Controlling Medicaid's Growth: Targeted Strategies Versus Across-the-Board Limits on Spending
Historically, federal and state policymakers have relied on targeted strategies to control Medicaid spending. These strategies zero in on specific drivers of cost, especially in areas where costs are escalating, and aim to reengineer services, making them more efficient and cost-effective. The approach reflects concerns that across-the-board spending limits would result in the denial of care to people in need. Notable examples of targeted cost-containment include reforms to lower outpatient prescription drug costs, expand access to preventive care, scale up managed care models, and create alternatives to long-term, costly institutional care. The approach also has included setting upper limits on certain expenditures, such as supplemental payments to hospitals that serve a disproportionate share of low-income people.
In addition, policymakers have imposed more stringent limits on the circumstances under which states can use health care provider taxes to finance their required share of Medicaid spending. These limits restrict the amount of money states have to invest in their Medicaid programs, which in turn restricts the amount of federal funding for which states can qualify.
Over the decades, these strategies have led to significant reductions in the cost of providing health care to individuals and eliminated unnecessary spending. For example, today about 80 percent of all Medicaid beneficiaries are served through some form of managed care.8 And cost-effective in-home and community-based care is now more common than long-term institutional care.9 The overall impact has been to make a growing national program more efficient, while still delivering quality health care.
Targeted cost-containment, however, does not address the primary source of increased spending on Medicaid: growing enrollment. Nor does it limit states’ ability to deploy new technologies to improve coverage or the quality of care (like offering new vaccines or drug treatments), or introduce new efficiencies like electronic health records or updated management information systems. As a result, Republican leaders are calling for a very different approach to cost control.
In particular, President-Elect Donald Trump and House Speaker Paul Ryan have proposed to repeal the Affordable Care Act (ACA) and restructure Medicaid. The president-elect wants to replace Medicaid with block grants to states.10 Ryan’s ideas are outlined in A Better Way: Our Vision for a Confident America, which devotes six of its 37 pages to Medicaid reform. While recognizing that Medicaid is a “critical lifeline for some of our nation’s most vulnerable patients,”11A Better Way nonetheless proposes to substantially scale back the federal contribution. The first step would be to roll back eligibility. States that had not already expanded their Medicaid programs by 2016 to cover nonelderly poor adults (19 states as of November 2016) would have no access to federal funds to support such expansion.12,13 States would then have a choice of complying with “default” limits on per capita spending set by the federal government or receiving support in the form of a block grant.
Other proposed changes include restricting the extent to which federal funds can be used to cover certain populations or services while eliminating federal funding for others. One example would be to withdraw federal funding for people who have served time in prison or in jail.
Block Grants as an Alternative to Flexible Spending
The federal government helps fund an array of public services—from housing to public health, education, and law enforcement—through grant programs that give states annual fixed amounts to spend on activities permitted under the terms of the program. Because the federal funds available to states are fixed amounts, they grow at a predictable, formula-driven rate from one year to the next—or not at all, if Congress does not appropriate funding increases. Such programs help support state health and social welfare activities; they do not entitle individuals to services, as does health insurance. Furthermore, they do not automatically take into account population growth, as would a per capita cap.
Providing federal funding for Medicaid using this type of approach (often referred to as a block grant) would disconnect the level of funding from the number of Medicaid beneficiaries and the cost of providing care. In other words, the federal contribution would remain the same, or grow only according to a preset formula, no matter how large the population in need becomes or how much a state actually must spend on health care for Medicaid recipients. To permit states to manage their Medicaid programs with a fixed amount of federal funding, the entitlement to coverage would need to be eliminated, and federal rules regarding eligibility, coverage, and payment would need to be substantially restructured or repealed. The Children’s Health Insurance Program (CHIP) provides an example: The federal contribution is fixed and states are free to scale back enrollment and coverage as needed to avoid budget shortfalls. (A special maintenance-of-effort provision in the ACA prevents participating states from changing CHIP eligibility before October 2019, but states can roll back benefits or increase cost-sharing.)
