CBO Releases Preliminary Analysis of Graham-Cassidy Bill

Preliminary Analysis of Legislation That Would Replace Subsidies for Health Care With Block Grants

September 2017

At the request of the Chairman of the Senate Budget Committee, the Congressional Budget Office and the staff of the Joint Committee on Taxation (JCT) have analyzed the direct spending and revenue effects of legislation sponsored by Senators Graham, Cassidy, Heller, and Johnson that would replace certain federal subsidies for health care with block grants to states. Specifically, the agencies analyzed H.R. 1628, an amendment in the nature of a substitute [LYN17744], posted on September 25, 2017, on Senator Cassidy’s website.¹

In the short time available, rather than provide the point estimates that are typical in such analyses, the agencies have been able to assess only whether any reductions in the deficit stemming from the legislation as a whole (and from its two titles individually) would exceed certain thresholds and to qualitatively assess its effects on health insurance coverage and market stability.

Over the 2017–2026 period, CBO and JCT estimate, the legislation would reduce the on-budget deficit by at least $133 billion, the projected savings from the House-passed reconciliation bill. (The effects on the deficit were estimated relative to CBO’s March 2016 baseline, as has been done for all legislation related to the 2017 budget resolution.) Those savings would occur mainly because, under the legislation, outlays from new block grants between 2020 and 2026 would be smaller than the reduction in net federal subsidies for health insurance. Funding would shift away from states that expanded eligibility for Medicaid under the Affordable Care Act (ACA) and toward states that did not.

The number of people with comprehensive health insurance that covers high-cost medical events would be reduced by millions compared with the baseline projections for each year during the decade, CBO and JCT estimate. That number could vary widely depending on how states implemented the legislation, although the direction of the effect is clear. The reduction in the number of insured people relative to the number under current law would result from three main causes. First, enrollment in Medicaid would be substantially lower because of large reductions in federal funding for that program. Second, enrollment in nongroup coverage would be lower because of reductions in subsidies for it. Third, enrollment in all types of health insurance would be lower because penalties for not having insurance would be repealed. Those losses in coverage would be partly offset by enrollment in new programs established by states using the block grants and by somewhat higher enrollment in employment-based insurance. Many of the new programs would probably cover people with characteristics similar to those of people made eligible for Medicaid by the ACA.

The decrease in the number of insured people would be particularly large starting in 2020, when the legislation would make major changes to federal funding for Medicaid and the nongroup market. CBO and JCT expect that market disruptions and other implementation problems would accompany the transition to the block grants created by the legislation — despite the availability of funding specifically designated to assist with that transition — given the short time for planning and making changes between now and then.

CBO and JCT would need at least several weeks to provide point estimates of the effects on the deficit, health insurance coverage, and premiums. During that time, the agencies would gather and analyze more information about states’ potential uses of the block grants and the extent to which states might modify rules governing the nongroup market.

Major Provisions of the Legislation

Upon enactment, the legislation would eliminate penalties associated with the requirements that most people obtain health insurance coverage (also called the individual mandate) and that large employers offer their employees coverage that meets specified standards (also called the employer mandate).

Starting in 2018, the legislation would reduce the federal share of Medicaid funding for adults made eligible for that program by the ACA to 90 percent for two years (compared with 94 percent in 2018 and 93 percent in 2019 under current law). It would also allow payments of premiums for certain types of insurance to qualify as medical expenses for health savings accounts and repeal a few of the tax provisions enacted as part of the ACA.

In 2019 and 2020, the legislation would make funding ($10 billion and $15 billion, respectively) available to health insurers to stabilize premiums and promote participation in the nongroup market.

In 2020, the legislation would set a limit, on a per — enrollee basis, on the amount of reimbursement the federal government provides to states for Medicaid, and the growth in per-enrollee payments would be limited to no more than the growth rates of certain price indexes. The following provisions would also take effect:

• Medicaid funding would be eliminated for adults made eligible for that program by the ACA.

• Tax credits for health insurance coverage purchased through the marketplaces established by the ACA and subsidies to reduce cost-sharing payments for certain low-income people would be repealed.

