Health Reform 2.0: Alternate State Waiver Paths Under the Affordable Care Act

NIHCR Research Brief No. 19

Some states might use a coordinated waiver to establish a single-payer system of tax-financed universal coverage. A key stumbling block would be the high tax rate required to fund a single-payer plan, particularly if it offers more comprehensive coverage than today’s norms. Other states might pursue a more-targeted coordinated waiver to replace the ACA’s health insurance marketplaces and Medicaid for children and nondisabled adults with income-related vouchers to buy private health insurance. In some states, projected funding levels for Medicaid and the marketplaces would be sufficient to support a broad-based voucher program. In other states, a voucher-based approach would face challenges in meeting the ACA Section 1332 coverage and affordability standards. States pursuing a voucher approach could potentially tap into additional federal funding if they agreed to implement the ACA Medicaid expansion as a base for the waiver. In general, states developing either a single-payer approach or a voucher approach would face an easier path if their waiver plans include substantial and effective cost-control components.

ACA 1332 Waivers: Side Door to Innovation?

The Affordable Care Act delegates several key roles to states, including defining the essential health benefits package, implementing an expanded Medicaid program and establishing health insurance marketplaces. So-called Blue states have embraced these roles enthusiastically, although with mixed success.1 Red states have taken a more winding path, mostly rejecting and attempting to undermine the ACA—but in some cases, reshaping the law more to their liking.2

The history of Medicaid, the joint federal-state health program for low-income people, offers guidance about what to expect with state ACA implementation. Medicaid was first signed into federal law in 1965, and federal matching funds were available for states beginning in January 1966.3 About half the states—mostly in the South and West, which today would be characterized as Red states—initially delayed and resisted Medicaid because of philosophical and political objections. The lure of generous federal funding eventually prevailed, however, and all states participate today. Moreover, mainstream conservative proposals today accept Medicaid’s existence and instead emphasize increased state flexibility to administer the program.4 It’s likely the ACA may follow a similar path into the mainstream over the coming years. And, as the ACA become less polarizing over time, state leaders likely will begin to focus on tailoring the ACA to their own circumstances and values.

The ACA opens several paths for state innovation through the Section 1332 waiver option. The provision was championed by Democratic U.S. Sen. Ron Wyden of Oregon and has been dubbed the “Wyden waiver,”5 though other senators provided key support, including Bernie Sanders of Vermont, Barbara Boxer of California and Mary Landrieu of Louisiana. Sen. Wyden, in promoting the waiver concept, rejected “one-size-fits-all” health reform, believing that Oregon could “come up with innovative solutions that the Federal government has never had the flexibility or will to implement.”6 Under Section 1332, states can opt out of or modify one or more of the following key ACA provisions beginning in 2017:

  • the individual coverage mandate and related penalties;
  • the employer coverage mandate and related penalties;
  • creation of health insurance marketplaces;
  • restrictions on premiums and benefit design in the nongroup and small-group insurance markets;
  • advance premium tax credits, or subsidies; and
  • cost-sharing reductions for low-income people.7

The ACA specifies that these provisions can be waived only if the state has an alternate way to cover at least as many people, provide benefits that are at least as comprehensive with the same or lower cost sharing, and without adding to the federal deficit. Section 1332 also lays out a process for states to submit a “coordinated waiver” that combines a 1332 waiver with one or more other waivers relating to Medicaid, the state Children’s Health Insurance Program (CHIP) or Medicare.

To illustrate the range of potential 1332 coordinated waivers, this analysis explores two far-reaching but very different approaches. The first coordinated-waiver approach would automatically enroll almost all state residents in a universal coverage plan financed by a federal lump-sum payment to the state and a new state payroll tax. The second coordinated-waiver approach would replace marketplace subsidies and the Medicaid and CHIP programs for children and nondisabled adults with a uniform voucher for all low- and moderate-income people to purchase a private health plan.

The goal is not to endorse either approach but rather to explore why states might pursue waivers, to quantify the level of federal and state funding potentially available to support states, and to identify what states would have to do to meet waiver requirements. Additionally, states that pursue a 1332 coordinated waiver would have to spell out many design details only hinted at here, or glossed over entirely, and they would face a new and untested federal approval process.

This analysis focuses on 2017 because it is the first year states could implement a waiver. The general approach is to use projected national spending totals from the Congressional Budget Office (CBO) and allocate the national totals to individual states using a variety of state-level data sources (see Data Source and Technical Appendix). Spending and financing levels for the nation as a whole are reported, as well as separately for each of the five most populous states—California, Florida, Illinois, New York and Texas. While important in their own right, the five states also reflect a range of political cultures and health care systems.

The analysis focuses only on spending for “medical care,” defined to include hospital care, physician and other professional services for physical and mental health care, prescription drugs, imaging, and laboratory services. The analysis excludes spending on care provided in nursing facilities or homes for the developmentally disabled, dental care, and over-the-counter drugs.

