|Health Policy Outlook logo 130|
No. 14, November 2009
Medicaid is the forgotten stepchild in the health reform debate. While Medicaid costs only about 76 percent of the cost of Medicare in the federal budget, it still covers 58.7 million Americans (20 percent of the population), accounts for 7 percent of the federal budget, and, on average, about 17 percent of state budgets. Government actuaries recently predicted that, under present law, Medicaid could be expected to grow at about 8.4 percent per year and increase expenditures from $352.1 billion today (2.3 percent of gross domestic product [GDP]) to $800.7 billion in 2018 (3.0 percent of GDP).
If Medicare is the eight-hundred-pound gorilla in the health reform room no one wants to talk about, then surely Medicaid is the six-hundred-pound gorilla. And, if this is indeed the year for health reform, it would be a tragedy for the poorest and most severely disabled people of this country if we did not seek to reform this troubled program.
But what are the chances for real reform? They are almost nil if we are to judge by the provisions Congress is currently considering. While the details of the House and Senate bills differ, they all expand Medicaid to include adults and children in families with incomes up to either 133 percent or 150 percent of the federal poverty line. Sliding-scale tax subsidies to help the uninsured buy health insurance from private companies or from state insurance exchanges have been proposed, but the lowest-income individuals would be required to participate in their state Medicaid program. These provisions expand the scope and expense of Medicaid but do nothing to address the fundamental problems with the present program.
Medicaid faces many problems, but the costs of the program and incentives built into its current financing that exacerbate these costs are the most fundamental. Quality and access to health care could be greatly improved if we could find a way to reform the current system of federal matching. Like Medicare, Medicaid is an open-ended entitlement program. This means the federal dollars supporting the program are not subject to congressional appropriations but are driven by the quantity and expense of the qualified claims submitted by the states. The federal government matches each claim by a formula designed to give a higher matching rate to poorer states than to wealthier ones. The state matching rates now vary from a minimum of 50 percent for the wealthier states (13 states would receive less than 50 percent if the per-capita income formula did not have a minimum) to 76 percent for the poorest state. The stimulus package has temporarily increased the rates of federal matching from a minimum of 65 percent for the wealthiest states to 83 percent for the poorest state.
Like with many things in politics, this system of financing has several unintended effects. The first is a ratchet effect on the federal and state budgets. Since no one state ever pays the full cost of its Medicaid program, each state has an incentive to expand its program when the economy is expanding and state revenues are increasing. It is the nature of a safety-net program that there are always unmet needs among the poor and disabled. When state revenues are declining, as has been the case in the current recession, each state is reluctant to cut a matched program like Medicaid since the state would have to share the savings with the federal government (for example, with a matching rate of 83 percent, Mississippi would have to cut its total program by $5.9 million to save $1 million in the state budget).
Another major flaw in the present system is the poor distribution of federal subsidies among the states; federal dollars are not flowing to the states where most of the poor people live. This is a direct result of the open-ended incentives mentioned above. The ratchet effect works on all states, but it works to a different degree in wealthier states than it does in poorer states, that is, states with higher levels of poverty. The higher-income states can afford to expand their programs in prosperous economic times to a greater extent than poorer states, but they cut Medicaid expenditures less when their economies contract. Even though wealthier states have lower matching rates, they have expanded their programs to a greater extent than the states that typically have higher proportions of poor people. (See the figure.)
Federal Medicaid Expenditures per Poor Person, 2007
To illustrate this effect, consider the following comparison of the fiscal year 2007 federal Medicaid dollars per poor person for the five wealthiest and five poorest states.
Five Wealthiest States (in descending order, by per-capita income)
Connecticut (highest), New Jersey, Massachusetts, New York, Maryland
Total federal Medicaid payments to the five states: $36.7 billion
Total number of poor people in the five states: 6.79 million (15.8 percent of the five states' populations)
Federal Medicaid expenditures per poor person: $5,405
Five Poorest States (in descending order, by per-capita income)
Utah, Arkansas, Kentucky, West Virginia, Mississippi (lowest)
Total federal Medicaid payments to the five states: $10.52 billion
Total number of poor people in the five states: 2.97 million (21.1 percent of the five states' populations)
Federal Medicaid expenditures per poor person: $3,547
While Nevada and Maine are not shown in the figure, the distribution of federal Medicaid expenditures per poor person ranges from a low of $2,014 in Nevada to a high of $7,753 in Maine, an almost fourfold difference. One of the objectives of the original Medicaid legislation was to allow states the flexibility to design a Medicaid program that reflected the preferences of their voters and their willingness to be taxed to support the program. Some states obviously prefer more extensive programs than others, but this choice has been influenced by the availability of federal, open-ended matching. This has had the effect of expanding the program in all states, while expanding it relatively more in higher-income states. The availability of federal funds has also reduced the incentive for the states to design cost-effective programs or to reduce the level of fraud and abuse.
