Proposal to Establish Federal Medicaid “Blended Rate” Would Shift Significant Costs to States
Would Be Hard to Set Fairly and Accurately; Would Likely Force Cuts to Children, People With Disabilities, Seniors, and Health Care Providers
June 24, 2011
An Obama Administration proposal that's on the table for budget negotiators would reduce federal Medicaid expenditures by reducing the federal share of Medicaid and CHIP costs, shifting costs to states and likely prompting states to cut payments to health care providers and to scale back the health services that Medicaid covers for low-income children, parents, people with disabilities, and/or senior citizens (including those in nursing homes). Reductions in provider payments would likely exacerbate the problem that Medicaid beneficiaries already face regarding access to physician care, particularly from specialists.
The proposal would replace the various matching rates at which the federal government reimburses states for their costs in insuring people through Medicaid and CHIP with a single "blended rate" for each state. A state's blended rate would be set at a level that provided the state with less federal funding than under current law, thereby saving the federal government money.
The blended-rate concept has two significant weaknesses.
- First, it would essentially shift costs to states, rather than constrain them. The proposal produces little administrative-cost or other efficiency savings, as explained below. States, which face their own budget problems, likely would compensate for the reduction in federal funding by scaling back the services that Medicaid and CHIP (the Children's Health Insurance Program) cover, cutting payment rates to health care providers, or both. Some Medicaid beneficiaries already have limited access to physician care, particularly from specialists, due largely to Medicaid's already-low reimbursement rates. The shift in costs to states under the blended-rate proposal would make that problem worse.
- Second, the federal government would find it extremely difficult to calculate each state's blended rate fairly and accurately. Under last year's health reform law, the federal government will pay a substantially higher matching rate (i.e., cover a much larger share of the costs) for covering the large numbers of low-income people whom that law makes newly eligible for Medicaid starting in 2014. To compute a blended rate for each state, federal officials would have to make a number of assumptions about each state's future Medicaid and CHIP enrollment and expenditures, including how many people in each state who become newly eligible for Medicaid under health reform will actually enroll in the program and how healthy or sick they will be (since that will affect the health services that they use and, hence, their average health care costs). Federal officials also would have to estimate how many people in each state who are now eligible for Medicaid but not enrolled will enroll after health reform's coverage expansion and insurance mandate take effect, as well as what their health status will be. These assumptions would be rife with uncertainty, because they would not be based on actual state experience. As explained below, federal officials would be making other estimates as well, for which hard data are not available, that would affect the blended rate a state is assigned.
Thus, the blended rates almost inevitably would produce intense controversy, with many states likely challenging the federal estimates and contending that their blended rate has been set too low.
The budget framework that the Administration issued in April envisions at least $100 billion in federal Medicaid savings over ten years. The savings would apparently come from three places: 1) a series of small but significant measures outlined in the President's fiscal 2012 budget to increase Medicaid efficiency and reduce Medicaid's costs in providing medical equipment, prescription drugs, and certain other items, which would save somewhere around $10 billion to $15 billion; 2) a proposal to sharply restrict or bar states from raising part of their matching contributions for Medicaid by taxing health care providers, which would save about $25 billion to $45 billion depending on how sweeping the proposal is; and 3) the blended-rate proposal.
By limiting how states can raise funds to help pay their Medicaid costs, the "provider-tax" proposals -- like the blended-rate proposal -- represents a cost-shift to states. The Congressional Budget Office and the Office of Management and Budget estimate these proposals would reduce federal Medicaid costs because many states would not find other revenues to replace the provider tax revenues -- and would have to cut back their Medicaid programs as a consequence, which in turn would lower federal costs.
Depending on the specifics of the provider-tax proposal that is adopted, the blended-rate proposal could well be the principal Medicaid cut in a deficit-reduction package and account for the majority of the federal Medicaid savings in the package.
How the Blended Rate Would Work
Under current law, the federal government generally pays a fixed percentage of each state's Medicaid costs, known as the Federal Medical Assistance Percentage, or FMAP. This percentage ranges across the states from 50 percent to 75 percent and averages 57 percent nationally. The federal government pays a larger share of the costs -- 70 percent, on average -- of covering low-income children through CHIP. 
Under the Affordable Care Act (i.e., the health reform law), the federal government will pick up 100 percent of the costs that state Medicaid programs will incur in providing health care to individuals whom the ACA makes newly eligible for Medicaid, for the first three years that the ACA's Medicaid expansion is in effect (2014-2016). The federal government will cover between 93 percent and 95 percent of these costs in 2017 through 2019 (the percentage declines over these years), and 90 percent of such costs after that. In addition, states that expanded coverage to childless adults (a group not traditionally covered under Medicaid) prior to enactment of the health reform law will qualify for a higher matching rate for some of these previously eligible childless adults. States will continue to receive their regular federal Medicaid matching rate for other current Medicaid beneficiaries, and the higher matching rate they now receive for children insured through CHIP.
