Health Policy for the Low-Income Population

This report is part of the Urban Institute's Assessing the New Federalism project, a multi-year effort to monitor and assess the devolution of social programs from the federal to the state and local levels. Alan Weil is the project director. The project analyzes changes in income support, social services, and health programs and their effects. In collaboration with Child Trends, the project studies child and family well-being.

The project has received funding from The Annie E. Casey Foundation, the W.K. Kellogg Foundation, The Robert Wood Johnson Foundation, The Henry J. Kaiser Family Foundation, The Ford Foundation, The John D. and Catherine T. MacArthur Foundation, the Charles Stewart Mott Foundation, The David and Lucile Packard Foundation, The Commonwealth Fund, the Stuart Foundation, the Weingart Foundation, The McKnight Foundation, The Fund for New Jersey, and The Rockefeller Foundation. Additional funding is provided by the Joyce Foundation and The Lynde and Harry Bradley Foundation through a subcontract with the University of Wisconsin at Madison.

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.

Note: This report is also available in the PDF format, which many users find more convenient when printing.

Contents

Introduction
Background on the Study States
Medicaid Enrollment and Expenditures

Enrollment
Expenditures

Private Insurance Reforms
Publicly Supported Health Insurance Programs
Welfare Reform and Medicaid

Medicaid Managed Care
Disproportionate Share Hospital Payments

Sources of Pressure on Safety Net Providers
The Safety Net: Strong or Vulnerable?

Increasing Private and Federal Funding
System Reform
Traditional Cost-Containment Mechanisms

Introduction

In the last several years, the United States has engaged in a major debate over health and other social policies for the low-income population. At issue is whether authority for major social programs, such as Medicaid, should be shifted from the federal government to the states. This thrust toward devolution, known collectively as the "New Federalism," would fundamentally reshape federal-state relationships in many policy arenas.

The debate on health care came to a head when Congress passed legislation in 1995 and 1996 that would have converted Medicaid from an open-ended entitlement program with many federal requirements to a block grant with few federal strings. Federal spending growth would have been held to levels substantially below what they would have been under existing law. President Clinton successfully vetoed this legislation and developed his own proposal, which would have given states considerably more program flexibility than existing law did, but much less than they would have under a block grant. This legislation, too, failed to be enacted.

Even without a block grant or President Clinton’s plan, states have increasing freedom in how they run Medicaid and other health programs for the poor. The federal Balanced Budget Act of 1997 repealed federal requirements governing state reimbursement of nursing homes, hospitals, and community health centers and eliminated the need for states to obtain waivers to enroll most Medicaid beneficiaries in managed care organizations on a mandatory basis. In addition, freedom of choice, home and community-based services, and research and demonstration Medicaid waivers have become much easier for states to obtain than in earlier periods. Moreover, although numerous federal rules govern Medicaid, states have always had a significant amount of flexibility in determining eligibility levels, covered services, and methods of provider reimbursement.

At the same time that states have gained increased flexibility, major changes have occurred in the health care marketplace. First, the number of uninsured has increased, reflecting an ongoing decline in employer-sponsored health insurance coverage, especially for low-income workers. In 1990, the number of uninsured was 35.6 million; by 1997, it had increased to 43.4 million.¹ Second, managed care has become much more important in the marketplace as a whole, and for Medicaid in particular. Enrolling more Medicaid beneficiaries in managed care organizations has become the centerpiece of state policy development for acute care, both to control costs and to improve access.

The problem, however, is that the combination of rising numbers of uninsured and growing managed care penetration is undermining the ability of providers, especially hospitals, to continue to provide care to the uninsured. The implicit American social contract that held that those with private insurance would overpay hospitals, which would use the surplus to provide health care to the uninsured, is eroding. Because private managed care organizations are increasingly aggressive in negotiating contracts, hospitals that fail to accept discounted rates are losing contracts and patients to the competition. Hospitals that lower payment rates, however, experience reductions in revenues that limit their ability to provide uncompensated care. Thus, at the time when the number of uninsured is likely to continue to increase, the ability of providers to meet their needs is becoming more limited.

Both the policy debate on devolution and market changes have focused the attention of policymakers on states and how they design and administer programs for the low-income population. Advocates of devolution contend that because states are closer to the people, they will be better able to design programs that reflect local conditions and values. Opponents of devolution worry that many states will not adequately meet the needs of the low-income population and that the lack of uniform national standards could result in a "race to the bottom" with adverse impacts on the low-income population.

The Urban Institute’s Assessing the New Federalism project is a large, multi-year research program examining state activity in a wide range of programs for the low-income population, including Medicaid, public health, cash welfare, child care, and job training. This paper summarizes the results of case studies of health policy for the low-income population in the 13 states that are the focus of the project. The 13 states are: Alabama, California, Colorado, Florida, Massachusetts, Michigan, Minnesota, Mississippi, New Jersey, New York, Texas, Washington, and Wisconsin. In each state capital, interviews were conducted with state officials, legislators, interest groups, and state-based researchers and experts. In addition, to investigate the impact of state policy and market change on the health care safety net, local officials and providers were interviewed in 17 counties or municipalities. The 17 communities (some of which are the state capitals) are: Alameda County, Birmingham, Boston, Denver, Detroit, El Paso, Houston, Jackson, Jersey City, Los Angeles, Miami, Milwaukee, Minneapolis, New York City, San Diego, Seattle, and Tampa. Site visits were conducted between mid-1996 and mid-1997.

Reports were prepared on health care policy for the low-income population in each state, and they are available through the Urban Institute’s website (http://www.urban.org). In addition, cross-state topical papers (on such issues as Medicaid managed care, safety net providers, and long-term care) were written, many of which appeared in the May/June 1998 issue of the journal Health Affairs.

In brief, the findings of the case studies shed light on five broad questions: First, how do states ensure that low-income individuals have third-party health coverage? Employer- and publicly sponsored health coverage varies greatly among the states. While coverage under Medicaid and other subsidized insurance programs has significant effects on uninsurance rates for people below 200 percent of the federal poverty level (FPL), the level of employer-sponsored coverage is the main determinant of the uninsurance rates for the overall population. States with high rates of employer-sponsored coverage had much lower uninsurance rates than states with low rates of employer-sponsored coverage. The latter face much larger problems than other states in financing health care for the uninsured, either through Medicaid and other subsidized insurance programs or directly through subsidies to public hospitals and clinics.