• Funds would be appropriated for block grants to states, with amounts based on historical federal Medicaid funding for adults made eligible for that program by the ACA and historical funding for subsidies provided through the marketplaces and the Basic Health Program.

States would be allowed to modify certain requirements in the nongroup insurance market if the new block grants directly provided some assistance to participants in that market. States could modify requirements that policies include what are known as essential health benefits; restrictions on insurers’ ability to vary premiums on the basis of health status, age, and other factors; and the requirement that insurance sold in the nongroup market generally rely on a single risk pool.

Effects on the Federal Budget Deficit

According to CBO and JCT’s analysis, the legislation would reduce the on-budget deficit over the 2017–2026 period by at least $133 billion, the projected savings from the House-passed reconciliation bill. The agencies made that assessment relative to the March 2016 baseline, which has been the basis for cost estimates related to the 2017 budget resolution. That effect on the deficit would arise mainly from a few budgetary flows that result from the net effect of provisions that would provide funding for block grants, eliminate subsidies and penalties, and reduce Medicaid spending.

The amount that would be appropriated for the new block grants — $1.2 trillion from 2020 to 2026 — is about $230 billion less than the amount in CBO’s March 2016 baseline for the major subsidies over that period that would be eliminated under the legislation. (Those subsidies are for adults made eligible for Medicaid by the ACA and for insurance obtained in the marketplaces and through the Basic Health Program.) That reduction in subsidies would be partly offset by the repeal of penalty payments from the individual and employer mandates, which totaled about $200 billion over the 2020–2026 period in CBO’s March 2016 baseline. The net result of those flows — providing the new block grants and eliminating those subsidies and penalties — would be to reduce the deficit even if all of the grant funding was spent by 2026. However, CBO estimates, at least $150 billion of the $1.2 trillion in budget authority for the block grants would not result in outlays by 2026, further reducing the deficit.

Several other budgetary flows would roughly offset one another. Eliminating the subsidies and penalties would have other effects that would work to increase the deficit significantly from 2020 to 2026. Among those other effects, according to CBO and JCT’s estimates, the largest would be a reduction in revenues because more people would obtain insurance coverage through their employer and hence would receive more of their income in nontaxable health benefits and less in taxable wages. Two other flows would work to decrease the deficit: Reductions in Medicaid spending over the 2020–2026 period for people besides those made eligible by the ACA and the total effect in 2018 and 2019 of the legislation’s provisions dealing with insurance coverage.

The provisions other than those directly affecting health insurance coverage would increase on-budget deficits by $22 billion, on net, over the 2017–2026 period (see Table 1, at the end of this document). Unlike the estimate for overall deficit reduction, the estimates of the effects of those noncoverage provisions are point estimates, which CBO and JCT were able to develop because the analysis was more straightforward.

Each title of the bill would, by the agencies’ estimates, reduce the on-budget deficit by more than $1 billion over the 2017–2026 period. Because no funding would be provided for the block grants after 2026, the annual reduction in the deficit would be much greater after that year. Enactment would not increase net direct spending or on-budget deficits in any of the four consecutive 10-year periods beginning in 2027.

Analytical Approach

In the short time available, CBO and JCT could not complete a full analysis of the effects of this legislation on the federal budget that was built up from effects on health insurance coverage and premiums (as the agencies typically construct their estimates). Nevertheless, building on work done earlier this year on many related proposals, the agencies assessed whether this legislation would reduce the deficit by an amount equal to or greater than the projected savings from the House-passed reconciliation bill. In particular, CBO and JCT drew upon their projections of the effects of related provisions of the Better Care Reconciliation Act, the Obamacare Repeal Reconciliation Act, the American Health Care Act, and other recent proposals.²

To assess how the new grant funding might be used and how quickly funds would be spent, the agencies relied on information provided by states and insurers across the country and by the Department of Health and Human Services, consultation with CBO’s Panel of Health Advisers, analysis of allowable uses of the funds by states, and historical experience with large federal grant programs.