The blended-rate proposal would replace this mix of matching rates with a single matching rate for each state, which would apparently apply to all of a state's Medicaid and CHIP expenditures, outside of administrative costs.
The blended rate would be set significantly below the combined effect of the various federal matching rates a state would otherwise receive. This would save money for the federal government -- the federal government would pay a lower percentage of overall Medicaid and CHIP costs than under current law, and states would bear a greater share. To compensate for the federal funding reductions, states would either have to contribute more of their own funds or, as is more likely, shift costs to beneficiaries and health care providers by scaling back benefits and already-low payment rates.
A related concern is that the blended rates could be "dialed down" as federal policymakers assemble a deficit-reduction package, in order to maintain the overall level of federal savings as other spending or tax proposals that negotiators have tentatively agreed upon come to light and ignite opposition from powerful interest groups. At the behest of interest groups, policymakers could face strong temptation to drop or scale back savings from, for example, reforming farm subsidies -- and to dial down the Medicaid/CHIP blended rates instead. Such actions would increase the cost shift to states.
Blended Rate Overwhelmingly a Cost Shift
The only way that the blended-rate proposal would not primarily shift costs to states would be if it produced large administrative savings for states that offset most of their loss of federal funds. But the proposal does not do so. There would be some administrative savings, but they would be slight.
Some policymakers or their staffs may have initially thought that the proposal would yield significant administrative cost savings on the assumption that under current law, states will have to evaluate each Medicaid applicant's eligibility twice -- under current eligibility rules and under the rules that will take effect in 2014 -- in order to determine the appropriate matching rate for that individual (since the rate would differ depending on whether the individual would have been eligible under current rules). Recent developments show, however, that such an assumption is not correct.
Limiting or Eliminating Provider Taxes Would Also Shift Costs to States
Currently, 46 states and the District of Columbia use provider taxes to help fund Medicaid. (Alaska, Delaware, Hawaii, and Wyoming are the only states that do not have provider taxes.a) Most states levy taxes on hospitals and nursing homes; a lesser number have placed a tax on managed care organizations and other types of health care providers.b Federal policy reforms enacted in 1991 and strengthened in subsequent regulations require that such taxes actually generate state revenues and not be used as a means to essentially borrow funds from providers for use as state matching funds. (Those rules prohibit states from assuring providers that the taxes they pay will be returned to them in the form of Medicaid reimbursements.)
The co-chairs of the President's fiscal commission proposed to save an estimated $44 billion over ten years by phasing down and eventually eliminating states' ability to use revenue from provider taxes as part of their Medicaid matching contribution.c The President's fiscal year 2012 budget would limit but not eliminate the use of provider taxes, saving $26.3 billion over ten years, according to CBO.
Limiting or eliminating the use of provider taxes to help fund Medicaid produces federal savings because unless states find other funds to replace the lost revenue, they would have to cut back their Medicaid programs, which would lower federal costs.
a National Conference of State Legislatures, "Health Care Provider and Industry Taxes/Fees," February 2011, www.ncsl.org/default.aspx?tabid=14359.
b Kaiser Commission on Medicaid and the Uninsured, "Medicaid Financing Issues: Provider Taxes," May 2011.
c The National Commission on Fiscal Responsibility and Reform, "The Moment of Truth," December 2010, http://www.fiscalcommission.gov/sites/fiscalcommission.gov/files/documents/TheMomentofTruth12_1_2010.pdf .
The U.S. Department of Health and Human Services announced several weeks ago that states will not have to determine applicants' eligibility twice. Instead, in forthcoming regulations that the Office of Management and Budget is now reviewing, HHS will enable states to use other, much less burdensome methods to ensure proper claiming, such as statistical sampling and possibly other reliable methods to determine the percentage of a state's Medicaid beneficiaries that falls into each matching-rate category.
The costs to states of sampling or other such approaches should be modest. As a result, the administrative savings from a blended rate would represent only a small fraction of the amount of federal Medicaid funds that states would lose if the federal government shifted to blended rates and set those rates at levels to produce sizeable federal savings. The blended-rate proposal thus would substantially reduce federal Medicaid payments to stateswithout appreciably lowering states' administrative costs, which is why it is overwhelmingly a cost shift. 
Blended Rates Would Reduce Incentives for States to Cover More of the Uninsured
From beneficiaries' perspective, the blended-rate proposal raises two primary concerns. First, by shifting costs to states, it could lead them ultimately to take actions that harm low-income beneficiaries. Second, the blended rate would substantially reduce states' incentives to enroll large numbers of the low-income people whom the Affordable Care Act makes newly eligible for Medicaid starting in 2014.
