Number A-3 in Series, Assessing New Federalism: Issues and Options for States
This brief draws heavily from previous and ongoing work sponsored by the Kaiser Commission on the Future of Medicaid.
The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders.
Medicaid, financed jointly by the federal government and the states, is the major source of health care funding for the low-income population in the United States.
Since states have considerable flexibility within federal guidelines, Medicaid programs vary across states in the numbers of people they cover and the amounts they spend on services.
This brief explores these variations and points out that the current federal matching formula, which provides more generous matching funds for poorer than for richer states, works to reduce interstate disparities. The two major Medicaid finance restructuring proposals currently on the table—fixed block grants and limits on per capita federal spending—could change this calculus, however.
The Medicaid Program
Medicaid programs covered 34 million low-income people in 1994 (including children and adults in families with children, disabled persons, and elderly persons). Total program spending amounted to almost $140 billion, split between state and federal sources. Children and adults in families with children together account for about three-quarters of the Medicaid population, with elderly and disabled persons making up the remaining quarter. The distribution of recipients notwithstanding, almost two-thirds of Medicaid expenditures go to the elderly and disabled.
Covered services include acute care for all eligible groups; prenatal care, delivery services, and postnatal care for low-income pregnant women; long-term care services for persons with disabilities and for low-income elderly; financial assistance to the low-income elderly to meet the cost-sharing requirements of Medicare; and financial assistance to hospitals that serve a disproportionate share of Medicaid beneficiaries and low-income persons without health insurance.
There are many reasons for variations in Medicaid coverage and expenditures. In part, these variations reflect incentives that are inherent in the federal matching formula. But they also reflect a state's ability to pay for these services, its political philosophy toward welfare families, and its recent efforts to expand coverage to pregnant women and children.
The states make very different decisions with regard to coverage and benefits in both acute and long-term care. Acute care spending per beneficiary depends on the composition of beneficiaries, the scope of a state's coverage of mandatory and optional benefits, its utilization controls, the cost of health care in the state, and its provider payment policies. Long-term care spending depends on the number of elderly and disabled people in need of long-term care services, the state's willingness to meet this need through population and benefit coverage, and its policies toward payment rates and utilization. Finally, states vary considerably in their use of disproportionate share hospital (DSH) payments. DSH payments often reflect complex financing arrangements that have allowed states to leverage federal dollars in order to make payments to hospitals serving disproportionate shares of low-income individuals. This spending is not linked to use by specific individuals.
Medicaid Spending Levels
Since Medicaid is a means-tested program, spending per low-income person provides a more accurate picture of program variation in relation to need than does spending per person in a state. State variation in expenditures per person living in a low-income family (income below 150 percent of the federal poverty line) is shown in
figure 1.
Medicaid spending averages slightly over $2,000 per low-income individual for the nation, with spending per low-income person in most New England and middle Atlantic states substantially greater than in the rest of the nation, particularly the states in the east south central, west south central, and mountain regions. Connecticut, Massachusetts, New Hampshire, New York, and Rhode Island spend over $4,000 per low-income person. At the other extreme, states such as Arkansas, Florida, Idaho, New Mexico, and Oklahoma spend less than $1,300 per low-income individual.
Medicaid covers a wide range of acute care services. Several services—hospital inpatient care; physician, laboratory, and X-ray services; outpatient and clinic services; early and periodic screening, diagnosis, and treatment (EPSDT); and payment to HMOs—form the bulk of acute care spending. Acute care spending per low-income individual averages $1,060 for the entire United States. Acute care spending varies from over $1,800 per low-income person in the District of Columbia, Massachusetts, New York, and Rhode Island to less than $700 per low-income person in Alabama, Arkansas, Idaho, Nevada, and Oklahoma,
Long-term care expenditures vary to a greater degree than acute care spending. Long-term care covers primarily institutional care—including nursing home care in nursing facilities and intermediate care facilities for the mentally retarded—which is a costly benefit, averaging over $30,000 per recipient per year. This is by far the largest component of long-term care spending, although home- and community-based services have grown in importance in many states.
Long-term care spending averages $730 per low-income person nationwide, varying from over $1,700 per low-income person in Connecticut, Massachusetts, New York, and Rhode Island to under $400 per low-income person in Arizona, California, Florida, Mississippi, Nevada, New Mexico, and Texas.
By far the greatest state-to-state variation in Medicaid spending is the use of the disproportionate share hospital (DSH) payments. DSH payments were originally designed to compensate institutions that served a disproportionate share of uninsured persons and Medicaid beneficiaries (for whom providers are frequently paid at lower than Medicare or private insurance rates).
States typically use provider taxes or intergovernmental transfers to pay the state share, which allows them to use provider funds, passed through the state treasury, to leverage federal dollars and then to return both amounts to DSH institutions. Some states make great use of this mechanism. In Connecticut, Louisiana, New Hampshire, and New Jersey, DSH payments amount to over $700 per low-income person. But a large number of states, including Arkansas, Idaho, Montana, North Dakota, South Dakota, and Wyoming, spend less than $10 per low-income person on DSHpayments.