Under this legislation, states would have enormous flexibility, and there are many possible responses to consider. To quickly estimate whether the deficit reduction stemming from this legislation would exceed the projected savings from the House-passed reconciliation bill, CBO and JCT focused on two sets of responses that would have a reasonable chance of occurring and that would produce deficit reductions that were on the small side of possible outcomes:

• Funds from the new block grant would be spent relatively quickly; and

• Conditions in the nongroup market would cause significantly more employers to offer nontaxable health benefits than under current law, reducing revenues substantially.

This approach differs from CBO and JCT’s usual approach because the agencies generally aim to provide point estimates in the middle of the distribution of possible outcomes. The agencies typically assess legislation making major changes to subsidies for health insurance by estimating the budgetary effects associated with different possible responses to the incentives in the legislation and produce a central point estimate by averaging the effects of those responses. In that average, a set of responses receives more weight when it applies to a higher proportion of people. The approach used in this analysis allowed the agencies to examine many fewer sets of possible responses. Providing a point estimate of the savings would take much longer.

Although CBO and JCT are confident that their estimates of the deficit reduction stemming from the legislation would exceed certain thresholds, the ways in which individuals, employers, states, insurers, doctors, hospitals, and other affected parties would respond to the changes made by this legislation are all difficult to predict. Thus, the estimates themselves are uncertain. (Point estimates would be even more uncertain.)

The baseline used for this analysis — CBO’s March 2016 baseline projections, with adjustments for legislation that was enacted after that baseline was produced — was chosen on the basis of consultation with the budget committees. Those projections underlie the 2017 budget resolution, which specified reconciliation instructions and was the basis for the deficit reduction goals stated in the resolution.³

Distribution and Use of Grants to States for Market-Based Health Care

In general, the allocation of the grants under the legislation would shift funding away from states that have already expanded eligibility for Medicaid under the ACA and toward states that have not. In 2020, both groups of states would receive about 10 percent less funding from the new block grants than the amount in CBO’s March 2016 baseline arising from two sources: Medicaid funding for people made eligible for that program by the ACA, and subsidies for insurance purchased through marketplaces or the Basic Health Program. By 2026, under the legislation, states that have already expanded Medicaid under the ACA would receive about 30 percent less funding than the amount projected in the baseline, and other states would receive about 30 percent more, CBO and JCT estimate. (Those estimates are averages in which each state receives equal weight; effects would differ among states.)

Under the legislation, grants would be allocated under a formula in which the basis shifts over time from historical spending to the share of low-income people in a state. In 2020, $146 billion would be allocated on the basis of how much states and their residents received from the two sources of subsidies during four recent calendar quarters. In 2026, $190 billion would be allocated to states according to their share of residents with income between 50 percent and 138 percent of the federal poverty level (FPL), with adjustments for factors related to the health of those residents and for other factors affecting states’ health care costs.⁴

Additional funds would be provided for states with low population density ($1.5 billion in 2020 and $1.25 billion in 2021), states that expanded eligibility for Medicaid under the ACA ($1.5 billion in 2020 and $1.25 billion in 2021), and states that did not undertake that expansion ($3 billion in 2020 and $2.5 billion in 2021). After 2026, under the legislation, funding for the new block grants would stop.

Allowable Purposes of the Grants

The grants to states could be used for the following seven purposes:

• To help people purchase nongroup coverage if they have or are projected to have high health care costs and do not have access to health insurance offered through an employer;

• To enter into arrangements with health insurers to reduce premiums in the nongroup market;

• To provide payments to health care providers;

• To provide assistance to reduce out-of-pocket costs, such as copayments, coinsurance, and deductibles, in the nongroup market;

• To help people (in addition to those targeted by the first purpose) purchase nongroup coverage;

• To provide Medicaid benefits through private insurers for people besides those made eligible by the ACA (with no more than 20 percent of the funds being used for this purpose); and

• To establish or maintain a program to provide health care services through arrangements with managed care organizations to people who are not eligible for Medicaid or the Children’s Health Insurance Program (CHIP).

At least half of the funds would be required to be used to provide assistance to people with income between 50 percent and 300 percent of the FPL.