While the importance of Medicaid cannot be overstated, more people with incomes below 200 percent of the FPL have private insurance than Medicaid coverage. Thus, recent state initiatives to reform the small-group and individual insurance market, prodded by the federal Health Insurance Portability and Accountability Act of 1996, have important implications for the low-income population.

Although extensive requirements exist for Medicaid eligibility standards, states differ in whether and how far they surpass the federal requirements, with several of our focal states not going much beyond the federal minimum standards. In an explicit attempt to maintain minimum Medicaid eligibility standards, the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996 broke the traditional link between cash assistance and Medicaid eligibility, a policy decision with critical implications for the future of Medicaid. Partly as a consequence of welfare reform, the number of low-income people on Medicaid is falling, despite state efforts to maintain enrollment for those leaving cash welfare rolls. Medicaid enrollment has not fallen as much as cash welfare rolls because of the many ways to remain eligible for Medicaid, especially for children. Nonetheless, the likely result will be higher uninsurance rates, particularly for adults, and lower increases in Medicaid program expenditures.

Second, how do states attempt to shape the financing and delivery of health care for the low-income population? Medicaid acute care policy is dominated by efforts to expand enrollment in managed care organizations. Drawn to managed care by the potential of cost savings and the promise of improved access for Medicaid enrollees, states face not only difficult implementation problems, but also a trade-off between cost savings and maintaining sufficient capacity in which to enroll beneficiaries. A combination of already low fee-for-service payment rates and concerns about protecting certain services and traditional Medicaid providers has limited cost savings. In addition, Medicaid managed care is limited predominantly to children and young adults, the least expensive populations; most states have had difficulty extending enrollment to elderly and disabled enrollees, where the greatest potential cost savings exist.

State efforts to support safety net providers, mainly through the disproportionate share hospital (DSH) provisions of the Medicaid program, have profoundly reshaped how care is financed. In many instances, however, states have turned the legislative provisions authorizing DSH payments to hospitals that serve a substantial number of Medicaid and uninsured patients into a "cash cow," enabling states to reduce or limit the amount of state funds they spend on Medicaid rather than greatly increasing the funding of safety net hospitals. States have been adept at minimizing the impact of federal efforts to control DSH spending, but the federal Balanced Budget Act of 1997, with its explicit reduction in maximum federal spending and other controls, will be a major challenge for the states. Where states have used DSH to provide significant additional funds to safety net hospitals, the federal budget cuts could have major implications for health care for the uninsured and Medicaid populations.

Third, how do state financing decisions and market conditions affect safety net providers in local communities? In all states, local governments play a critical role in providing health care for the uninsured through support of public hospitals, and a few states have sought to shore up safety net hospitals through the use of bad debt and charity care pools. Because of the growth in hospital competition and Medicaid managed care, increases in the number of uninsured, and reductions in DSH payments, the health care safety net is under enormous stress. Despite these strains, the safety net is surviving better than many observers predicted, in part because of the resourcefulness of the providers and in part because state and local governments have not been willing to let the system collapse.

Fourth, how are states attempting to control costs and improve services in long-term care, especially given that long-run demographic changes ensure huge future increases in demand? While long-term care is often downplayed at the national level, it is a critical element in state health policy and accounts for a substantial share of overall Medicaid spending. In some of our study states, long-term care for the elderly and younger people with disabilities accounts for a majority of Medicaid expenditures. States have many options to control their spending on long-term care—shifting state costs to Medicare and the private sector, expanding home and community-based services, integrating acute and long-term care services through managed care, cutting nursing home reimbursement rates, and limiting nursing home supply—but no one approach seems to dominate policy initiatives. While several of these initiatives aim at changing the balance of care between institutional and noninstitutional services, Medicaid long-term care spending for the elderly remains dominated by nursing homes in all but a small number of states. A greater shift toward community-based care has occurred in services for younger people with disabilities.

Fifth, and more philosophical than the other questions, how do states view federalism and how do their views affect state policy choices for the low-income population? States generally want more flexibility, but some federal rules chafe more than others. A key issue, which dominated the debate over the proposed Medicaid block grant, is whether the drive to reduce taxes as part of an interstate competition for businesses would result in a "race to the bottom" in publicly subsidized health care if federal rules were removed. Despite the contention that federal rules require a "one size fits all" approach to social policy, the case studies found great variation among the states, but little evidence of a "race to the bottom" in health care. Rather, state commitments to health care for low-income populations seem to be much more dependent on the favored status of health care and the state’s wealth, fiscal effort, values, and political culture.

Background on the Study States²

The Assessing the New Federalism project selected 13 states (Alabama, California, Colorado, Florida, Massachusetts, Michigan, Minnesota, Mississippi, New Jersey, New York, Texas, Washington, and Wisconsin) for in-depth study based on several considerations. The project sought states that included a large proportion of the nation’s population yet also represented a broad range with respect to geography, fiscal capacity, citizens’ needs, and traditions of providing government services. This was done to maximize the likelihood that the states selected would choose to adopt a diverse set of policies over the next few years.

All states were categorized into groups by high or low ranking on three factors: (1) child well-being (a composite of the Annie E. Casey Foundation’s Kids Count indicators); (2) fiscal capacity (total taxable resources divided by population and adjusted for cost of living); and (3) spending on Aid to Families with Dependent Children (AFDC) and Medicaid (adjusted for regional prices and divided by population under 150 percent of the FPL). States were selected from across this range of categories, with preference given to large states, and with geographic diversity taken into consideration. In general, states’ rankings on the three factors tend to cluster, with more states falling in groups ranked low or high on all three factors than in groups with mixed rankings. The 13 states selected also follow this pattern. Finally, states considered to be leaders in health (Minnesota and Washington) and welfare (Michigan and Wisconsin) reform were intentionally included.

The 13 states include three eastern states, four southern states, three western states, three midwestern states, and the four states with the largest populations (California, Florida, New York, and Texas). Populations range from more than 31 million in California to fewer than 3 million in Mississippi. Rankings on child well-being range from eighth to forty-ninth; rankings on AFDC and Medicaid spending range from second to forty-ninth; and rankings on fiscal capacity range from seventh to forty-eighth. Table 1 describes the 13 states’ rankings on the three factors used in selecting them, as well as additional characteristics of each state.

At the time the study began, the 13 states together contained half of the country’s population and more than half of its poverty population. They accounted for 57 percent of AFDC recipients, 65 percent of AFDC spending, 53 percent of Medicaid enrollees, and 54 percent of Medicaid spending.