State Motivations for a 1332 Waiver

State policy makers could use 1332 coordinated waivers to move their health care systems in many different directions, depending on their perceptions of problems with the current system. Those perceived problems generally fall into three areas:

  • Too much income redistribution and regulation. The ACA was designed to build a health insurance system where nearly everyone is enrolled in a relatively comprehensive and affordable health plan.8 To accomplish this vision, the ACA imposed significant new regulations on nongroup and small-group insurers (e.g. plans must cover essential health benefits, insurers must offer guaranteed issue without medical underwriting); on individuals (e.g. tax penalties for remaining uninsured); and on large employers (e.g. tax penalties for not offering coverage). Making coverage affordable also entailed expanding Medicaid eligibility and offering significant new premium and cost-sharing subsidies for low- and moderate-income people to buy private coverage in the marketplaces. Some view these requirements as overly restrictive and the premium subsidies as overly redistributive.
  • Too little progress toward universal access to health care. The ACA adds new enrollees and federal subsidies to the current system of private health insurance and private, often for-profit providers. This puts the burden of enrolling in and paying for a health plan on individuals and endorses a competitive dynamic where willingness and ability to pay play key roles in determining who gets care. Some view this approach as deepening the reliance of the health care system on the market and falling short of the promise of universal access to needed care.
  • Too distortionary and too little reward for efficiency. Under the ACA, eligibility for Medicaid and premium subsidies phase out at higher incomes, and, as a result, individuals whose income rises will lose some or all of their health care subsidies.9,10 This subsidy phase-out increases what economists refer to as the effective marginal tax rate on labor, which disincentivizes, or “distorts,” individual decisions about whether to work and how much to work.11,12 According to CBO estimates, for a single parent enrolled in a marketplace plan, the phase out of premium and cost-sharing subsidies increases marginal tax rates by 10 to 18 percentage points. The employer penalties in the ACA also have so-called cliffs—discrete cut points—based on employees’ hours worked per week and firm size. These cliffs can distort firms’ hiring and staffing choices. The clearest example of a cliff-driven distortion is employers following the “29er” strategy, meaning that they limit their employees’ weekly hours to 29 to avoid being subject to the ACA employer coverage mandate penalty.13

In addition to labor-market distortions, the ACA expands the “beggar-thy-neighbor” system of health care financing. Employers, states and local governments have a range of options available that could slow health care spending growth and increase the efficiency of the health care system. But the rewards for implementing these changes are blunted, or reversed entirely, under the current system of financing health care. For example, a municipality that prevents overbuilding of health care facilities will end up losing federal inflows from Medicare and Medicaid.14 And, employers that reduce spending on health care for their employees give up the value of the tax exclusion for employer-sponsored insurance. States also face strong financial incentives to increase the flow of funds through their Medicaid program to draw down larger federal matching payments.15 For the most part, the ACA leaves each of these rewards for inefficiency in place and adds a new one: marketplace premium subsidies are based on local premiums, and, as a result, states that manage to reduce marketplace premiums will reduce federal funding inflows for marketplace subsidies.

Current Federal Funding Flows for Health Care

Section 1332 coordinated waivers would entail some reorganization of the roles of the federal and state governments in financing and regulating health care.16 To understand the potential for 1332 coordinated waivers, the first step is to identify and quantify the major existing flows of health care funding. States considering a 1332 waiver also would have to take these flows into account in designing a waiver plan and meeting the requirement for federal budget-neutrality.

Historically, the federal government has funded medical services through four major channels (see Figure 1):

  • Medicare, the federal health program for the elderly and disabled ($603 billion in projected federal outlays in 2017);
  • Medicaid, the joint state-federal health program for the poor ($312 billion in projected federal outlays for medical services in 2017);
  • health coverage for current and former federal employees and veterans, through the Veterans Health Administration, Department of Defense and the Federal Employees Health Benefit Program (VHA/DoD/FEHBP) ($166 billion in projected federal outlays in 2017); and
  • the tax exclusion of employer-sponsored health benefits ($268 billion in projected tax expenditures in 2017). Employer health benefits receive two types of favorable tax treatment: They are excluded from workers’ taxable compensation, and they are treated as a deductible business expense for the employer. This tax treatment creates a tax expenditure, meaning that it reduces federal tax revenues,17 which represents an indirect federal subsidy of employer-sponsored health benefits.

The ACA has increased the flow of federal funds through the Medicaid program, and the ACA creates two new federal flows of funds for health care that may be central to a 1332 waiver proposal:

  • premium and cost-sharing subsidies for individuals enrolled in qualified health plans through health insurance marketplaces ($95 billion in projected federal outlays in 2017);
  • individual mandate penalties, which will be paid to the federal government by certain uninsured individuals ($4 billion in projected federal revenues in 2017), and employer mandate penalties, which will be paid to the federal government by certain large employers that do not offer an affordable health plan to their workers ($12 billion in projected federal revenues in 2017).