States’ Uses of the Grants

The flexibility afforded by the new block grants would allow states to experiment with different ways to cover health care costs. CBO and JCT expect that states would learn from one another, and over time, they would adopt practices found to be successful elsewhere that supported their goals for subsidizing coverage (in terms of both the people receiving subsidies and the types of coverage subsidized), regulating coverage, and so on. Although further analysis is needed to estimate what proportion of the funds would be used for which purposes and how quickly the funds would be spent, the agencies have begun to assess the likely uses.

In CBO and JCT’s estimation, most states would eventually make changes in the regulations for their nongroup market in order to stabilize it and would use some funds from the new block grants to facilitate those changes. In the agencies’ assessment, the nongroup market would become unstable if current-law regulations remained in place without substantial subsidies for insurance (and penalties for not having insurance).⁵ For example, if regulations prohibited premiums from being based on one’s health status and healthy people’s insurance was not subsidized, the people who wanted to buy coverage at any offered price would have average health care expenditures so high that offering the insurance would be unprofitable to insurers. Insurers would then not participate in the nongroup market. Therefore, lacking substantial subsidies for the nongroup market, most states would eventually want to modify various insurance market requirements. Under the conditions specified in the legislation, to be allowed to modify market rules, states would have to pay at least small amounts as subsidies for purchasing nongroup coverage.

CBO and JCT expect that many states that have expanded Medicaid would use funds from the new block grants to establish a new program. Because people enrolled in Medicaid as a result of the expansion would, under the legislation, lose eligibility starting in 2020, those states would aim to continue to provide coverage to a similar group of people to the extent allowed by available funds. States would be able to leverage their existing systems to provide that coverage. They would have great flexibility to determine the eligibility criteria, covered benefits, premiums, cost sharing requirements, and other aspects of such a program. Those alternative programs could be structured similarly to current Medicaid programs or very differently.

In addition, to the extent they had additional funds available, some states that have expanded Medicaid would probably use other mechanisms to help people purchase nongroup coverage if they had or were projected to have high health care costs. However, by 2026, the amount of funding for states that expanded Medicaid would roughly equal those states’ federal Medicaid costs (as projected under current law) for people made eligible for the program under the ACA. Those states would not have enough federal funding to both provide similar benefits to people in an alternative program and extend support to others. In particular, CBO and JCT expect that most of those states would then choose to provide little support to people in the nongroup market because doing so effectively would be the more difficult task.

In states that have not expanded Medicaid, CBO and JCT anticipate, funds from the new block grants would be used partly to help people purchase nongroup coverage if they had or were projected to have high health care costs (through what are known as high-risk pools). Several other uses would help satisfy the requirement to use at least half of the funding to provide assistance to people with income between 50 percent and 300 percent of the FPL and potentially reduce pressure on state budgets. For example, states could fund some of their programs that would have operated under current law. They could make payments to health care providers, primarily for services provided to low-income people, such as paying for uncompensated care. And states could use funds to increase Medicaid payment rates or benefits for people who remain eligible under the legislation.

Effects on Health Insurance Coverage

CBO and JCT expect that, if this legislation was enacted, millions of additional people would be uninsured compared with CBO’s baseline projections each year over the 2018–2026 period. (Adopting a well-established definition, the agencies categorize people as uninsured if they are not covered by a policy or enrolled in a government program that provides financial protection from major medical risks.) That increase would stem mainly from lower enrollment in Medicaid and the nongroup market.

CBO estimates that many fewer people would be enrolled in Medicaid for three primary reasons: The expansion of the program established by the ACA would be repealed starting in 2020, federal reimbursement to states for Medicaid would be capped on a per-enrollee basis beginning in 2020, and the individual mandate penalty (which induces some people to enroll in Medicaid) would be repealed upon the legislation’s enactment.