Medicaid Enrollment and Expenditures

Health policy for the low-income population is largely, although not entirely, about the Medicaid program. Medicaid is by far the dominant public financing program for acute and long-term care services for the low-income population. In 1996, the program spent $161.0 billion (table 2) on behalf of 41.3 million people. Medicaid covered 30.5 million adults and children, 6.7 million people with disabilities, and 4.1 million elderly. The program covers a wide range of acute and long-term care services, well beyond what is usually covered in commercial health insurance policies. In addition, Medicaid DSH payments provide direct financial support for acute care hospitals and institutions for mental disease that serve high percentages of uninsured and Medicaid beneficiaries. The 13 selected states vary considerably in the proportion of their low-income population covered and their level of spending.

Enrollment

The importance of Medicaid in providing health care coverage varies across states, partly a function of the generosity of eligibility criteria and partly a function of the proportion of the population that is poor. The proportion of the population enrolled in Medicaid varies from 9.4 percent in Colorado to 20.8 percent in California and 20.7 percent in Mississippi; the average of the 13 states is 16.2 percent, slightly above the national average (table 3).

Between 1990 and 1992, Medicaid enrollment grew by an average of 11.3 percent a year for both the nation and the 13 states, with faster growth in the southern and western states. Enrollment growth was fueled by federally imposed eligibility expansions for children and pregnant women, the recession of the early 1990s, and changes in Supplemental Security Income (SSI) eligibility rules for children. Between 1992 and 1995, enrollment growth slowed to an average of 4.5 percent per year for the 13 states and 5.3 percent for the nation. This slowdown is attributable to the improving economy, declining legal immigration, the lack of new federal mandates, and tougher state welfare programs. As will be discussed below, declines in cash welfare rolls between 1995 and 1996 resulted in absolute reductions in Medicaid enrollment.

Expenditures

Total Medicaid expenditures in the 13 states in 1996 varied from $1.7 billion in Mississippi to $25.7 billion in New York (table 2). Expenditures on benefits also varied considerably across states, from $6,776 per enrollee in New York to $2,295 per enrollee in California (table 3). Differences in spending per enrollee are a function of the richness of the benefit package and reimbursement rates, as well as the extent to which the state funds its long-term care system. States where acute care is more than 65 percent of service spending—California, Florida, Mississippi, and Texas—generally have the lowest levels of expenditures per enrollee (table 4). In contrast, states where long-term care receives a relatively high share of program spending—Massachusetts, Minnesota, New Jersey, New York, and Wisconsin—have the highest levels of spending per enrollee.

Although Medicaid spending grew at a very rapid rate at the beginning of the decade, the rate of growth has slowed considerably since then (table 2). Between 1990 and 1992, Medicaid spending grew by an average of 27.1 percent a year for the nation as a whole and by 24.1 percent in the 13 states, with Colorado increasing by 44.9 percent per year and Minnesota by 15.3 percent. A substantial portion of this very high growth rate was caused by states maximizing DSH payments, but even without DSH expenditures, spending on health benefits increased by an average of 18.8 percent between 1990 and 1992 (15.1 percent in the 13 states). In contrast, between 1992 and 1995 total Medicaid expenditures grew by an average of 9.7 percent for the nation as a whole and 10.1 percent in the 13 states, with Florida growing by 13.4 percent a year and Colorado by 3.1 percent (table 2). A substantial proportion of the slowdown was caused by federal curbs on DSH payments. Between 1995 and 1996, Medicaid expenditures increased only 2.3 percent for the country and 3.0 percent for the 13 states. Almost all of the decline in total expenditure growth can be attributed to a reduction in DSH payment growth and to a decline in Medicaid enrollment.

State Efforts to Expand Third-Party Coverage

Because of the high cost of health care, most Americans have some sort of third-party coverage that pays for at least part of their medical care. Employer-sponsored health insurance is the cornerstone of the American health care system for the nonelderly population. Nearly 160 million children and adults (66 percent of the nonelderly population) were covered by employer-sponsored health insurance in 1995.³ Employer-sponsored insurance covers a much smaller proportion of the low-income population (below 200 percent of the FPL) than it does for the population as a whole, and varies much less across states. Nonetheless, more low-income persons are covered by private insurance than by Medicaid.

Given that employer-sponsored private insurance coverage is voluntary and purchase of individual policies is unaffordable for the low-income population, states operate a variety of programs to increase third-party coverage of medical care. One major approach is to shore up the private insurance system by regulating the issue and prices of insurance policies. The other strategy is to expand publicly funded health insurance, the largest being Medicaid.⁴ In addition, several states have used general assistance medical programs and have created state-subsidized health insurance to serve individuals ineligible for Medicaid.

Private Insurance Reforms⁵

Most regulation of private health insurance is done at the state rather than federal level. Typically, states regulate the content, terms of issue, and pricing of these policies. However, federal rules established by the Employee Retirement Income Security Act (ERISA) largely exclude self-insured plans from state oversight, thus precluding regulation of much of the insured population. An exception to the usual state dominance in health insurance regulation is the federal Health Insurance Portability and Accountability Act (HIPAA) of 1996, which establishes a federal floor to the issuance and renewability of health insurance and extends these provisions to self-insured plans. The law requires that insurance companies give credit for prior coverage and waive or limit waiting periods and preexisting condition exclusions, guarantee renewal of policies for small groups and individuals, and offer policies to small groups regardless of claims experience or health status. In addition, insurers are required to issue policies to certain individuals (those with 18 months of continuous health insurance coverage and no access to health insurance through a former employer, spouse, or public program). Because these requirements are quite restrictive, most observers believe that the number of people helped by HIPAA’s group-to-individual portability provisions will be small. Moreover, HIPAA did not impose any restrictions on the premiums insurers could charge the newly eligible. Thus, while HIPAA guaranteed access, it did not ensure affordability.

Though some states have been leaders in particular aspects of health insurance reform, most of the selected 13 states had to significantly strengthen their regulations in order to comply with HIPAA, especially in the individual health insurance market. Of the 13 states, only New Jersey, New York, and Washington were in general compliance with HIPAA at the time of its passage, having already enacted significant small-group and individual insurance market reforms. These state reforms, of course, did not apply to firms exempted by ERISA. California, Colorado, Florida, Massachusetts, Minnesota, and Texas had HIPAA-type reforms in the small-group market or needed to make only small modifications to be in compliance. But none of these six states had implemented significant individual insurance market reforms. Thus, all were required to enact reforms to implement the group-to-individual-market portability provisions of HIPAA, ending the laissez-faire environments for individual market insurers. This kind of regulation often leads to a reevaluation of traditional insurer business strategies and renewed attempts to select good risks by other means.

