visited Jan. 24, 2018. interchangeably in the text. 3. Under recently enacted federal legislation, beginning in 2019, the tax penalty for 18. In fact, Section 1115, enacted in 1962, predates the Medicaid program itself. For individuals who do not maintain minimum essential coverage is reduced to zero, more information regarding state waiver efforts under Section 1115, see, e.g., E. effectively eliminating the individual mandate. Hinton, M. Musumeci, R. Rudowitz, et al. “Section 1115 Medicaid Demonstra- 4. J. Giovannelli and K. Lucia, “Status of State ACA Innovation Waivers,” The Com- tion Waivers: The Current Landscape of Approved and Pending Waivers,” Kaiser monwealth Fund, Nov. 10, 2017. Family Foundation, Dec. 13, 2017. 5. 42 U.S.C. § 18052(b)(1)(A)-(D). 19. Medicaid and CHIP Payment and Access Commission, “Waivers,” last visited Jan. 24, 2018. 6. An applying state must also identify a state law that provides the state with au- thority to implement its proposed waiver. 42 U.S.C. § 18052(a)(1)(C); 45 C.F.R. § 20. CMS, “About Section 1115 Demonstrations,” last visited Jan. 24, 2018. 155.1320(f)(3)(ii); 31 C.F.R. § 33.120(f)(3(ii). 21. 45 C.F.R. § 155.1320(f)(1); 31 C.F.R. § 33.120(f)(1). 7. Formally, the Secretary of the Treasury is responsible for reviewing any portion 22. H. Howard and G. Benshoof, “Section 1332 Waivers and the Future of State of an application that requests to waive a provision or provisions of the Internal Health Reform,” Health Affairs Blog, Dec. 5, 2014. Revenue Code, while the Secretary of HHS is responsible for reviewing all other aspects of the application. 42 U.S.C. § 18052(a)(6). In practice, however, the 23. 45 C.F.R. § 155.1328(a); 31 C.F.R. § 33.128(a); see also 42 U.S.C. § 18052(a)(4) impact of a waiver is evaluated as a whole and the Departments jointly assess a (B)(v). state’s full application. 9