The ACA gives states strong incentives to reach and enroll individuals made newly eligible for Medicaid because the federal government will pick up all of states' costs in providing health coverage to these people in the initial years and 90 percent of the costs over the long run. By contrast, under the blended-rate proposal, states would have to pay a much greater share of the cost of each newly eligible individual who enrolls (and a smaller share of the cost of those already eligible under current state Medicaid rules). States would likely be less energetic in their outreach efforts as a consequence, with the result that fewer newly eligible individuals likely would apply and enroll, and the number of people who remain uninsured would be higher.
For example, CBO assumes that only about 50 percent of newly eligible childless adults will enroll in Medicaid in the expansion's early years, even with a 100 percent federal matching rate. Enrollment likely would be considerably lower under a blended rate.
Calculating States' Blended Rates Would Be Difficult to Do Accurately
The second major concern regarding the blended-rate idea stems from the difficulty of calculating state blended rates accurately. The federal government would have to set each state's rate based on a number of assumptions about the state's future Medicaid and CHIP enrollment and expenditures. In the case of the health reform law's Medicaid expansion, these assumptions would not be based on actual experience and could well turn out to be off the mark. As a result, when the federal government set the blended rates, many states likely would strongly dispute them, arguing that the methodology was flawed because the assumptions were faulty.
While the Administration has not disclosed how it would determine the blended rates, the process would likely consist of something along the lines of the following two-part calculation:
- Federal officials would determine a "current-law" blended matching rate for each state, taking into account the state's regular FMAP, the higher matching rates for newly eligible individuals under the Affordable Care Act, the higher matching rate for some previously eligible childless adults (in states where that is applicable), and the state's CHIP matching rate. The state's current-law blended matching rate would be the rate estimated to provide the same amount of federal funding as the existing mix of matching rates.
- This current-law blended rate would then be reduced by a specified percentage, which would be applied to all states to ensure that states receive less federal funding than under current law. The size of the percentage reduction would be set to produce the overall amount of reduction in federal Medicaid expenditures that federal policymakers have agreed on.
This second part of these calculations would be straightforward. But the first part -- determining each state's current-law blended rate -- would not be. Federal officials would have to project the exact level of Medicaid expenditures in each state under the health reform law's Medicaid expansion, as well as future expenditure levels for each state's existing Medicaid program, its CHIP program, and the coverage of certain currently eligible childless adults in states that have already expanded coverage to this group. This would require making numerous assumptions, in some cases without any actual experience.
- Federal officials would have to estimate how many newly eligible people are likely to enroll in each state under the Affordable Care Act's Medicaid expansion and the expected costs in that state per newly eligible beneficiary. This would require federal officials to make assumptions about the newly eligible population in the state, the likely participation rate among newly eligible beneficiaries, and the makeup of the expansion population -- whether they will be healthier or sicker than current comparable beneficiaries and thus have lower or higher costs per beneficiary. Federal officials also would have to make assumptions about the Medicaid benefit packages that states will provide to newly eligible beneficiaries (the health reform law allows states to offer regular Medicaid benefits or a less comprehensive package) and about pent-up demand for health care among newly eligible beneficiaries. Uninsured individuals tend to be healthier than average and have lower costs, but they also tend to delay obtaining needed care; as a result, when they first gain coverage, they may use more health care services than those in comparable health who already have health insurance.
- In the handful of states that have already provided coverage to childless adults up to 100 percent of the poverty line, federal officials would have to estimate enrollment and expenditures among the currently eligible childless adults who will qualify for a higher matching rate.  Federal officials would have to make assumptions about whether enrollment among currently eligible childless adults would increase in these states, and to what degree, once the expansion takes effect because publicity and outreach efforts surrounding the Medicaid expansion could lead more currently eligible childless adults to sign up for Medicaid. 
- Federal officials would have to estimate expected enrollment and expenditures for the rest of the state's Medicaid program (the part reimbursed under the state's regular FMAP) -- and to make assumptions about the extent to which the enrollment of currently eligible individuals will increase as a result of the Medicaid expansion, the individual mandate, and other features of the Affordable Care Act. Federal officials would need to make assumptions about the relative health of such new enrollees, as well as about future changes in the mix of a state's beneficiary population -- including the impact on the beneficiary mix of the aging of a state's population, as well as the increased incidence of disability that occurs among a population as it grows older.
- Federal officials would have to estimate expected enrollment and expenditures in the state's CHIP program, including assumptions about whether the state's CHIP enrollment may differ from current enrollment levels after the Affordable Care Act takes effect and whether per-beneficiary costs will change over time.