Medicaid Coverage of the Low-Income Population
A major source of variation in Medicaid spending among states is differences in coverage of their low-income population. These differences occur in part because states have considerable discretion when establishing financial eligibility criteria for Aid to Families with Dependent Children (AFDC). (Medicaid eligibility still depends on July 1996 AFDC rules, even after welfare reform.) There are also a number of options for states in extending Medicaid coverage to low-income pregnant women and children. In addition, states can apply for Section 1115 waivers, which allow even more flexibility in extending coverage to other population groups and to higher-income individuals. Finally, states can choose to cover the medically needy (i.e., persons with incomes above cash assistance standards but with extremely high medical expenses).
The number of individuals covered by Medicaid for every 1,000 state residents is compared with the number of low-income persons per 1,000 state residents in figure 2. The whole line for each state is the number of low-income persons per 1,000 population. The darker part of the line shows how much of a state's low-income population is covered by the state's Medicaid program. Nationally, 46 percent of all individuals below 150 percent of poverty are covered by Medicaid.
States in the New England region tend to have fewer low-income people, while those in the southern regions tend to have the highest proportions of individuals in poverty. Some states, such as California and the District of Columbia, have large proportions of low-income people despite having above-average per capita incomes (not shown).
The District of Columbia, Mississippi, Tennessee, and West Virginia have the largest percentages of their populations on Medicaid. In the District of Columbia, Mississippi, and West Virginia, the large number of covered beneficiaries reflects the large numbers of people in poverty. In Tennessee, the large number of covered individuals reflects the recent expansion through the state's Section 1115 waiver.
The difference between the darker line and the whole line for each state represents the number of low-income individuals not covered by Medicaid. The states with the smallest gaps and highest coverage rates tend to be high-income states in the Northeast. For example, Connecticut, Massachusetts, New York, Rhode Island, and Vermont all cover more than 60 percent of their low-income population. States with the largest gaps (lowest coverage) are lower-income states in the South and the West. Alabama, Arkansas, Florida, Idaho, Nevada, Oklahoma, South Dakota, and Utah all cover less than 40 percent of their low-income population. Most of the high-coverage states have relatively small proportions of low-income persons. The low-coverage states have larger proportions of low-income persons. Thus, poor states like Alabama and Arkansas are covering a higher percentage of their total population than are Connecticut, Massachusetts, and New Hampshire.
Medicaid Spending per Beneficiary
Spending per beneficiary is another source of state variation in total Medicaid spending. Medicaid spending per beneficiary exclusive of DSH payments is shown in
figure 3, with the darker line showing the federal contribution.
The relative sizes of the federal and states' shares in figure 3 measure the variation in the federal matching percentage. High-spending states tend to have high per capita incomes, and thus lower federal matching contributions. Conversely, low-spending states tend to have lower per capita income and therefore a higher federal match. This serves to somewhat even out federal spending per beneficiary, which, as a result, varies less than total spending per beneficiary.
Considerable differences in federal spending do remain, however. For example, high-income states such as Connecticut, Massachusetts, and New York—which have the minimum possible federal match rate of 50 percent—have the highest expenditures per beneficiary in the nation and, thus, well above average federal payments. In contrast, states such as Florida and Tennessee have much higher federal matching rates, but lower Medicaid expenditures and federal contributions per beneficiary.
State Medicaid Spending
Variation in the amounts states spend out of their own funds per beneficiary (
figure 3) or per low-income individual (
figure 4) is vastly greater than variation in total Medicaid spending. Connecticut is at the top of the range, spending over $2,400 per low-income individual in 1994—the same as the federal share since Connecticut receives the minimum 50 percent match. This is about 9 times as much per low-income person as is spent by states at the bottom of the range, such as Arkansas, Idaho, Mississippi, and Oklahoma. Total Medicaid spending per low-income person (see
figure 1), in contrast, varies from a low of just under $1,000 in Oklahoma to a high of over $4,800 in Connecticut, about 5 times as much as Oklahoma spends. Lower levels of spending occur in low-income states despite the fact that one dollar of state spending brings in about three dollars of federal funds in those states, compared with only one dollar in high-income states.
Implications for Policy
This policy brief documents that Medicaid is really 51 different programs, with threefold variations in spending per low-income individual.
There are extensive differences in coverage of the low-income populations. There is also extreme variation in the level of states' Medicaid spending per low-income person from their own revenues. Some high-income states in the Northeast spend about 9 times as much per low-income individual as some low-income states in the South and West. Finally, federal contributions vary inversely with state per capita incomes. These federal payments partially even out the level of aggregate spending per low-income individual across states.
These variations have significant implications for public policy. The block grant proposals discussed during 1995–1996 would have distributed federal payments based on existing (1994 or 1995) spending levels. Thus, block grant distributions would have reflected current state differences in both coverage and spending per beneficiary. A per capita cap proposal, likely to be introduced in the 105th Congress, would limit the growth in spending per beneficiary. Unlike block grants, such a cap would not lock in existing coverage differences among states when determining the total amount allocated to each state. Thus, while some of the current differences across states would necessarily remain under a per capita spending cap, the long-term distributional issues would be less sharp than under block grants.
Charts & Graphs
About the Authors
John Holahan is director of the Urban Institute's Health Policy Center. His areas of special interest are health policy, health system reform, health care cost containment, Medicare, and Medicaid. Recent publications include Medicaid since 1980: Costs, Coverage, and the Shifting Alliance between the Federal Government and the States (with T. Coughlin and L. Ku) Urban Institute Press 1994.
David Liska is a research associate at the Urban Institute. His recent research has focused on the Medicaid program and state and national health care reform.
Latest Research
PDF
Print this page