Finally, Alabama, Michigan, Mississippi, and Wisconsin did not have significant group or individual market reforms and had to enact substantial changes to conform with HIPAA. Alabama and Mississippi passed new laws that restricted insurer freedoms for the first time in both the group and individual markets. Michigan and Wisconsin passed the required group market reforms with relative ease, given their prior unregulated markets, but failed to make the required though relatively minor modifications to their long-standing mechanisms for uninsurable individuals to have access to health insurance (Blue Cross as a last resort in Michigan, a high-risk pool in Wisconsin).

The greatest challenge to states from HIPAA is implementing the group-to-individual portability provisions. HIPAA allows three options. The first provides for guaranteed issue of all private insurance products. New Jersey, New York, and Washington have required guaranteed issue, and will continue to do so. HIPAA also allows states to limit guaranteed offerings to two products that meet certain standards. Colorado, Florida, Massachusetts, and Michigan are likely to adopt this option. This could lead to some risk selection between products meeting HIPAA’s guaranteed issue requirements and other insurance products, thus possibly resulting in higher premiums for eligible individuals.

The third alternative for ensuring group-to-individual portability is to establish high-risk pools for those seeking individual insurance products. This approach would combine eligible individuals with otherwise uninsurable persons and would minimize the impact of HIPAA on private premium costs outside the risk pools. Some version of the high-risk pool option will be employed in Minnesota, Texas, Alabama, Mississippi, and Wisconsin. HIPAA required that states using pools limit premiums to no more than twice the "standard rate," defined as the premium charged in the nongroup market for a person of average risk. This provision requires the state to subsidize high-risk pool premiums with other funds, such as general tax revenues, and thus could result in more risk-spreading than in states that require private insurers to sell policies to all who apply. General tax revenue funding spreads the risk across all citizens, whereas private-market guaranteed issue spreads the risks only to those who buy nongroup policies.

The effect of HIPAA on premiums in the individual market and within high-risk pools is unknown. Increased uncertainty in insurance markets is rarely good, because insurers that are unaccustomed to guaranteed issue may price policies quite high at first, leading some currently insured persons to drop coverage. It may take individual markets a few years to reach a new equilibrium. There are also questions of how many people with health problems will obtain coverage as a result of HIPAA, and at the same time, how many others will no longer choose to purchase insurance because of premium increases. Moreover, some observers fear that insurance carriers and health plans may leave certain states or leave the business altogether because HIPAA has made their profitable risk-selection strategies more difficult.

Publicly Supported Health Insurance Programs⁶

In structuring a system of publicly funded health insurance, state policymakers have a number of options, which can be broadly classified as Medicaid, general assistance (GA) medical care, and other state-subsidized health insurance programs. Each of these presents different opportunities and obligations for state governments. Federal financial participation makes Medicaid relatively appealing; however, because Medicaid is an entitlement program, states are limited in their ability to control enrollment and expenditure levels. States must also abide by federal regulations, including provision of a mandated list of services. GA medical programs and other state-only health insurance programs allow states more flexibility in terms of program design but generally place all of the financial responsibility on state and local governments.

Historically, Medicaid eligibility has been connected to requirements for government cash-assistance programs such as AFDC and SSI. Since the mid-1980s, however, the federal government has mandated eligibility for certain low-income populations, especially children and pregnant women, with no link to AFDC or SSI. At their discretion, states may extend coverage to additional populations via optional Medicaid coverage rules. The major optional eligibility categories include the "medically needy"—people who satisfy Medicaid’s nonfinancial eligibility standards but have incomes that exceed the state’s AFDC standards—and broader poverty-related coverage for pregnant women and children. States may also expand Medicaid eligibility through federal research and demonstration waivers.

State GA programs provide cash or in-kind assistance to temporarily disabled and other low-income individuals who do not qualify for federal assistance programs. Many state GA programs provide medical assistance, which typically consists of limited acute care benefits, although some programs also cover long-term care.

Some states have established other state-subsidized insurance programs. These programs differ from GA programs in that they target a broader, more moderate-income population and are not coupled with cash assistance. The programs generally have fixed budgets and receive funding from state general revenues, special taxes (for example, health care provider or cigarette taxes), and local tax revenues. Securing and maintaining the necessary financial and political support for state-only programs can be difficult; nonetheless, the freedom from federal oversight and mandates has attracted state policymakers to such programs. States have used this flexibility to cover a wide range of populations, limit enrollment, and impose beneficiary cost-sharing requirements. Further, many states have designed their programs to promote self-sufficiency and avoid the welfare stigma often associated with Medicaid and GA programs.

Typology of states’ publicly supported health insurance systems
To assess each of the selected states’ overall approaches to addressing the problem of the uninsured, a typology was developed on the programs described above that classifies states as comprehensive, moderate, or limited. The "comprehensive" category includes states that tend to maximize Medicaid eligibility standards and thus take advantage of federal financial participation. These states have set AFDC income thresholds above the national average and have expanded Medicaid eligibility for children and pregnant women beyond federal mandates. These states also have medically needy programs. Moreover, comprehensive states use state funds to finance insurance coverage through GA medical care, state-subsidized programs, or both. The "moderate" category includes states that also tend to implement broad Medicaid eligibility standards but do little to provide insurance coverage using only their funds. Finally, the "limited" category includes states that have extended Medicaid coverage only minimally beyond federal requirements and thus have not taken full advantage of federal matching funds (despite the high federal matching rate in many of these states). In addition, these states typically have not used state or local monies to support health insurance programs for the uninsured beyond Medicaid.

Comprehensive. Massachusetts, Minnesota, New York, and Washington fall into the comprehensive category. Not only do they meet the criteria outlined above, but their commitment to creating a strong health care system is also evident in other efforts, including large hospital uncompensated care pools in New York and Massachusetts and comprehensive health care reform legislation passed in the early 1990s in Minnesota, Massachusetts, and Washington. Among state Medicaid programs, Massachusetts and New York have two of the highest AFDC income thresholds, exceeding 50 percent of the FPL. Massachusetts and New York also have two of the most generous medically needy income limits. Eligibility expansions for poverty-related pregnant women and children have been most dramatic in Washington and Minnesota. Children under age 21 and pregnant women with family incomes up to 275 percent of the FPL are eligible for Minnesota’s Medicaid program, and children under age 19 with family incomes below 200 percent of the FPL are eligible for Medicaid in Washington.