Several other ACA provisions increase federal revenues—for example, the additional Medicare Hospital Insurance tax and the excise, or Cadillac, tax on high-premium employer-sponsored health plans, but these revenues are not obviously within the scope of a 1332 coordinated waiver and so are not quantified in this analysis.18

Three facts are worth noting regarding federal health care funding flows under the current system:

Total federal funding for medical services is projected be $1.44 trillion in 2017, or about $4,400 per person (see Figure 2), accounting for 47 percent of all spending on medical services (see Table 1). Among the five most-populous states, the federal funding share varies over a narrow range, from 43 percent in New York to 50 percent in California.19 The federal share is similarly high across these states, despite their different political climates and health care systems, reflecting that states with high federal spending through one type of financing may have low spending through another.

Florida and Texas, for example, have chosen not to implement the ACA’s Medicaid expansion for childless, nondisabled adults, and, as a result, their federal spending per capita on Medicaid is lower than in the other states. But, partly as a result of not expanding Medicaid, Florida and Texas have relatively high federal spending per capita on marketplace premium and cost-sharing subsidies.

The tax exclusion for employer-sponsored health plans represents an enormous flow of federal funding for health care—$268 billion in 2017—but is prone to being overlooked and misunderstood. The tax exclusion exemplifies what Suzanne Mettler has dubbed the “submerged state,”20 because it is buried in the tax code and largely invisible to average Americans. While the ACA individual mandate penalties are highly visible—uninsured individuals in 2017 will owe about $700, or 2.5 percent, of income, whichever is greater—the tax expenditure related to employer-sponsored health benefits is defined by an absence of taxation of a major form of worker compensation.

In the context of 1332 coordinated waivers, the size of the federal tax exclusion for employer health benefits is important in at least three ways. First, the tax exclusion represents a major potential source of federal revenue. A state seeking to offset the costs of waiving the employer mandate, for example, could easily achieve federal budget-neutrality by limiting the tax exclusion for employer-sponsored insurance. The ACA already limits the tax exclusion somewhat by applying the Cadillac tax beginning in 2018, but a state could strengthen that provision. The 1332 coordinated waiver could limit the tax exclusion and raise federal revenues without directly amending the federal Internal Revenue Code or other federal tax provisions. For example, a state could implement an excise tax on high-premium plans, similar to the Cadillac tax, but with a lower attachment point. Federal budget analysts typically assume that employers would respond to that type of provision by reducing employer-sponsored health benefits and increasing taxable wages, resulting in higher federal tax revenues.

Second, for states that impose a personal income tax, limiting the tax exclusion for employer-sponsored plans will increase state tax revenues, which could provide some funding for a 1332 waiver plan—these impacts on state tax revenues have not been estimated in this analysis, but they would be important to quantify when developing a coordinated waiver plan.

Third, the size of the federal tax expenditure reflects the degree to which the current financing of health care is entwined with employment-based benefits. There have been major policy proposals that would shift financing of health care away from employment-based benefits, such as the Healthy Americans Act of 2009.21 Such proposals entail significant upheaval, however, and make it nearly impossible for policy makers to promise that currently insured individuals would be able to keep their health plan and their doctor.

The ACA’s individual mandate and employer penalties have significant impacts on whether and how individuals obtain coverage,22 but they represent minuscule shares of the overall flows of federal financing. For a state developing a 1332 waiver plan, modifying or eliminating the individual and employer penalties would not require a major new funding source. For a state seeking to waive these penalties, the challenge would be designing an alternate approach that would cover the same number of individuals as under the ACA.




Two Flavors of State Options

The state-based single-payer plan. One approach that states might take under a section 1332 coordinated waiver is to establish a system of tax-financed universal coverage for all residents. To estimate the financing required for such a plan, the following assumptions were made:

  • The state would automatically enroll almost all residents in a comprehensive plan covering medical services but excluding long-term care, dental care and over-the-counter medications. The only residents who would not be enrolled in the state plan would be federal employees, veterans choosing to receive care through the VHA, and members of the military and their dependents—these excluded populations would continue to receive their current federal health benefits.
  • Individuals would not pay a premium to enroll in the basic state plan.23
  • Employer-sponsored health plans for state residents would be discontinued, although employers could choose to continue to sponsor benefits outside the scope of the state plan, such as dental care and long-term care insurance.
  • The state plan would enroll individuals of all ages, including Medicare beneficiaries, the uninsured and the disabled.
  • Individual and employer mandates and penalties would be waived in the state.
  • The federal government would provide lump-sum financing to the state equal to what the federal government would have spent in that state under the current system for Medicare and Medicaid (excluding long-term care) and marketplace subsidies, plus the value of the tax expenditure for employer-sponsored health plans.
  • Federal lump-sum financing would be reduced by an amount equal to the lost federal revenues from waiving the individual and employer penalties.
  • To finance the plan, the state would use its existing spending on Medicaid and CHIP and also impose a new payroll tax on state residents (and possibly employers) to cover the difference between state plan costs and the lump-sum federal funding.
  • The state would provide low-income individuals with an expanded scope of benefits, including services such as non-emergency transportation, consistent with each state’s current Medicaid benefit package.24

For two key features of the state plan—actuarial value of the basic plan, or the average share of covered medical services paid for by the plan, and whether there is any reduction in total medical spending—the analysis includes a base-case scenario and alternative scenarios. (See Table 2 for an illustration of funding for the base-case scenario and Figure 3 for a summary of the payroll tax rate required to finance the base-case and alternative scenarios.)