Total enrollment in the nongroup market would be lower because the current-law subsidies for coverage in that market would be eliminated and the individual mandate would be repealed. CBO and JCT expect that the funds from the block grants would do little to offset the effects on the nongroup market of that elimination other than to facilitate modifying market regulations. For example, the distribution of those funds among states would differ substantially from the federal funding under current law, and many states that expanded eligibility for Medicaid and have particularly high levels of coverage under current law would receive the largest reductions in funding under the block grants. Consequently, those states would find it particularly challenging to reach current enrollment levels using the available subsidies. Because supporting the nongroup market would be more difficult, states would probably expend less in grant funding to do so. Also, states that received a large increase in funding in 2020 relative to the amount under current law would spend the money slowly, CBO and JCT expect, and not be able to immediately boost coverage because they would have insufficient information technology and related infrastructure to establish their own system for administering subsidies. In addition, states would probably use some of the available funding for purposes that would not be geared toward increasing health insurance coverage, such as for payments to providers who deliver health care services to low-income people, state programs that would have operated under current law, or both.

Other factors affecting the total number of people uninsured would partially offset the lower enrollment in Medicaid and the nongroup market. In particular, some people who would have been covered by Medicaid under current law would be covered by the alternative programs established by states that would have their expansion of Medicaid rolled back. (If expansion states’ alternative programs spent an equal amount per enrollee for a population similar to the projected population made newly eligible under the ACA, the total cost would roughly equal the total block grant amount for those states by 2026.) In addition, the agencies expect, more people would be covered by employment-based insurance because some employers that would not have done so otherwise would respond to the reduction in subsidies for nongroup coverage by offering coverage to their employees.

Because the legislation would not provide funding for the block grants after 2026, the increase in the number of uninsured people compared with the number under current law would be significantly greater after that year.

Effects on Medicaid

All told, federal spending on Medicaid would be reduced by about $1 trillion over the 2017–2026 period under this legislation, and the program would cover millions fewer enrollees. The largest effect would stem from eliminating funding for adults made eligible by the ACA. Depending on how states used their new grant funds, many of those people could receive assistance in other ways. Other changes to Medicaid, such as capping Medicaid spending on a per-enrollee basis and allowing work requirements, would also occur under the legislation.

The Legislation’s Caps on Federal Medicaid Spending

Beginning in 2020, the federal government would establish a limit on the amount of reimbursement provided to states for Medicaid on a per-enrollee basis. For each state, that limit would be set on the basis of the average cost of medical services for most enrollees who received full Medicaid benefits in a recent period, although no limit would apply for disabled children. (Adults made eligible for Medicaid by the ACA would be excluded from those calculations because their coverage would be eliminated.) Those per-enrollee costs would be allowed to increase by no more than the growth in certain price indexes:

• For nondisabled children and nondisabled adults enrolled in Medicaid, the medical care component of the consumer price index (CPI-M) during the 2020–2024 period and the consumer price index for all urban consumers (CPI-U) thereafter, and

• For most enrollees who are disabled adults or age 65 or older, the CPI-M plus 1 percentage point during the 2020–2024 period and the CPI-M thereafter.

If a state spent more than the limit on federal reimbursement, the federal government would provide no additional funding to match that spending.

In general, states would not have substantial additional flexibility under the per capita caps. A few states would probably obtain additional flexibility to make changes to their Medicaid program by participating in the Medicaid Flexibility Program, an option to receive a block grant of a fixed amount rather than a per-enrollee amount for nondisabled adult enrollees. Under that option, states could alter cost-sharing requirements and, to a limited degree, benefits. However, because funding under that program would grow over time at the rate of the CPI-U, CBO anticipates that it would be attractive mainly to the few states that expect to decline in population and would have little effect on enrollment in Medicaid. It would not be attractive in most states experiencing population growth, as the fixed block grant would not be adjusted for such growth. In those states, population growth would constrain federal reimbursement per enrollee.

Effects of Capping Medicaid Spending

By CBO’s estimates, in most states, capping federal Medicaid spending would result in less total reimbursement than would occur under current law. As a result, states might decide to commit more of their own resources to maintain current-law levels of spending for people who would remain eligible for the program — particularly for nondisabled children and nondisabled adults after 2024. Alternatively, they might decide to reduce spending in various ways: by cutting payments to health care providers and health plans, eliminating optional services, restricting eligibility for enrollment through work requirements and other changes, or (to the extent feasible) finding more efficient methods for delivering services. In some states, CBO anticipates, a portion of the new block grants would be used to boost spending on Medicaid. All in all, CBO expects that states would adopt a mix of those approaches. If those approaches reduced enrollment, federal spending would be reduced.