The politics of fixing this conundrum are indeed daunting since it pits one region of the country against another, but to ignore this effect means Medicaid will continue to do a poor job of helping the vast majority of the poor and disabled. Expanding Medicaid eligibility to more adults and to higher-income families with 100 percent federal financing may have a marginal effect to narrow this distribution, but it will not change the basic incentives now built into the system of federal financing.
As authors Thomas W. Grannemann and Mark V. Pauly have argued in their new book, Reform Medicaid First (AEI Press, 2009), a strong case can be made that we should be addressing the problems in Medicaid first. They point out that there is substantial support for providing public assistance to the poorest and most disabled people in society, and that providing more consistent care for this segment of the population would go a long way toward increasing access and coverage across all states. They reason that improving the value of our expenditures on Medicaid would make it easier to cover the larger segment of the population with incomes too high to qualify for Medicaid.
Recognizing that the expected future growth of Medicaid will eventually force us to consider more fundamental ways to reform it, there are several ideas that might be considered. All of these ideas retain the present system of partial federal financing with state administration of the program, and they all attempt to correct the incentive problems inherent in the current system. Any of these reforms could be phased in as a way of giving each state time to adjust to the new system.
Give Block Grants to the States. The simplest and most straightforward way to reform Medicaid would be for the federal government to give each state an annual block grant to use in any way it wished to provide care for its poor and disabled. In addition to reducing the regulatory burden on both the federal and state governments, this would give the federal government a means to control future expenditure growth and make an explicit decision each year about how much of the federal budget should be allocated for Medicaid. At the state level, this removes the state's incentive to try to qualify for more federal funds through clever accounting schemes, while giving states more flexibility to tailor their programs to meet local needs. The federal government would still have to develop a more equitable formula for the allocation of federal funds among the states.
Reform the Federal Matching-Rate Formula. A large number of academic and government reports criticize the present formula, which is based on a state's per-capita income, and suggest alternative formulas based on target populations (the poor, the mentally and physically disabled, and long-term care populations) and each state's taxing potential. Basing the formula on these factors would reallocate federal support for Medicaid more toward states with larger poor and disabled populations and lower income-tax bases to support their programs. While this reform would reduce the present disparity in federal support per poor person, it would not necessarily reduce a state's incentive to expand (or game) the program unless it were combined with reforms that would reduce the marginal incentive to spend federal dollars.
Reduce Marginal Incentives to Expand Medicaid. Without switching to block grants or reforming the matching formula, the incentives in the present program could be improved by eliminating, or at least limiting, the open-ended nature of the present system. This could take several forms, but the basic idea would be to eliminate or reduce the federal match for certain optional benefits or for income levels above some minimum standard. An example of the latter might be no federal match (or a lower federal match) for those with incomes above 300 percent of the federal poverty line. At present, only the wealthiest states can afford these expansions, but they do so using federal funds for at least half of their expenditures. People with much lower incomes in other states are left without coverage.
Other Reform Ideas. There are a host of other reform ideas that might do some good but that do not deal with the fundamental incentive problems discussed above. These usually involve special grants or enhanced matching rates for specific programs or purposes. These include additional federal support for such things as adopting electronic medical-records systems more quickly, encouraging the use of care management in older and disabled populations, and encouraging states to rely more on managed-care plans that receive a set fee per enrollee and assume the risk of managing their care. Some of these ideas are written into the health-reform proposals now under consideration by Congress.
Tucked away in the 1,990-page House reform bill is a requirement that the Government Accountability Office (GAO) conduct a study advising Congress on the implications of removing the present federal medical assistance percentage (FMAP) upper and lower limits (50 percent to 76 percent after the stimulus bill expires in 2011) and on how to reform the formula to reflect each state's ability to pay. The cynic may say this is just the usual way for members of Congress to ignore an issue, but at least someone recognizes that there is a problem. Over the years the GAO has produced many studies criticizing the present FMAP formula, but this should be viewed as a new opportunity to inform policymakers about the potential to improve Medicaid so it does a better job of providing assistance to the poor and disabled.