Proposals to fund Medicaid through block grants have a long history. In 1981, President Ronald Reagan proposed state-specific block grants based on historical levels of spending in each state. Congress rejected the proposal but did temporarily tighten the federal funding formula. With the country in the midst of a recession, even this relatively modest downward adjustment in federal funding triggered widespread reductions in enrollment as well as benefits at a time when the opposite was needed.14 This temporary spending reduction was repealed in 1984 through bipartisan budget legislation.
A little more than a decade later, in 1995, both the House and Senate passed a bill that would have funded Medicaid through block grants to states based on historic average levels of spending nationally, coupled with a complex growth formula that would set future spending levels well below the expected rate of growth in Medicaid. President Clinton vetoed the legislation in the face of widespread evidence regarding its adverse financial impact on state Medicaid programs and underlying state economies.15
Since that time, block grant proposals have appeared intermittently. Most recently, in 2015, Senators Richard Burr of North Carolina and Orrin Hatch of Utah, both Republicans, and Congressman Fred Upton, a Republican from Michigan, introduced bills to repeal the Affordable Care Act.16 Both bills would have ended the ACA’s Medicaid expansion funding for low-income adults and created block grants to states based on levels of spending prior to 2014.
The Congressional Budget Office (CBO) estimated that the proposed legislation would reduce federal spending by $1 trillion over 10 years.17 Much of the savings would come from denying access to Medicaid for roughly 14 million people—the estimated number of low-income Americans who would have been eligible for Medicaid by 2026. Additional savings would be achieved by reducing federal spending for the traditional Medicaid program by 4.3 percentage points. By 2026, according to the CBO, federal spending on Medicaid was expected to be one-third below projected spending levels. Although the House bill offered no details regarding the level of flexibility states would have in order to absorb the significant reductions in federal funding, it had enough support to be incorporated into the 2017 fiscal year budget that was released in 2016.
A Better Way offers no formula for how block grants would be calculated or trended forward, or what growth factors would be considered, other than to note that the (undefined) base year for purposes of calibrating the block grant would exclude the ACA expansion population and would transition beneficiaries in expansion states to “other sources of coverage.” As a block grant, the formula presumably would be divorced from actual rises in enrollment and the cost of coverage, relying instead on a formula designed to produce predictable savings over time. Assuming that a new block grant proposal might mirror the 2017 House budget proposal, federal Medicaid funding could be expected to fall by a third in the tenth year of the proposal’s implementation.18
Per Capita Limits on Spending
Another way to control spending on Medicaid is to establish limits on per capita spending—per capita caps. These caps have the advantage of allowing funding to increase along with enrollment and underlying need, while setting an annual upper limit on federal spending per enrollee, and are supported by many advocates of Medicaid finance reform.19 President Clinton suggested this kind of cap prior to vetoing the 1995 congressional block grant proposal discussed above, but Congress rejected the idea.
Within this approach there are options: the federal government could set a single per-enrollee cap that applies to all Medicaid recipients, including children, adults, the elderly, and persons with disabilities; it could set different caps for each group; or it could exempt certain groups from the cap. However, since spending on elderly people and people with disabilities accounts for nearly two-thirds of total Medicaid spending,20 the per-enrollee limits would need to apply to these populations in order to generate significant savings. In addition, the cap or caps could be structured to apply to all Medicaid services or only certain services, with others such as prescription drugs being exempt. And how much growth over time to allow in the caps themselves is also an open question.
Limits on per capita spending are more accommodating, at least in theory, to increases in enrollment reflecting underlying need, but a fundamental trade-off remains: To save money at the federal level, the caps must keep spending below projected levels—in effect shifting the burden to states in much the same way that block grants do. Under caps as well as block grants, states will face a gap between the costs of providing coverage and the federal funds available to offset those costs. And as with block grants, federal rules pertaining to eligibility, coverage, and payment to providers would have to be altered, allowing states to narrow their programs and avoid significant budget deficits.
The effects of per capita caps could have significant consequences for people’s health care and for insurers. For example, states might reduce already-low provider payment rates, forcing out many current providers and thus limiting access to care, a shift that research suggests would be especially detrimental for people who need specialized treatment and long-term care.21 If federal spending updates lag rising health care costs, states might reduce managed care payments below actuarially sound levels, triggering the demise of managed care plans. Or states might narrow eligibility to control costs, perhaps even eliminating coverage for the most needy and costly individuals.