Some of the choices made by states could reduce enrollees’ access to care. If states reduced payment rates, fewer providers might accept Medicaid patients — especially because, in many cases, Medicaid’s rates are already significantly below those of Medicare or private insurance for some of the same services. If states reduced payments to Medicaid’s managed care plans, some plans might shrink their provider networks, curtail quality assurance, or drop out of the managed care program altogether. If states reduced covered services, some enrollees might decide either to pay out of pocket or to forgo those services entirely. And if states narrowed their categories of eligibility or used administrative procedures that made enrolling more difficult, some people would lose access to Medicaid coverage. (However, some might become eligible for similar services provided through a new block grant program, if enough money was available in their state.)

Alternatively, if states chose to leave their Medicaid program unchanged and instead found other ways to offset the loss of federal funds, enrollees would notice little or no change in their Medicaid coverage. States might also find ways to deliver services at a lower cost without affecting access to care for enrollees.

Work Requirements for Medicaid

Under the legislation, CBO anticipates, some states would use work requirements — allowed starting October 1, 2017 — to reduce enrollment and the associated costs. Under current law, states may not condition the receipt of Medicaid on any criteria related to a person’s employment status. This legislation would permit states to impose a work requirement for an adult as long as the person is not disabled, elderly, pregnant, or exempted for another allowed reason. The definition of work would be the same as the Temporary Assistance for Needy Families program’s, which includes activities such as unsubsidized employment, subsidized employment, vocational training, and educational activities. The legislation would provide states with broad discretion to define how many hours of work each week were required; how long enrolled people would have before needing to meet the requirements; and, if they failed to meet the requirements, when benefits would cease.

Effects on the Nongroup Market

CBO and JCT would need at least several weeks to provide point estimates of the effects on health insurance coverage in the nongroup market, but the direction of the effects is clear. Overall, CBO and JCT estimate, millions fewer people each year would be enrolled through the nongroup market under the legislation than would be under current law.

Because this legislation would, upon enactment, eliminate penalties associated with the individual and employer mandates, fewer people would enroll in health insurance obtained through the nongroup market (as well as through employment-based coverage and Medicaid) starting in 2018. The estimated savings from the reduced subsidies stemming from lower enrollment exceed the estimated loss of revenues from eliminating the mandate penalties in 2018 and 2019. Starting in 2020, the grants and the ability to modify market rules would change the operation of the nongroup market significantly in most states. Insurers’ anticipation of those changes would probably affect their decisions to participate in the nongroup market before 2020, also affecting enrollment.

To use the grants provided by the legislation to aid people purchasing insurance in the nongroup market, states would face the difficult task of implementing new systems by 2020 (regardless of any market rules changed), so their options would initially be limited. Implementing broader changes to the nongroup market would probably take states several years. In 2020 or in later years, some states might eliminate many regulations on health insurance and use grant funds to subsidize insurance for people facing high prices in an unregulated market. However, in many states, the transitions starting in 2020 would be difficult — and some areas would probably have no insurers offering policies in the nongroup market until the new market rules were clear and insurers had enough time to adapt to them.

Difficulties in Providing Income-Based Assistance

Providing income-based assistance to people to purchase insurance in the nongroup market would be especially difficult. To establish its own system of subsidies for coverage in the nongroup market related to people’s income, a state would have to enact legislation and create a new administrative infrastructure. A state would not be able to rely on any existing system for verifying eligibility or making payments. It would need to establish a new system for enrolling people in nongroup insurance, verify eligibility for tax credits or other subsidies, certify insurance as eligible for subsidies, and ultimately ensure that the payments were correct. Those steps would be challenging, particularly if the state chose to simultaneously change insurance market regulations. Insurers would also need time to develop plans under the new system. And accomplishing those steps before 2020, when the tax credits available under current law would be eliminated, would be hard.