The four comprehensive states have also been very active in establishing their own state programs. Massachusetts operates three separate programs, targeted at children, unemployed adults, and disabled persons, with a combined enrollment of 53,200 as of April 1997. The MinnesotaCare program is even larger; it insured 93,000 children, single adults, and childless couples in 1995. New York’s Home Relief program stands out as being the most generous of the GA programs studied, in terms of both eligibility and service coverage. The Home Relief program had an average monthly caseload of 279,200 persons in 1995. In addition, the state’s Child Health Plus program covered 124,000 low-income children as of July 1997. Washington’s Basic Health Plan, established in 1988, was the first of its kind and has served as a model for other states. The program subsidizes, on a sliding scale, the purchase of insurance coverage from a specified group of private health plans. In July 1996, 151,000 individuals were insured through this program.

Moderate. The moderate category includes California, Michigan, New Jersey, and Wisconsin. Among these states, California has the most generous Medicaid eligibility criteria. The state’s AFDC income limit is well above the national average, and the state has extended coverage to pregnant women and infants with incomes up to 200 percent of the FPL. Michigan and Wisconsin also provide fairly generous Medicaid coverage, as does New Jersey, although somewhat less so. As discussed above, less ambitious efforts by these states to cover low-income populations with state-only funds set them apart from the comprehensive states. California and Wisconsin operate programs that cover a small number of people. Michigan’s GA program (cash assistance and medical care) suffered major cutbacks in the early 1990s. A residual medical program was established that covered an average of 11,500 persons per month in 1995. Wayne County (Detroit) runs its own program, which served 39,000 persons in 1996. New Jersey operates a relatively modest GA program, with an average monthly caseload of 22,600. In addition, the state funds a small state-subsidized health insurance program.

Limited. The remaining study states—Alabama, Colorado, Florida, Mississippi, and Texas—meet the criteria for the limited group. In particular, all five states have set AFDC income standards lower than the mean national level. Alabama and Texas have the lowest limits, at 20 percent of the FPL. These five states are not uniformly "limited" in their efforts to provide health insurance to the low-income population. Florida, Mississippi, and Texas have expanded eligibility to pregnant women and infants beyond the federally mandated levels; and Florida and Texas have medically needy programs, although income limits are well below the national mean. Florida also operates a state-subsidized health insurance program that covered 39,000 children in 17 counties in 1997.

Relationship of the typology to state uninsurance rates
The pattern of uninsurance rates is consistent with the typology discussed above: States with more comprehensive public coverage have lower uninsurance rates for the lower-income population than do states with limited coverage. Table 5 shows that the average uninsurance rate (23.4 percent) for states with comprehensive public programs is slightly lower than the average for the moderate group (24.8 percent). States with limited coverage have an average uninsurance rate (30.5 percent) for low-income individuals that is substantially higher than the average for the nation. Uninsurance rates vary considerably within the group, but much of this variation can be explained by differences in employer-sponsored coverage.

Differences in insurance coverage across the three categories show the same patterns for low-income adults and children, although the uninsurance rates of adults are much higher than those for children (table 6). Low-income adults in states with limited public health insurance programs are more likely to be uninsured (38.2 percent) than their counterparts in states with comprehensive programs (32.0 percent). The rates of employer-sponsored coverage are comparable across the three categories, but Medicaid coverage of adults in the comprehensive and moderate categories is about 40 percent higher than in the limited category.

The relationship between public coverage and the percentage of low-income children without insurance is particularly striking. The uninsurance rate ranges from 8.7 percent in the comprehensive group of states to 18.7 percent in the limited group. Rates of employer-sponsored coverage of children were similar across the comprehensive, moderate, and limited groups, and were comparable to employer-sponsored coverage for adults. However, Medicaid covered a much higher percentage of low-income children than did employer-sponsored coverage and, as with adults, Medicaid coverage rates were also highest in the comprehensive and moderate categories.

Welfare Reform and Medicaid⁷

The data in the previous section showed variations in Medicaid coverage in 1994-1995. Welfare rolls were beginning to fall during this period, often dramatically. Enrollment in cash assistance programs—AFDC, which was replaced by Temporary Assistance for Needy Families (TANF)—dropped by 26 percent from 1993 to 1997. Most of these declines began before the enactment of federal welfare reform, PRWORA, in 1996. The reductions are usually attributed to either the improved economy or state welfare reform policies that preceded PRWORA.

PRWORA established strict work requirements, limited the duration of welfare recipiency, increased state flexibility in determining eligibility, and limited Medicaid benefits for immigrants arriving after the enactment of the legislation. Adult recipients must be in work activities within two years of enrollment in TANF and may not receive TANF benefits for more than five years over a lifetime. States may impose shorter time limits and school attendance requirements and may establish benefit levels that do not vary with family size. States may also increase work incentives, raise or reduce benefits, or expand child care.

While PRWORA sought to transform cash assistance, it attempted to prevent Medicaid eligibility rules from changing. The law required states to continue to use AFDC rules in effect in July 1996 for Medicaid eligibility; at state option, the income threshold for Medicaid eligibility could increase at the rate of the Consumer Price Index (CPI). States were not permitted to use time limits in determining Medicaid eligibility. They were, however, permitted to use higher earnings disregards to determine Medicaid eligibility and more generous asset rules for both TANF and Medicaid than under AFDC.

These provisions mean that TANF and Medicaid eligibility requirements can diverge significantly from one another. In practice, states appear to be making efforts to simplify eligibility requirements and ease enrollment procedures to maintain or increase participation in Medicaid. For example, the majority of the selected states effectively raised Medicaid eligibility levels by increasing earnings disregards because of changes in TANF, and they continued to use joint application forms for TANF/Medicaid beneficiaries (rather than require separate application for each program).

Despite these efforts, Medicaid enrollment is falling, although not as rapidly as for AFDC/TANF. Between 1995 and 1996, enrollment of AFDC beneficiaries in Medicaid declined nationally by 7.3 percent, while enrollment in noncash groups (adults and children) increased by 3.2 percent. This equaled a net reduction of 2.4 percent in the total number of children and nondisabled adults in Medicaid, the first decrease in almost a decade.

Table 7 shows changes in Medicaid enrollment along with changes in monthly AFDC caseloads. The decline in AFDC/Medicaid enrollment mirrored the decline in AFDC caseloads between 1995 and 1996. The overall Medicaid caseload fell less than the welfare caseload because families leaving AFDC because of higher earnings can receive transitional medical coverage or can qualify for medically needy coverage, and children can qualify through the poverty-related criteria that do not require receipt of cash assistance. Moreover, the phased-in expansion of poverty-related coverage of older children was continuing.