- Federal officials would then have to use all of these assumptions and estimates to construct a current-law blended rate for each state. Basically, they would look at their estimates of expected federal Medicaid and CHIP expenditures in a state in future years, taking into account the different matching rates for various areas of Medicaid and CHIP spending under current law, and estimate the federal share of total federal and state expenditures for Medicaid and CHIP in the state under current law. This would be the blended rate that the state is assumed to receive under the current mix of Medicaid and CHIP matching rates.
Many of these assumptions would be subject to considerable uncertainty, particularly those related to the Medicaid expansion. As a result, a number of states likely would argue that the federal government set their current-law blended rate too low -- for example, by underestimating enrollment in the state for people newly eligible under the ACA's Medicaid expansion or by underestimating the per-beneficiary cost of serving the newly eligible population. A state whose current-law blended rate was too low would experience a larger cut in federal funding than federal policymakers had intended when they adopted the blended-rate concept (as well as a larger cut than states whose blended rates were closer to being accurate).
Further complicating matters, it is unclear when and how the federal government would initially set the blended rates. What base year would it use? What data would it use to derive the necessary assumptions, and how would it adjust those data to reflect the relevant future time periods? Would there be a mechanism to "correct," after the Medicaid expansion takes effect, the assumptions used to set a state's blended rate, and if so, how would that be done?
Consider what would happen if a state enrolled many more newly eligible individuals than federal officials had assumed. Under current law, the state will receive a matching rate of 100 percent for those individuals for the first three years the expansion is in effect. Under the blended-rate proposal, however, the state could be locked-in to the original participation estimates. States in this situation would argue that the federal government should adjust their current-law blended rates upward in later years to take their higher enrollment into account, when the data demonstrating the higher caseloads become available. The absence of such an adjustment mechanism would likely discourage states from enrolling more of the newly eligible individuals, with the result that more low-income people would remain uninsured.
One way to ease these problems would be to defer instituting the blended rate until enough time has passed after the ACA's Medicaid expansion has taken effect for the increased enrollment from the expansion to reach a steady state, so that federal officials could use actual state experience with the expansion in setting the blended rates. That would substantially reduce the federal savings in the ten-year budget window -- unless the federal government set the blended rates quite far below what states would receive under current law in order to generate as many savings in two or three years as it would generate (over the ten-year budget window) with more immediate implementation. Such an approach would shift even greater costs to states in the decades ahead and cause more harm to beneficiaries over the long run.
The blended rate proposal appears flawed. It would be used as a mechanism to institute significant cuts in federal Medicaid funding, relative to current law. It would thereby shift significant costs to states, and inevitably to beneficiaries and providers. Administrative savings gleaned by states would be modest. In addition, the calculation of the blended rate for each state would be extremely complicated and difficult to compute with accuracy and precision, because it would require numerous assumptions and estimates for which solid data are not yet available.
 Each state's CHIP matching rate is equal to its Medicaid matching rate plus 30 percent of the difference between the Medicaid matching rate and 100 percent. In other words, the CHIP matching rate is set at a level that effectively reduces the state's share of costs by 30 percent, relative to what the state is required to pay under Medicaid.
 To qualify for the higher matching rate for currently eligible childless adults, states must have expanded coverage to both childless adults and parents up to 100 percent of the poverty line prior to enactment of the Affordable Care Act.
 A related question is whether a blended rate is needed to ensure that states place beneficiaries in the correct reimbursement rate category. The answer to this question appears to be "no." Since 1997, states have managed significantly different matching rates for their Medicaid and CHIP programs, without evidence of a problem. In 2006, the Government Accountability Office (GAO) reported on audits of enrollment in CHIP programs conducted by the HHS Office of Inspector General. The GAO found few instances where children were enrolled in CHIP when they should have been enrolled in Medicaid. HHS has said that it is designing methods to similarly ensure proper accounting by states under the different Medicaid matching rates scheduled to take effect in 2014. See Government Accountability Office, "Children's Health Insurance: Recent HHS-OIG Reviews Inform the Congress on Improper Enrollment and Reductions in Low-Income Uninsured Children," March 9, 2006.
 The matching rate for currently eligible childless adults in states that already cover some of them (and parents) up to at least 100 percent of the poverty line would be increased each year under a formula included in the Affordable Care Act, until it equals the matching rate for newly eligible individuals by 2019.
 States that have used Medicaid waivers to expand coverage to poor childless adults typically have caps in place on the number of childless adults allowed to enroll. Childless adults who newly enroll in Medicaid in or after 2014 in numbers that exceed these enrollment caps are considered (under the Affordable Care Act) to be newly eligible individuals. As a result, in expansion states, HHS would also have to make assumptions about how many childless adults who are eligible under existing criteria, but are not served due to the enrollment caps, would qualify for the matching rate for newly eligible individuals.