Reforming the Medicaid program will not be easy. We cannot know exactly what the future will bring, but it seems clear that the present rate of growth of federal Medicaid expenditures cannot continue for many more years. Despite this fact, the proposals Congress is considering will expand Medicaid in ways that will make the present incentive problems worse. This may be good for short-term political gain, but it is not in the best interest of the nation, and it is especially not in the best interests of the millions of poor and disabled people not presently being served. These populations are being forgotten in the versions of health reform Congress is considering now.
Robert B. Helms (firstname.lastname@example.org) is a resident scholar at AEI. AEI research assistant Rohit Parulkar and editorial assistant Victoria Andrew worked with Mr. Helms to edit and prduce this Health Policy Outlook.
1. Office of the Actuary, Centers for Medicare and Medicaid Services (CMS), National Health Expenditure Projections 2008-2018, 110th Cong., 2nd sess. (Washington, DC, July 7, 2009), table 4, available at www.cms.hhs.gov/NationalHealth
ExpendData/downloads/proj2008.pdf (accessed October 29, 2009).
2. Kaiser Family Foundation, "United States: Total Medicaid Enrollment, FY2006," Kaiser State Health Facts, available at www.statehealthfactsonline.org/profileind.jsp?ind=198&cat=4&rgn=1 (accessed October 29, 2009).
3. Office of the Actuary, CMS, 2008 Actuarial Report on the Financial Outlook for Medicaid, 110th Cong., 2nd sess. (Washington, DC, October 17, 2008), available at www.cms.hhs.gov/
ActuarialStudies/downloads/MedicaidReport2008.pdf (accessed October 29, 2009).
4. Kaiser Family Foundation, "The Role of Medicaid in State Economies: A Look at the Research, Executive Summary," in Kaiser Commission on Medicaid and the Uninsured (Washington, DC, January 2009), 1, available at www.kff.org/medicaid/upload/
7075_02_ES.pdf (accessed October 29, 2009).
5. Office of the Actuary, CMS, National Health Expenditure Projections 2008-2018, table 4.
6. Department of Health and Human Services, Office of the Secretary, "Federal Financial Participation in State Assistance Expenditures; Federal Matching Shares for Medicaid, the State Children's Health Insurance Program, and Aid to Needy, Aged, Blind, or Disabled Persons for October 1, 2009 through September 30, 2010," Federal Register 73, no. 229 (Washington, DC, November 30, 2008), available at http://aspe.hhs.gov/health/
fmap10.htm (accessed October 30, 2009).
7. Bureau of Economic Analysis, "Per Capita Personal Income 2/(Dollars) [SAI-3--Personal Income Summary]," April 2009, available at www.bea.gov/regional/spi/pdf.cfm?account=
REMD&page=Drill&seltable=SA1-3&selLineCode=400&selYears=2007&rformat=display&selSort=0&end_year=2007&start_year=2007&draw=false&printable=true&selFips=00000 (accessed November 2, 2009).
8. Kaiser Family Foundation, "Federal and State Share of Medicaid Spending, FY2007," Kaiser State Health Facts, available at www.statehealthfactsonline.org/comparetable.jsp?ind=636&cat=4&sub=47&yr=30&typ=4 (accessed November 2, 2009).
9. U.S. Bureau of Labor Statistics and U.S. Census Bureau, "Current Population Survey, 2008, Annual Social and Economic Supplement," March 2009, available at www.census.gov/hhes/www/cpstables/032009/pov/new46_100125_01.htm (accessed November 2, 2009).
10. Total number of federal Medicaid payments to the five wealthiest states divided by the total number of poor people in the five wealthiest states. This represents a weighted average of federal Medicaid expenditures per poor person.
11. Kaiser Family Foundation, "Federal and State Share of Spending, FY2007."
12. U.S. Bureau of Labor Statistics and U.S. Census Bureau, "Current Population Survey, 2008, Annual Social and Economic Supplement."
13. Total number of federal Medicaid payments to the five poorest states divided by the total number of poor people in the five poorest states. This represents a weighted average of federal Medicaid expenditures per poor person.
14. See their AEI Press book, Thomas W. Grannemann and Mark V. Pauly, Reform Medicaid First: Laying the Foundation for National Health Care Reform (Washington, DC: AEI Press, June 2009), available at www.aei.org/book/100016.
15. Affordable Health Care for America Act, HR 3962, 111th Cong., 1st Sess., Open Congress (October 29, 2009), § 1747, available at www.opencongress.org/bill/111-h3962/text (accessed November 3, 2009).