On balance, however, Medicaid participation by nondisabled adults and children in 1996 declined in all of the case study states except for Washington, which enacted a major expansion of coverage through its health plan for the uninsured (Washington’s Basic Health Plan). Table 7 also shows that the welfare caseload fell by an average of 13.6 percent in 1996-97 for the 13 states. While no data are yet available on changes in Medicaid caseloads for 1997, it is likely that Medicaid experienced a greater drop in 1997 than in the previous year.

PRWORA also eliminated Medicaid benefits for legally admitted immigrants who entered the United States after August 1996. These immigrants are ineligible for Medicaid coverage during their first five years in the country. Most are unlikely to qualify for Medicaid after the first five years because the income of their sponsors must be deemed as available to them. Most people who immigrated to the country before enactment continue to be Medicaid-eligible. Some states, including California, Massachusetts, and Washington, use state funds to provide insurance to immigrants losing federal benefits, while others, such as Florida, New York, and Texas, mostly follow the federal legislation.

Several of the 13 states—especially California and New York—will be substantially affected by PRWORA’s immigration changes. The impacts will appear slowly, as immigrants newly entering the United States will be ineligible for benefits. About 7.5 percent of national Medicaid enrollees were noncitizen immigrants, but 24.9 percent of all enrollees in California were noncitizen immigrants, as were 12.6 percent in New York. Thus, while these states will not be immediately affected, the ultimate impact could be significant. If immigration continues at past rates, then a large portion of new entrants will be uninsured and ineligible for Medicaid, which could impose large costs on the states and localities in which they reside as well as on the immigrants themselves.

Financing and Delivery of Health Care

Financing and delivery of health care for the low-income population are changing rapidly. Medicaid programs are expanding enrollment in managed care organizations, although states are finding that achieving major cost savings is difficult and can conflict with other policy goals. In particular, aggressive cost savings may adversely affect safety net providers and state and local governments, limit enrollment capacity, and hurt the ability of providers to serve the sickest patients. In addition, during the 1990s, disproportionate share hospital (DSH) expenditures have grown dramatically and have been a major source of federal-state tensions as well as being a major source of revenue for safety net providers. As the federal government has tightened the rules on DSH, states have been forced to look to other mechanisms to maximize federal revenue and to aid safety net hospitals.

Medicaid Managed Care⁸

Medicaid managed care is growing quickly in the United States, and it dominates the acute care policy agenda of most Medicaid programs. Forty-nine states now rely on some form of Medicaid managed care, and enrollment has grown from 9.5 percent of total Medicaid enrollment in 1991 to 40.1 percent in 1996. In addition, managed care is moving from limited primary care case management (PCCM) approaches to more comprehensive, fully capitated managed care systems.

Despite interest in extending managed care more broadly, the 13 selected states have primarily enrolled only the AFDC/TANF and related populations into Medicaid managed care, and then not always on a statewide basis. The AFDC/TANF and related populations are by far the largest number of Medicaid beneficiaries as well as those that are most attractive to Medicaid managed care plans. But limiting enrollment to these lowest-cost groups also limits the Medicaid savings potential. States that restrict enrollment to these groups and include only acute care services leave approximately 75 percent of Medicaid spending outside of managed care.

The SSI population, the medically needy, and Medicaid-Medicare dual-eligibles are the groups with the greatest per-enrollee costs and the most likely to benefit from care management. For example, in 1996, average Medicaid acute care spending per enrollee for the disabled was $5,058 per year versus $1,821 for nonelderly adults and $1,070 for children. However, states face severe problems in enrolling people with disabilities in managed care because this group includes individuals with a wide variety of complex illnesses such as AIDS, tuberculosis, physical disabilities, developmental disabilities, and mental illness. As a result, it is difficult to establish contract provisions that ensure appropriate care, access to specialty care, and risk-adjusted payment rates. Some states (Mississippi, New Jersey, Texas, and Wisconsin) permit SSI recipients to enroll in managed care on a voluntary basis; others (Colorado, Florida, and Michigan) mandate enrollment but give beneficiaries a choice between a PCCM program and a fully capitated health maintenance organization (HMO). Washington began mandatory enrollment for its SSI population but found costs were too high and soon reverted to fee-for-service.

Medically needy and dual-eligible beneficiaries are also difficult to enroll in managed care because of their high costs and complex needs. In addition, medically needy beneficiaries may go on and off the Medicaid program as their health needs change over time. For dual-eligible beneficiaries, coordinating with the Medicare program is extremely difficult, and mandatory enrollment in capitated plans is prohibited except for the delivery of Medicaid services (see the discussion below on the integration of acute and long-term care services).

Medicaid managed care programs often started as PCCM arrangements but are evolving into fully capitated plans. PCCM programs match beneficiaries with providers who do not bear financial risk but take responsibility for providing primary care and making referrals for specialist care and hospitalizations. Full-risk HMOs, in contrast, are capitated for a comprehensive set of services and have greater incentives to control utilization and medical costs. The 13 study states employ various models. At one extreme, Alabama and Mississippi, which are only beginning to implement Medicaid managed care, have expanded PCCM programs because neither has viable HMOs in most geographic areas. Washington is at the other extreme, with virtually all of its AFDC/TANF and related beneficiaries enrolled in fully capitated HMOs. A number of other states are in the process of shifting enrollees from PCCMs to HMOs.

States also vary in the extent to which they are contracting with commercial or mainstream HMOs or relying upon plans formed by or otherwise centered on traditional Medicaid providers, often public hospitals. Medicaid programs originally hoped that mainstream managed care plans would offer beneficiaries broader access to providers. Some observers believe that mainstream or commercial plans became interested in Medicaid managed care when enrollment was voluntary and HMOs could benefit from favorable selection. This viewpoint holds that mainstream plans are becoming less interested as managed care becomes mandatory for more enrollees, in part because there is less opportunity to benefit from favorable selection. Others believe that mainstream plans find Medicaid more attractive under mandatory arrangements because there are many more beneficiaries. But it is generally agreed that if states attempt to limit the growth of capitation rates while at the same time increasing the regulation of plan performance, mainstream plans could become increasingly reluctant to participate in Medicaid.

