Expect the Unexpected: The Impact of Medicaid Caps
One of the most sweeping changes in the House-passed American Health Care Act and in the Senate’s draft bill, The Better Care Reconciliation Act of 2017 is the end to the federal government’s commitment to share all Medicaid costs with states. The House bill replaces that commitment with a cap that would limit—and significantly reduce—federal Medicaid spending. The cap would apply to virtually all spending under the program, including for the care provided to children, pregnant women, people with disabilities, and seniors. It would apply to acute care, as well as long-term care, like nursing home and home- and community-based services and supports. The Senate draft adopts the House approach, exempting certain children from the cap but including other features that make the cap even “tighter,” meaning the cuts that result would be even greater, particularly toward the end of the next decade.
According to the Congressional Budget Office (CBO), the projected Medicaid spending reductions resulting from the House version of the AHCA and Better Care caps are deep, but as important is the uncertainty and risk associated with the caps. Health care costs are inherently difficult to project, but one of the few things that is certain is that unexpected health care needs and costs will inevitably arise—whether due to new cancer therapies, sharp increases in drug costs, or an epidemic like the current opioid crisis. By definition, a cap on federal funding shifts the risk of all costs above the caps to states. States will either reduce their spending to come under the caps or they will be responsible for 100 percent of the cost over the caps. We have analyzed states’ risk by modeling the impact, by state, of the per capita cap in the House bill under various scenarios.
Let’s start with a standard scenario, using the Centers for Medicare and Medicaid Services’ Office of the Actuary’s estimate of Medicaid spending per enrollee growth under current law and CBO’s projection of the trend rates that the House AHCA bill would use to adjust the caps each year. Once the caps are in place, the federal government would adjust spending each year according to a “trend rate” of growth. The Senate bill’s trend rate would be equal to the rate of inflation in medical prices, or the medical consumer price index (CPI), plus 1 percentage point for the disabled and elderly groups and medical CPI for all others. CBO projects medical CPI would be 3.7 percent during this period.
Under this scenario, the cap is expected to result in $295 billion in cuts between 2020 and 2026. To stay within the caps, states would need to reduce total (federal and state) spending in the program by this amount. The nearly $300 billion reduction in funding for the program would be in addition to even larger reductions in Medicaid funds for states that had expanded Medicaid under the Affordable Care Act; the AHCA would end the enhanced matching rate for the expansion in 2020 except for a quickly dwindling group of “grandfathered” enrollees. The Senate bill would also eliminate the enhanced match through a three-year phase out. If states maintained their own spending at current law levels even in the absence of federal matching payments, the federal reduction from the per capita cap would be $172 billion.
The impact of the cap on states varies considerably, driven by the size of a state’s program, its historical spending base, and the distribution of enrollees by eligibility group. The caps are derived by multiplying the number of people enrolled in each eligibility group by the cap established for each group. (See here for a full explanation of the cap calculation.) The disparities across states will likely be greater than these estimates suggest because we are using national projections for growth. In practice, state growth rates will vary. Indeed, one of the major risks to states under the AHCA’s per capita cap is that no one knows which states are going to be hardest hit by a public health crisis, price increases, demographic changes, or other factors that will drive health care spending pressures.
A June 2017 analysis of inpatient stays and emergency department visits related to opioid use illustrates this point: nationally, opioid-related admissions grew by 64 percent and emergency room visits nearly doubled between 2005 and 2014, but some states saw rates of increases far in excess of the national average. Maryland had the highest rate of opioid-related inpatient admissions in 2014, with 5.5 times as many admissions (per 100,000 people) as Iowa, the state with the lowest rate. Today, Medicaid financing ensures the federal government shares these costs wherever they arise; under a cap, the risk falls solely on states.
But that’s just one possible scenario. What if the trend rate applied to adjust the cap each year is just a little below the CBO projection? The medical CPI is an unpredictable metric. It has ranged from as low as 2.4 percent to as high as 4.7 percent in recent years. CBO projects medical CPI will average 3.7 percent for the period between 2020 and 2026, but if it turns out to be just half a percentage point lower (and actual Medicaid costs do not similarly drop) the total cut (federal and state) due to the per capita cap will jump to $466 billion for that period of time. Arizona’s cuts grow from $8.8 billion to $12.8 billion; Ohio’s from $12.3 billion to $19.1 billion; and Texas’s from $16.5 billion to $27.8 billion. And again, for the expansion states, these cuts are in addition to the losses states will experience when the enhanced matching rate for the expansion ends in 2020.
Even more striking is the impact if Congress decides to change the trend rate it uses to set the caps. The Senate is considering adopting the overall CPI, the index for general prices, beginning in 2025 and continuing thereafter. Growth in the overall CPI is projected by CBO to be 2.4 percent between 2020 and 2026; if it is substituted for medical CPI for the entire 2020–2026 period, total cuts resulting from the per capita cap more than double to $729 billion.
If the medical CPI or the overall CPI accurately tracked state Medicaid costs, the volatility in the trend rates would be less of a problem, but neither measure is designed to track the appropriate or expected level of Medicaid spending nationally, let alone by state. Both measure consumer spending trends, and even the medical CPI gives relatively little weight to the cost of key Medicaid services, such as nursing home care, and neither address local variations in costs.
The problem, however, is not with the measure but with the concept that a cap based on spending in a state in a prior year trended forward using a national trend rate will provide any assurance that federal funding won’t fall far short of actual need. No trend rate can accommodate differences in spending pressures across states whether due to changes in utilization attributable to a state’s demographics such as its aging population, a rise in demand for care due to a public health crisis, or the need over time to adjust provider payments to ensure access to care for children or quality of care in nursing homes.
And that brings us back to the concept of shifting risk. The cap on federal Medicaid funding is intentionally designed to reduce federal costs and make those costs predictable to the federal government. In doing so, the harm as well as the uncertainties and risks shift to states, health care providers, and the people served by the Medicaid program.