In the base-case scenario, the state plan would have an actuarial value of 80 percent, which is equivalent to a gold plan in the current marketplaces and roughly equal to the actuarial value of the Medicare benefit. Alternative scenarios include a platinum-plus basic plan with an actuarial value of 94 percent and a silver basic plan with an actuarial value of 70 percent.25 Under all of the scenarios, the actuarial value of the plan was assumed to be boosted for lower-income people, consistent with ACA marketplace cost-sharing subsidies and Medicaid cost-sharing limitations, and so the differences would only affect middle- and higher-income people.26

In the base-case scenario, total spending on medical care would be the same as under the current system. The base-case scenario can be thought of as assuming no change in administrative costs, no change in the quantity of medical services provided and no change in the average rates paid to providers. In the alternative scenarios, total medical care spending would be reduced by 10 percent across the board. This analysis does not specify the mechanism to achieve the reduction and instead focuses only on the implications for funding the single-payer plan.27

There are many, many additional details that any state considering such an approach would have to work through. These include whether and how to enroll noncitizen legal residents and undocumented immigrants; whether to operate a single state-run plan or a marketplace with competing private plans or some hybrid; and whether and how to offer supplemental coverage, or “buy-ups,” to achieve actuarial values above the basic plan (i.e. 80% in the base-case scenario).

Providing a tax-financed universal coverage plan would make it unnecessary for employers to sponsor and contribute to health plans for their employees. As a result, employers would likely increase employee wages, and, at least in the short term, retain some of the savings on health benefits as increased corporate earnings. This analysis assumes that the additional federal taxes paid on higher wages and corporate earnings would be returned to the state as part of the lump-sum federal financing.

Employees would face a new payroll tax to finance the state-based plan, with rates in the base-case scenario ranging from 8.6 percent in California to 17.6 percent in Florida. The very high rate in Florida reflects the fact that that state has a relatively small earnings base over which to spread the financing of a state-based plan. Those tax rates may sound shockingly high, but it is important to keep in mind that they are calculated to finance exactly the same level of medical care spending as under the current system. In other words, state residents are already paying the amount needed to finance a single-payer system, but they are paying through less-transparent mechanisms—for example, employer premium contributions—and less redistributive than a payroll tax.

The rate of the new payroll tax could be substantially higher or lower under different single-payer plan designs (see Figure 3). For example, increasing the actuarial value of the base plan from gold (80% actuarial value) to platinum-plus (94% actuarial value) would increase the required tax rate by 3 to 4 percentage points, depending on the state. On the other hand, limiting the base plan to silver (70% actuarial value) would reduce the required tax rate by 2 to 3 percentage points. If the base plan were silver, and the state managed to achieve a 10 percent across-the-board spending reduction on medical care, the required payroll tax rate would be reduced by 4 to 7 percentage points.

One important question regarding the single-payer plan is whether the payroll tax would be levied just on employees, or just on employers, or on both. This analysis assumes the payroll tax would be levied solely on employees. This differs from the Medicare Hospital Insurance tax, which is levied equally on employees and employers. It also differs from the Vermont single-payer financing plan, which would have combined a premium tax on individuals with a payroll tax levied on employers. There are advantages and disadvantages to levying the tax only on employees. The advantage is that it eliminates entirely the tax advantage for health care spending, which addresses the tax incentives for inefficiency and is consistent with including the full value of the federal tax exclusion in the federal lump-sum funding to the state.28 The disadvantage is that it would be more disruptive than a payroll tax applied both to employers and employees

The state-based voucher plan. A very different state-based approach would be to combine and convert marketplace premium subsidies and Medicaid funding for children and nondisabled adults into a unified voucher program for low-income individuals and families. One of the rationales for this approach would be to lessen the institutional barriers separating Medicaid and private coverage and the distortionary tax provisions designed to encourage employer-sponsored coverage. The vouchers, therefore, could be applied equally to the purchase of a nongroup health plan or an employer-sponsored plan, and the employer mandate would be eliminated. To estimate the available funding for such an approach, the following assumptions were made:

  • Each state would maintain its existing Medicaid program for the aged and disabled, and the Medicare program would continue as is, along with existing federal health programs for veterans, federal employees, and members of the military or their dependents.
  • For children and nondisabled adults, the Medicaid program would be converted into a voucher program for the purchase of private health insurance.
  • Federal funding for Medicaid and CHIP for children and nondisabled adults, along with funding for marketplace premium and cost-sharing subsidies, would fund the voucher program.
  • Low-income people would be auto-enrolled in a private nongroup health plan, with the voucher amount credited toward the premium.
  • Individuals could switch to a competing nongroup health plan or an employer-sponsored plan and apply the voucher to the cost of the plan.
  • Voucher-eligible individuals who enroll in an employer-sponsored health plan could apply the voucher amount to employer and employee premium contributions.

Voucher amounts in excess of premiums would be automatically credited to a dedicated health care savings account.