An issue with mainstream plans is that they have been less likely to contract with safety net providers, seeking instead to move beneficiaries to less expensive hospitals. Safety net providers, including public hospitals, community health centers, and local health departments, need Medicaid revenues not only to provide care to Medicaid beneficiaries but also to subsidize care for the uninsured. Because of concerns about the threat of managed care to the continued survival of safety net providers, states have established a variety of protections for them. These protections include higher capitation rates for plans contracting with safety net providers and favorable treatment in the assignment of beneficiaries. These policies have made it increasingly difficult for commercial plans to compete. The result could be, at least in many states, a reliance on relatively few plans organized around traditional Medicaid providers, a return to historical patterns of care with a change in payment arrangement, and perhaps some improvement in the management and efficiency of care delivery.

States clearly expect to see improvements in access and quality of care as a result of increased use of managed care. In fact, Medicaid authorities believe that managed care is making providers more accountable for quality of care than they have been in the past under fee-for-service. States are establishing mechanisms to deter poor-quality care, monitor plan performance, and provide recourse for beneficiaries when problems arise. They are setting standards in managed care contracts related to the adequacy and appropriateness of provider credentialing, appointment availability, and beneficiary appeal procedures. States are also requiring reports based on the Medicaid HEDIS (Healthplan Employee Data and Information Set) system for assessing performance, as well as member surveys to identify problems associated with access to care and patient satisfaction.

However, enforcement of these quality requirements is another issue. Some of the study states had limited staff capabilities to monitor quality. Other states had so few plans that they were sometimes reluctant to apply sanctions for fear of losing capacity. In other cases, states were concerned about the impact of financial sanctions on plans tied to safety net providers, some already in precarious financial situations.

It was also evident that a major goal of Medicaid managed care is cost savings. Early research showed that Medicaid managed care, even without relying on capitated payments, could result in savings of up to 15 percent relative to fee-for-service, and some people have envisioned even higher savings targets.⁹ In the case-study states, savings expected from managed care were on the order of 5 to 10 percent, relative to fee-for-service. These lower savings expectations in part reflect historically low Medicaid payment rates to hospitals and physicians, which make it difficult to achieve the price reductions often seen in the private sector. States seem to be increasingly reluctant to attempt to achieve greater savings because they recognize the need to maintain a large Medicaid managed care capacity to promote access and competition in the long run.

A key question is how plans are achieving savings (if any), given that their rates are based on (and lower than) already low fee-for-service levels and that they must cover plan administration expenses as well as services. States expected that excess capacity of hospitals and specialists, along with the recent decline in private payment rates, would make providers more willing to accept Medicaid patients at lower rates. In fact, plans have succeeded in reducing hospital payment rates. Medicaid policymakers also believed that managed care plans were reducing inpatient utilization and inappropriate emergency room visits, lowering the use of specialty care and shifting patients to more efficient providers.

The 5 to 10 percent savings assumption could overstate the amount of true savings to the state from Medicaid managed care because of state Medicaid maximization initiatives. In the last decade, states have used Medicaid to fund a broad set of services and providers, in part to bring in federal funds for these services. Over time, school-based services, case management, and other social services, as well as state and local public health and mental health programs, have become increasingly financed by Medicaid. There is growing concern that Medicaid managed care plans will not actually provide these services or will not contract with government providers to do so. As a result, government agencies may still have to provide the services, but without the benefit of federal Medicaid reimbursement. Thus, to some extent, the state may actually lose money on Medicaid managed care.

Finally, Medicaid programs are increasingly aware that managed care poses a threat to safety net institutions. Safety net providers are expected to form managed care plans or join existing ones. In either case, safety net providers will most likely have to accept lower payments than they had under fee-for-service and consequently reduce costs in order to compete for managed care patients. Thus they face the threat of lower volume and lower revenues per patient for Medicaid. To the extent safety net providers become more efficient or patient-oriented, a benefit clearly exists, but to the extent that lower Medicaid revenues reduce their ability to fund care for the uninsured, states and localities may need to finance this care through other mechanisms. As a result, the savings from managed care may be less than anticipated.

Disproportionate Share Hospital Payments¹⁰

The Medicaid DSH program was established by Congress to increase funding of "hospitals that serve a disproportionate number of low-income patients with special needs." Expenditures for DSH increased sharply in the early 1990s, from $1.3 billion in 1990 to almost $18 billion in 1992. DSH expenditures have not grown much since then, but they still account for almost 10 percent of federal and state expenditures on Medicaid. The 13 states accounted for about 60 percent of total Medicaid DSH expenditures (table 8). As measured by the amount of DSH payments per low-income individual or the amount of DSH payments per uninsured individual, Alabama, California, Colorado, Massachusetts, New Jersey, and New York were high-expenditure DSH states, and Florida, Minnesota, and Wisconsin were low-expenditure DSH states.

The DSH legislation was enacted in the early 1980s to help hospitals that were losing money on Medicaid patients because of low Medicaid reimbursement rates. These hospitals often served many indigent patients and had high levels of uncompensated care; they were less able to cross-subsidize uncompensated care because they also had few privately insured patients. A revision to the DSH legislation in the Omnibus Budget Reconciliation Act (OBRA) of 1981 allowed states to pay rates above the Medicare "upper payment limit" to hospitals rendering high volumes of care to the poor.¹¹

To generate federal DSH payments, states had to find a way to finance their share of the payments. Provider taxes and voluntary donations permitted them to do so with no use of state general revenues, spurring the growth of DSH expenditures. States required hospitals to make a donation or pay a tax on Medicaid services. Hospitals received federal DSH funds, along with the return of all or most of the donation or tax payment. States quickly found that these arrangements could leverage large amounts of federal dollars at no cost to the state.

States were also able to benefit fiscally from these transactions if not all of the federal DSH money was paid to hospitals. That is, if some of the funds were retained by the state, the states could make their treasuries better off at the expense of the federal government. For example, assume a state collected $100 million in provider taxes paid by hospitals. Then the state makes a $150 million DSH payment to tax-paying hospitals. Assuming that the state’s Medicaid match rate is 50 percent, the state collects $75 million in federal reimbursement. At the end of the transactions, the hospitals have netted $50 million in DSH payments, and the state has netted $25 million in federal money. The federal government has made $75 million in DSH payments; however, only $50 million was paid to hospitals, with the balance retained by the state.

The ability to provide additional revenue to hospitals as well as fiscal relief to states made the use of DSH payments highly attractive to states, especially in a period when states faced new federal mandates and declining economic conditions. The rapid rise in DSH payments, however, generated considerable controversy among federal policymakers who felt that the states were abusing the program. In 1991, Congress enacted the Medicaid Voluntary Contribution Provider Specific Tax Amendments, which banned provider donations, limited provider taxes, and required that hospitals not be held harmless for their tax payments. The legislation also capped DSH payments at 1992 levels. States whose DSH payments were 12 percent or more of total Medicaid expenditures (high-DSH states) in 1992 could not exceed this dollar level in the future. States whose DSH payments were less than 12 percent (low-DSH states) could increase them at the same rate as their overall Medicaid expenditures. This law essentially stopped DSH payment growth and required states to find new sources of revenue for their share of DSH outlays.