Financial constraints would also limit states’ options. Under current law, on average in any month during 2016, 8.4 million people had income between 100 percent and 400 percent of the FPL and received income-based subsidies for nongroup coverage. That figure represents roughly half of the estimated total number of enrollees in the nongroup market in that year. Given the difficulties in establishing a new income-based subsidy for nongroup insurance, those people could potentially receive assistance from a new state program similar to Medicaid. But if such a program covered people made eligible for Medicaid under the ACA but no longer eligible under the legislation, then funding to cover additional people would probably be limited.

In CBO and JCT’s assessment, the states that expanded Medicaid eligibility under the ACA would be the most likely to try to establish a new program or provide new subsidies to aid people with income between 100 percent and 400 percent of the FPL who would enroll in the nongroup market under current law. However, CBO and JCT expect that many of those people would receive little or no support from the block grants to pay for health care expenses. States that expanded Medicaid would be facing large reductions in funding compared with the amounts under current law and thus would have trouble paying for a new program or subsidies for those people.

Effects of Retaining Market Rules

In states that did not modify the rules governing the nongroup market, its operation would differ depending on how the new block grants were used. If a state used its block grant funds entirely to create a program for the people made eligible for Medicaid under the ACA (but no longer eligible under the legislation) and people with somewhat higher income, for example, then many of those people would receive benefits through that program. However, people with income too high to be eligible for that program would probably face an unstable nongroup market.

Without subsidies — and with insurers required to accept enrollees having preexisting health conditions and with premiums varying only by age, geography, and smoking status — premiums would be high, and few people would enroll. Not only would enrollment decline, but the people most likely to remain enrolled would tend to be less healthy (as they would be more willing to pay higher premiums). Thus, average health care costs among the people retaining coverage would be higher, and insurers would have to raise premiums in the nongroup market to cover those higher costs. Anticipating such an unsustainable spiral, some insurers would not participate in the nongroup market. In those areas with insurers, according to CBO and JCT’s analysis, enrollment would continue to drop, and premiums would continue to increase year by year. Under such circumstances, over time, fewer and fewer insurers would continue to offer insurance in the nongroup market.

Without any changes to the rules governing the nongroup market, if a state used a combination of funding from the new block grants and its own resources to provide subsidies in that market similar to those under current law, then the nongroup market would probably be stable. However, insurance plans would attract less healthy people and premiums would be higher than under current law. If a state required individuals to have insurance, some healthier people would enroll, and premiums would be lower.

Effects of Modifying Market Rules

Starting in 2020, under the legislation, states could modify certain existing rules governing the nongroup market. For people who received direct assistance through the block grant, states could specify the extent to which insurers could vary premiums for enrollees and the benefits that were required, with certain exceptions. The assistance could consist of a small flat amount per enrollee provided to insurers or some other direct subsidy. (States would probably need to use other funding to satisfy the legislation’s requirement that at least half of the block grant amount provide assistance to people with income between 50 percent and 300 percent of the FPL.)

CBO and JCT anticipate that most states would eventually modify various rules to help stabilize the nongroup market. Most states would lack the stabilizing mechanisms that exist under current law: significant subsidies for nongroup health insurance and the requirement to purchase insurance. Their eventual modifications to market rules — covering, for example, how premiums could be set and what benefits policies would have to provide — would increase the number of insured people by 2026 above what would occur under this legislation if states did not modify the market rules, CBO and JCT expect. Nevertheless, with the modifications, coverage for people with preexisting conditions would be much more expensive in some of those states than under current law. (Without such modifications, as discussed in the previous section, coverage could become unavailable or more expensive for many more people than it would be under current law.)

States could expand the limits on how much insurers in the nongroup market could vary premiums on the basis of age, for example. Under current law, a 64-year-old can generally be charged premiums that cost up to three times as much as those charged to a 21-year-old. CBO and JCT anticipate that some states would increase that ratio — to, say, 5 to 1, instead of 3 to 1 — which would reduce premiums for younger people and increase premiums for older people and tend to somewhat increase insurance coverage, on net.

In addition, CBO and JCT expect that some states would alter requirements that policies include the essential health benefits specified by the ACA and instead allow a narrower scope of benefits. For some people, their premiums would be lower, but their insurance would cover fewer medical services. CBO and JCT expect insurance covering certain services not included in the scope of benefits to become more expensive — in some cases, extremely expensive.