  • The employer mandate would be eliminated.
  • Vouchers would only be available to U.S. citizens.
  • The voucher amount would be set based on the individual’s family income and would be set equal to a sliding-scale proportion of the premium of a benchmark plan (see Table 3).
  • The premiums for benchmark plans in each state were projected based on the premiums for the second-lowest cost marketplace silver plan in 2015. Marketplace premiums were assumed to grow from 2015 to 2017 at a 4 percent annual rate and were adjusted based on the age and sex of the projected population receiving the vouchers.
  • In some states, the available federal and state funding is not adequate to provide vouchers for the entire population below 400 percent of the federal poverty level. For the purposes of modeling vouchers in these states, an income cutoff was identified up to which vouchers could be offered with the available funding.29

It is important to note that the voucher approach described in this analysis differs in key ways from the approach taken in the ACA. Under the ACA, individuals are prevented from receiving a premium tax credit and at the same time also benefiting from the tax exclusion for employer-sponsored coverage (so-called “double dipping,” see Internal Revenue Code 36B(c)), and firms that attempt to use pre-tax contributions to reimburse employees’ marketplace premiums face stiff penalties.30 The primary purpose of these prohibitions is to limit the federal costs of the program. In contrast, the approach described here provides vouchers for individuals enrolled in employer-sponsored coverage even if premium payments are excluded from taxable income.31 The rationale for taking this approach is twofold. First, it creates a more progressive approach to subsidizing health insurance for low-income workers, by layering the voucher on top of existing tax benefits rather than forcing employers into an either-or choice. Second, allowing vouchers for employer-sponsored plans does not create additional federal costs because federal outlays are fixed based on what they would have been without the 1332 waiver.

The ACA is also designed to encourage employers to continue to offer health coverage by penalizing large employers that do not offer affordable coverage. The voucher approach described in this analysis takes a different approach and eliminates the penalties for employers dropping coverage for some or all of their workers. The rationale for eliminating the employer mandate is that it avoids the burden of administering the employer mandate, it increases the progressivity of the subsidies for health care (note that the economic incidence of employer mandate penalties falls on workers), and it increases the fluidity between employer-sponsored and nongroup coverage.

The following examples illustrate how voucher amounts likely would vary across states. For instance, the voucher amount for a 40-year-old nonsmoker below the poverty level would range from $4,500 in Illinois to $5,700 in New York (see Table 4). The variation in vouchers reflects the differences in the premiums for marketplace plans currently offered in these states.

In two states—Illinois and New York—projections indicate that funding would be sufficient to provide vouchers up to 400 percent of poverty. In California, the available funding is projected to fall slightly short, covering 92 percent of the amount required to provide vouchers up to 400 percent of poverty ($47 billion out of $51 billion). Compared to New York and Illinois, California is similar in having implemented the ACA Medicaid expansion, but California has a larger share of its population in lower-income groups. In Texas and Florida, the available funding falls significantly short and would only be sufficient to cover vouchers up to 190 percent of poverty.

New York could fund vouchers up to 400 percent of poverty, while Texas could not (see Figure 4). The difference reflects the fact that New York currently has a much more expansive Medicaid/CHIP program and is projected to draw much more federal Medicaid/CHIP funding for children and nondisabled adults than Texas in 2017 ($12 billion vs. $8 billion). New York’s larger federal Medicaid funding occurs despite the state having a smaller overall population than Texas, a smaller share of its population in poverty, and receiving a lower federal Medicaid matching rate.

New York and Illinois, if they chose to pursue a state-based voucher plan, would likely have a much easier time than other states meeting the Section 1322 coverage adequacy requirements. The other three states would have to take additional steps to design a voucher system that would extend up to 400 percent of poverty and meet Section 1332 requirements without additional funding. One approach would be to limit eligibility for the vouchers to individuals who are not enrolled in or do not have access to employer-sponsored coverage. This approach would limit spending on vouchers but would have notable disadvantages: It would require administration of a complex regulatory scheme, and it would undermine the progressivity of the subsidies by forcing employers into an either-or choice between benefitting from the employer tax exclusion vs. making vouchers available to their employees. Another approach states could take is to generate additional federal or state funding, either by limiting the tax exclusion for employer-sponsored plans or generating new state revenues. Health care cost-containment efforts, if successful, could also allow funding for vouchers to extend further up the income scale. The voucher amounts were simulated based on 2015 marketplace premiums, and it is possible that health plans could reduce premiums below those currently offered in the marketplaces. Health plans could, for example, negotiate lower provider-payment rates—closer to Medicaid, for instance, than commercial rates—and develop ultra-narrow provider networks, although consumers’ appetite for these products remains uncertain.

Among the five largest states, Texas and Florida have chosen not to expand Medicaid programs for childless, nondisabled adults as called for under the ACA, and, as a result, these two states have the most limited ability to provide vouchers with current funding. The motivations behind not expanding Medicaid expansion are complex, but at least in part it reflects a desire to avoid excessive federal control over a state’s health care system. Paradoxically, by choosing not to expand Medicaid limits, a state limits the federal funds flowing to the state and narrows options for using Section 1332 to remake its health care system more to its liking.