Under Ryan’s plan outlined in A Better Way, states that choose to operate their Medicaid programs within the federal caps (as opposed to receiving a block grant) would transition to a new funding formula. That formula would take effect in 2019 but would be calculated based on enrollment and costs in 2016—three years earlier. The plan would apply separate caps to each of the four major beneficiary categories (children, adults, elderly people, and people with disabilities), which would be permitted to grow, but at an unstated rate below “current law.”22 Each state’s allotment would apply the federal cap formula to the sum of its 2019 enrollment, adjusted for full-year equivalency (what the cost would be if every beneficiary remained enrolled in Medicaid for the full year) across all eligibility categories.
This plan does allow for population growth. But it fails to take into account that even within a single beneficiary category, some individuals are much more expensive to cover than others. In particular, the formula would treat people who are enrolled in Medicaid for part of the year as less expensive than full-time enrollees when, in fact, providing coverage to them can be more expensive if they enrolled because of a single, high-cost health episode. Nor does the plan explain how the high number of part-year enrollments would be taken into account in reaching an accurate picture of growth over time. Because the plan proposes to generate a predictive enrollment figure, rather than use actual enrollment, it could undercount enrollment. It also could fail to adequately adjust for short enrollment periods, which carry extremely high costs.
While A Better Way notes that the caps would reflect each state’s expenditures for medical assistance and “non-benefit” expenditures, exactly which expenditures would be counted in the calculation is unclear. This is because the proposal notes that “[r]ecognizing the complexity of Medicaid financing, certain payment categories would be excluded . . . and would be calculated through a separate funding stream, such as payments to states for disproportionate share hospitals, graduate medical education payments, and other appropriate exclusions.”
The proposal also would replace the actuarial soundness principles used to set managed care rates under current law with a new (undefined) “reasonable enforceable” premium test for nondisabled adults, as well as replace Medicaid’s specific benefit and payment rules with state flexibility to adopt coverage designs that “promot[e] personal responsibility and healthy behaviors and encourag[e] a more holistic approach to care.” The proposal does not explain which aspects of Medicaid’s current coverage design would be eliminated or what an alternative design might look like.
What Counts as State Spending: An Unaddressed Issue
An important aspect of any proposal to reengineer federal funding for Medicaid is what will count as state spending for purposes of qualifying for federal funds. In fiscal year 2012, 69 percent of state Medicaid spending came from general revenues. States met their remaining obligations through local government contributions (16%), permissible health care–related taxes (10%), and other sources such as special dedicated revenues (5%).23 If block grants or caps designated any of these forms of financing as impermissible, states would be in a position in which they would not quality for every federal dollar otherwise available to them, causing federal outlays to fall even more than predicted. While easily overlooked, this crucial issue should be addressed in any proposal to create block grants or limit per capita spending—it remains unanswered in A Better Way.
As the country’s largest insurer, Medicaid is subject to the same cost drivers that affect all providers of health insurance: population growth and demographic trends that increase enrollment, health trends that influence how often people need care and what kind of care they require, and advances in technology that drive up costs, among other factors. But unlike commercial insurers, government-funded Medicaid, in its role as first responder and safety net, is more vulnerable to these trends and to cost increases. For more than 50 years, Medicaid has been rooted in a flexible federal–state partnership, constantly restructured over time to meet current challenges.
Any attempt to restructure federal financing for Medicaid and replace flexibility with strict spending limits—whether in the form of block grants, per capita limits on spending, restrictions on what counts as state expenditures, or a combination of all three—would divorce funding considerations from the real-life needs that have informed federal and state Medicaid policy for half a century. Crucially, a per capita cap would permit population growth to occur. But the limit of lawmakers’ appetite for continued growth in enrollment is unclear. Given how states responded to the relatively mild and temporary funding reductions the federal government enacted in 1981, sweeping changes like those currently under consideration are likely to produce far more substantial fallout.