The scope of benefits could be modified to, among other things, exclude coverage of services that have high costs and are used by few people. If so, then coverage could be difficult to obtain for mental health care, rehabilitative and habilitative treatment, and certain very expensive drugs. Such modifications would lower premiums for many people and increase the number of people with coverage for a narrower set of benefits. But on the basis of historical experience, CBO and JCT anticipate that the funding available to help provide coverage for excluded high-cost services would be insufficient in some cases even if a special program was designed for that purpose. Also, states would probably be conservative in setting eligibility rules for such a program to ensure that costs did not exceed the available federal funds. Therefore, the agencies expect that insurance coverage for high-cost services would become extremely expensive in those areas, as it was in some places before the enactment of the ACA in 2010.

CBO and JCT also anticipate that some states would allow insurers to set premiums on the basis of an individual’s health status. That is, the state would eliminate the requirement for what is termed community rating for premiums charged to such people, and they would be charged premiums based on their own expected health care costs (medically underwritten premiums). People with lower expected health care costs would have lower premiums, and more of those people might buy such insurance than would do so if premiums were not based on an individual’s health status. However, the higher the expected health care costs, the higher the premiums would be; for some people, premiums would be a very large share of their income. As a result, some people who would have been insured in the nongroup market under current law would be uninsured, and others would obtain coverage through a family member’s employer or through their own employer.

Budgetary Effects of Noncoverage Provisions

This legislation would make changes other than those directly affecting health insurance coverage, increasing the deficit by a total of $28 billion over the 2017–2026 period, with an on-budget increase of $22 billion. Many provisions would have the same effects as those estimated for prior versions of the legislation, as explained in CBO’s earlier estimates.⁶ Those provisions address, among other things:

• The Better Care Reconciliation Implementation Fund,

• Medicaid and CHIP quality performance bonus payments,

• The Prevention and Public Health Fund, and

• The Community Health Center Program.

Some noncoverage provisions differ substantially from those in prior versions of the legislation. They would, among other things:

• Decrease the period for which Medicaid benefits may be covered retroactively from up to three months before a recipient’s application to the second month before the month in which a recipient makes an application;

• Lower the threshold for the amount of taxes that states can collect from health care providers from 6 percent to 4 percent by 2025;

• Permit states to provide inpatient psychiatric services to adult enrollees for up to 90 days per calendar year;

• Provide 100 percent federal reimbursement for services for Medicaid enrollees who are members of an Indian tribe;

• Reduce the cuts to allotments to states for payments to hospitals that treat a disproportionate share of uninsured and Medicaid patients — scheduled to occur from 2018 to 2025 — by the amount of any annual reductions (with the effects of inflation removed) in a state’s funding from the new block grants; and

• Increase the federal matching rates for the two states, Alaska and Hawaii, that have a separate poverty threshold by a portion of the average matching rate for all other states that have one poverty threshold.

The legislation would also repeal several revenue-related provisions of the ACA that are not directly related to health insurance coverage. Effective beginning in 2018, the provisions with the largest budgetary effects would increase the maximum contribution to health savings accounts, repeal the medical device excise tax, and allow money from health savings accounts and flexible spending arrangements to be used to purchase over-the-counter drugs.

This document was requested by the Chairman of the Senate Committee on the Budget. In keeping with CBO’s mandate to provide objective, impartial analysis, the document makes no recommendations. Kate Fritzsche, Jeffrey Kling, Sarah Masi, Kevin McNellis, Allison Percy, Lisa Ramirez-Branum, Robert Stewart, and the staff of the Joint Committee on Taxation prepared the document, with guidance from Jessica Banthin, Chad Chirico, and Alexandra Minicozzi. Theresa Gullo, Mark Hadley, Robert Sunshine, and David Weaver reviewed the document; John Skeen edited it; and Casey Labrack prepared it for publication. An electronic version is available on CBO’s website (www.cbo.gov/publication/53126).

Keith Hall
Director

September 2017