Others have projected that in 2022 Texas and Florida each would be giving up around $10 billion in federal funding if they do not implement the ACA Medicaid expansion.32 Additional federal funding of that magnitude would close much of the gap in financing the voucher plan in Texas and Florida. In negotiating the terms of a coordinated waiver, states could request that federal funding be set as if they had implemented the ACA Medicaid expansion—whether federal officials would approve such a shortcut is uncertain.

The voucher plan would require working through and specifying numerous important policy details. For example, one key question is how to handle people auto-enrolled into nongroup coverage who opt out and become uninsured, or who fail to pay the balance of the premium owed after the voucher is applied. The assumption in this analysis is that the federal individual mandate penalties would continue to apply to discourage the practice. But, the state might consider applying a multi-year lockout period to individuals who decline coverage or fail to pay premiums—during that time, individuals would not be eligible for guaranteed-issue open enrollment.

Another key question is whether the voucher amounts would be adjusted based on age, health status, local health care costs and premiums, smoking status, and so on. And, would the voucher amounts be fixed before each tax year or adjusted during the year based on changes in income? Would insurers be required to offer guaranteed-issue coverage without exclusions for preexisting conditions, or would these regulations be rolled back?







The two state-based waiver options outlined in this analysis are likely to appeal to polar opposites of the political spectrum. Nevertheless, they share key features.

Both state-based plans would convert open-ended federal payments under the current financing system into a lump-sum federal contribution. Such an approach would better position a state and its residents to reap the rewards of system-wide cost-containment efforts. Ignoring political polarities for a moment, it is possible to imagine combining the state-based single-payer plan or the voucher plan with any number of cost-containment approaches. These approaches, for example, might include stricter enforcement of antitrust law and increased competition among health plans and providers, elimination of the tax exclusion for employer-sponsored health benefits, increased transparency in prices, crackdowns on fraudulent providers and exploitative billing practices, state-based rate setting for hospitals and physicians, and expanding the scope of practice for non-physician providers.

Both state-based plans would waive the employer mandate and penalties, and the universal coverage plan also would waive the individual mandate. The penalty payments from these mandates represent only minor flows of financing for health care, but they have drawn a disproportionate share of the criticism and ill will directed at the ACA.

Sources of Uncertainty for States

Arkansas, Hawaii and Vermont have initiated the process of developing a 1332 waiver application.33,34,35 But, no state has actually submitted a waiver application and negotiated the terms with the federal government. Vermont was the state furthest along the path to submitting a 1332 waiver application but recently abandoned the effort because of perceived negative economic effects.36,37 Without any precedent to go by, states face significant uncertainties in how the federal government will manage the 1332 waiver process.

One obvious source of uncertainty is the conceptual and technical approach that federal agencies will take in determining whether a waiver application meets the requirements of section 1332. The Office of Management and Budget (OMB) will be tasked with determining whether a waiver would be budget-neutral for the federal government. Whether a state meets that test will depend in part on the proposal details and on the methods and assumptions OMB uses to score the proposal. For example, will states get credit for increased federal income and payroll tax revenues from increased wages if spending on employer-sponsored health plans drops? (In this analysis, the assumed answer is yes.) Will states get credit for spending on expanded Medicaid if they haven’t yet expanded their program? (The assumed answer is no.) Will a “dynamic scoring” approach be used that takes into account any changes in labor force participation? (The assumed answer is no.)

Another source of uncertainty is how the Centers for Medicare and Medicaid Services (CMS) Office of the Actuary will determine whether a state plan meets requirements for the scope of benefits covered, the share of total costs paid by the plan vs. out of pocket by the patient, and number of individuals covered.38 For example, would the voucher plan be deemed to provide the same protection to a person below 100 percent of poverty if it covered the premium of a silver plan and provided a contribution to a health savings account? In general, it is unclear whether each of the tests applies on an individual-by-individual basis or at a more aggregate level.

One last source of uncertainty is the federal Employee Retirement Income Security Act (ERISA), which preempts state-based regulation of employer-sponsored health benefits.39 ERISA “clearly preempts state laws that mandate employer-provided health coverage, dictate what services should be included in such coverage, or to which employees coverage must be available.”40 Neither of the state-based waiver plans described in this analysis would, on their face, violate the ERISA preemption in these ways. But, modifications to ERISA might be necessary to allow state-based plans to proceed.

For example, a multi-state employer might object if a state-based voucher is made available for self-funded plans but only if they meet certain conditions. Or, a multi-state employer might object if a state-based single-payer plan creates economic incentives that more or less compel the employer to drop its employee health benefits for residents of one state. The path that would provide the most certainty to states would be for Congress to pass legislation delineating a “safety zone” within which 1332 waiver plans could not be challenged under ERISA. The administration, in reviewing and approving a 1332 waiver plan, could also offer a legal rationale for the plan being consistent with ERISA. The most uncertain path through the ERISA minefield would be for a state to propose and implement a 1332 waiver plan, and then deal with ERISA-based court challenges as they arise.


1. Johnson, Jenna, and Mary Pat Flaherty, “As Md. Rebuilds Its Health Insurance Exchange, Lots of Pressure and Little Room for Missteps,” The Washington Post (May 3, 2014).

2. Arkansas Center for Health Improvement, PPACA Section 1332: Waiver for State Innovation (2015).

3. Kaiser Commission on Medicaid and the Uninsured, A Historical Review of How States Have Responded to the Availability of Federal Funds for Health Coverage (August 2012).

4. U.S. Senators Richard Burr, Orrin Hatch and Fred Upton, The Patient Choice, Affordability, Responsibility, and Empowerment Act (2015).

5. McDonough, J. E., “Wyden’s Waiver: State Innovation on Steroids,” Journal of Health Politics, Policy and Law, Vol. 39, No. 5 (2014).

6. Wyden, Ron, Letter to Bruce Goldberg, MD (Aug. 24, 2010).

7. U.S. Department of Health and Human Services (HHS), “Application, Review, and Reporting Process for Waivers for State Innovation,” Federal Register, Vol. 77, No. 38 (Feb. 27, 2012).

8. Note that “nearly everyone” does not include undocumented immigrants and only includes legal residents who have met residency requirements. For a discussion of the ACA and provisions relating to immigrants, see Kenney, Genevieve M., and Michael Huntress, The Affordable Care Act: Coverage Implications and Issues for Immigrant Families, HHS, Office of the Assistant Secretary for Planning and Evaluation (April 2012).

9. For a narrative account, see Goodnough, Abby, “Under Health Care Act, Many Tax Filers Are Discovering Costly Complications,” The New York Times (March 21, 2015).

10. The ACA, as written by Congress, required that states expand Medicaid eligibility to 138 percent of the federal poverty level (FPL) and made exchange subsidies available for individuals above that point. But, as a result of the 2012 U.S. Supreme Court decision, the Medicaid expansion was made optional for states. In states that have not expanded Medicaid, a cliff exists at 100 percent FPL—individuals with incomes below that point are ineligible for exchange subsidies. The result of that cliff is to encourage workers to increase their incomes to become eligible for the exchange subsidies.

11. Congressional Budget Office, Effective Marginal Tax Rates for Low- and Moderate-Income Workers (November 2012).

12. Mulligan, Casey B., Average Marginal Labor Income Tax Rates under the Affordable Care Act, National Bureau of Economic Research, Working Paper No. 19365 (August 2013).

13. Dettling, Aaron L., So, You’re Thinking About Running the 29er Strategy, Balch and Bingham, LLC (June 17, 2013).

14. White, Chapin, and Paul B. Ginsburg, Working at Cross Purposes: Health Care Expansions May Jumpstart Local Economies but Fuel Nation’s Fiscal Woes, Center for Studying Health System Change, Washington, D.C. (August 2011).

15. U.S. Government Accountability Office, Medicaid Financing: States’ Increased Reliance on Funds from Health Care Providers and Local Governments Warrants Improved CMS Data Collection (July 2014).

16. The flows of funds among households, the federal government and state governments are extraordinarily complex. For a detailed discussion of these flows in one small state, see Eibner, Christine, et al., The Economic Incidence of Health Care Spending in Vermont, RAND Health (January 2015).

17. Joint Committee on Taxation, Estimates of Federal Tax Expenditures for Fiscal Years 2012-2017 (Feb. 1, 2013).

18. The ACA increases the Medicare Hospital Insurance tax, establishes the new “Cadillac” tax on high-premium employer-sponsored plans beginning in 2018, and establishes new fees on health insurance plans and prescription drug and device manufacturers. For details, see Congressional Research Service, Health-Related Revenue Provisions in the Patient Protection and Affordable Care Act (PPACA) (May 3, 2010).

19. The federal funding share calculated (47%) is much higher than the federal government’s share of national health expenditures projected for 2017 by the Centers for Medicare & Medicaid Services (CMS) (29%). That difference in federal funding shares is due to two key differences in methodology. First, CMS’ calculation of federal spending only includes Medicare outlays from general revenues and treats Medicare outlays that are funded by payroll taxes as spending by households and employers. Second, CMS’ calculation does not take into account the value of the federal tax exclusion for employer-sponsored health plans. CMS’ calculations also include a broader set of spending (national health expenditures, rather than spending on medical care), although that difference does not necessarily significantly affect the federal government’s share. CMS projects that total national health expenditures will equal $3.6 trillion in 2017 and that the federal government will sponsor $1.0 trillion of that amount (29%). See Table 16, CMS, NHE Projections 2013-2023 � Tables (2014).

20. Mettler, Suzanne, The Submerged State: How Invisible Government Policies Undermine American Democracy, University of Chicago Press, Chicago, Ill. (2011).

21. U.S. Congress, Healthy Americans Act (Feb. 5, 2009).

22. Auerbach, David, et al., “Will Health Insurance Mandates Increase Coverage?: Synthesizing Perspectives From Health, Tax, And Behavioral Economics,” National Tax Journal, Vol. 63, No. 4, Part 1 (December 2010).

23. An alternative approach to financing the single-payer plan would include premium payments by state residents, similar to Medicare, which is financed with revenues from taxes on wage income plus premiums paid by enrollees. Requiring premium payments for a single-payer plan would be similar to imposing a head tax on state residents, which would reduce the tax progressivity of the financing of the plan. Requiring premium payments raises several policy questions, among them: 1) how large would premium revenues be compared to payroll-tax revenues? 2) would premiums be collected through the tax system and, if so, how would premiums be collected from non-filers? 3) would nonpayment of premiums result in individuals being dropped from the single-payer plan? 4) would premiums be adjusted for age, income or health status, and, if so, how? Those policy questions are clearly answerable, but they make the modeling too complex for the scope of the current study.

24. In all states, the Medicaid Early and Periodic Screening, Diagnosis and Treatment (EPSDT) program provides some benefits beyond the essential health benefits package, including dental care for children under age 21. Other additional Medicaid benefits include non-emergency medical transportation. See Baumrucker, Evelyne, and Bernadette Fernandez, Comparing Medicaid and Exchanges: Benefits and Costs for Individuals and Families, Congressional Research Service (June 26, 2013).

25. Vermont’s proposed single-payer plan would have provided an actuarial value of 94 percent. For details, see Shumlin, Peter, Green Mountain Care: A Comprehensive Model for Building Vermont’s Universal Health Care System, State of Vermont (Dec. 30, 2014).

26. A base plan with 70 percent actuarial value would entail significantly higher cost sharing relative to the current Medicare benefit package and relative to typical employer-sponsored plans. That gap in actuarial values could result in the coordinated waiver failing to meet the requirements in the ACA and would undoubtedly raise objections among current enrollees in more-generous plans.

27. For examples of the policy options that a state might consider to reduce overall medical spending, see Eibner, Christine, et al., Controlling Health Care Spending in Massachusetts: An Analysis of Options, RAND (August 2009).

28. To see that the tax advantage is eliminated entirely under the employee-only approach, suppose the residents of a state are considering whether to increase the generosity of the state-based plan and increase the payroll tax accordingly to cover the additional spending. Under the employee-only approach, increasing health spending and the payroll tax would not reduce taxable income or generate any savings for state residents in federal income tax payments or corporate tax payments. Suppose, instead, that the payroll tax is split evenly between the employer and employee. If total compensation is assumed to remain, then increasing the payroll tax rate would result in a reduction in taxable compensation to the employee, which, in turn, generates savings for state residents in federal income tax payments.

29. Setting an income limit for vouchers below 400 percent of the federal poverty level simplifies the modeling of the voucher approach but would likely not be acceptable to the federal government under the terms of a coordinated waiver.

30. See Internal Revenue Service, Employer Health Care Arrangements (April 16, 2015).

31. There is a precedent, in the Medicare Part D prescription drug benefit, for providing two forms of subsidies for the same health care spending. When originally enacted in the Medicare Modernization Act of 2003, Part D provided substantial subsidies to employers who provided prescription drug benefits to their retirees and, at the same time, allowed employers to deduct total spending on retiree drug benefits (including the subsidies) for purposes of corporate income taxes. Additionally, the value of retiree drug benefits was excluded from retiree income. Under the ACA, employers continue to receive Part D subsidies, but they are no longer permitted to deduct the Part D subsidies from their corporate income.

32. Glied, Sherry, and Stephanie Ma, How States Stand to Gain or Lose Federal Funds by Opting In or Out of the Medicaid Expansion, Commonwealth Fund (December 2013).

33. State of Hawaii, State Innovation Waiver; State Innovation Waiver Task Force; Appropriation (2014).

34. Public Consulting Group, Section 1332 Waivers and the Future of Arkansas Healthcare Innovation, Arkansas Health Insurance Marketplace Board and Legislative Oversight Committee (Dec. 19, 2014).

35. Ollove, Michael, “Vermont Tries to Go Beyond Obamacare,” Governing (Aug. 11, 2014).

36. McArdle, Megan, “If Single Payer Can’t Work in Vermont…” BloombergView (Dec.23, 2014).

37. Shumlin, Peter, Gov. Shumlin Details Health Care Financing Report to Business and Consumer Advisory Councils (2014).

38. From Section 1332(b): “(1) IN GENERAL.—The Secretary may grant a request for a waiver under subsection (a)(1) only if the Secretary determines that the State plan—(A) will provide coverage that is at least as comprehensive as the coverage defined in section 1302(b) and offered through Exchanges established under this title as certified by Office of the Actuary of the Centers for Medicare & Medicaid Services based on sufficient data from the State and from comparable States about their experience with programs created by this Act and the provisions of this Act that would be waived; (B) will provide coverage and cost sharing protections against excessive out-of-pocket spending that are at least as affordable as the provisions of this title would provide; (C) will provide coverage to at least a comparable number of its residents as the provisions of this title would provide; and (D) will not increase the Federal deficit.”

39. Congressional Research Service, Regulation of Health Benefits Under ERISA: An Outline (Oct. 20, 2008).

40. Monahan, Amy, Pay or Play Laws, ERISA Preemption, and Potential Lessons from Massachusetts, University of Missouri-Columbia School of Law Legal Studies, Research Paper (2007).

-Columbia School of Law Legal Studies, Research Paper (2007).