Another federal concern with the DSH program was that states were making DSH payments to providers that were not significant providers of care to the poor and that some providers were receiving DSH payments that substantially exceeded their financial losses in serving Medicaid and uninsured patients. Typically, in these cases, DSH payments were substituting for payments the states or localities otherwise would have made, and thus were means of providing fiscal relief to state and local governments. In response, OBRA 1993 legislation contained provisions that prohibited DSH payments to hospitals with less than 1 percent Medicaid utilization and required that total DSH payments to a hospital not exceed that hospital’s losses on Medicaid and uninsured patients. The latter provision is sometimes referred to as a hospital-specific cap.

The case study states varied as to whether they were classified as high- or low-DSH states by the 1991 legislation, how they financed the state’s share of DSH payments, and the share of the federal DSH payments that was retained by states or passed on to hospitals. These states responded to the 1991 and 1993 legislation in several ways. In an immediate response to the 1991 DSH law, most of the 13 states made fundamental changes in the way they financed their share of the DSH program. Because donations and most provider taxes were no longer eligible for federal reimbursement, several states responded by raising the state’s share through intergovernmental transfer (IGT) programs. Typically, the IGTs were made by state and local public hospitals or by another agency of state government to the Medicaid agency and, correspondingly, DSH payments were made back to the same hospitals. Some states, such as New York and Massachusetts, that generated the state’s share of DSH expenditures through hospital payments into an uncompensated care pool had provider taxes that met the requirements of the 1991 legislation, and therefore did not change their financing mechanisms. Minnesota, with a very small DSH program, continued to finance its state share with general revenues.

A second response by many states to the 1993 OBRA legislation was to begin making or to expand DSH payments to "institutions for mental diseases (IMDs)" (mostly mental hospitals). The use of IMDs allowed states to fully spend their DSH allotments while being consistent with the 1993 OBRA facility-specific caps (because of the large percentage of uninsured persons in IMDs). Moreover, because many IMDs are public institutions owned by state or local governments, the same entities that make the IGTs benefit from the DSH payments. Federal lawmakers are concerned, however, that these DSH payments seem to circumvent the long-standing prohibition of using Medicaid to pay for IMD services for persons between the ages of 22 and 64. Through DSH payments to IMDs, federal dollars once more have replaced state dollars and provided fiscal relief to states, clearly not the intention of the original DSH program.

A third response to the 1991 and 1993 legislation is that some states, including Colorado, Michigan, and Minnesota, found it necessary to reduce DSH spending. In Colorado, for example, legislators cut funding, fearing that the state would be adversely affected by future federal cutbacks and politically would be obligated to replace federal spending with state dollars. Colorado was also faced with a state revenue limit that curtailed its ability to raise the state share.

Perhaps the most important response to the 1991 and 1993 DSH provisions was that some states began making supplemental payments to selected hospitals and other providers. Typically these payments are made to public facilities as add-ons to regular Medicaid reimbursement for services. These supplemental or enhanced payments do not count against the state’s DSH allotment or an individual hospital’s DSH cap. In using this mechanism, the state may not pay more than the Medicare upper payment limit, but if Medicaid reimbursement rates are sufficiently low, the supplemental payments allow states to make substantial increases to selected hospitals without exceeding the 1991 or 1993 DSH caps. Michigan was particularly active in employing this strategy, making supplemental payments to state-, county-, and city-owned public hospitals. The state’s share of these payments came from IGTs paid by the institutions receiving the supplemental payments.

States used several other strategies to avoid the federal restrictions. In response to the 1993 legislation imposing hospital-specific caps, Alabama developed a network of prepaid health plans (PHPs) that received capitated payments for all Medicaid inpatient care in their geographic area. The state sets the capitation rates for the PHPs and includes DSH payments in the calculation of rates. The PHPs then allocate the DSH funds. By including DSH payments in the PHP capitation rates, the state avoids the 1993 hospital-specific caps that were making it difficult for the state to fully spend its DSH allotment.

Several states received specific legislative or regulatory exemptions or supplemental funds to lessen the impact of the 1991 and 1993 legislation. California, through a Medicaid "research and demonstration" waiver, received additional federal funds for the Los Angeles County health system, with the condition that services be restructured to provide more ambulatory care. New York was granted a similar provision in its Partnership Plan research and demonstration waiver, with the funds intended to help safety net hospitals make the transition to Medicaid managed care.

Massachusetts, through its research and demonstration waiver, was able to use federal funds to support two major Boston-Cambridge hospitals that provide large amounts of uncompensated care. These federal funds replaced money that hospitals had received from the state’s uncompensated care pool, allowing the state to use pool funds to assist other hospitals. The extra payments made to the two Boston-area hospitals do not count as DSH payments. Instead, these two hospitals established capitated managed care plans for previously uninsured individuals who were extended Medicaid coverage as part of the research and demonstration waiver. The new managed care plans were paid an enhanced capitation rate, making up for the hospitals’ loss in DSH and pool funds.

States are now faced with the Balanced Budget Act (BBA) of 1997, which included several important DSH provisions. One provision changed the DSH allotments allowed by the 1991 law and replaced them with new state-specific allotments. All states face some reduction from current law; the Congressional Budget Office estimates reductions of $10.4 billion over the 1998-2002 period. In addition, the new law places limits on how much of a state’s federal DSH allotment can be made to IMDs.

The impact of the BBA is uncertain. To the extent DSH payments have been paid to hospitals, the BBA will mean a reduction in revenues to safety net providers, but many DSH expenditures have meant fiscal relief to states or have replaced funds that states would have otherwise spent financing health care institutions, particularly IMDs. In those states where Medicaid DSH payments were going to safety net providers, the BBA could result in significant hardship for many hospitals. These safety net providers are under financial stress from many forces. It is also important to remember that while federal DSH payments often replace state funds, and thus provide fiscal relief to states, this replacement of state expenditures occurred in the past, and federal dollars now make major contributions to safety net providers. The key question is whether a reduction in federal DSH payments will be offset by increases in state and local payments to these facilities.

Safety Net Providers: Pressures and Responses¹²

In every community visited as part of the Assessing the New Federalism study, a significant number of ambulatory and inpatient providers served the low-income population, both Medicaid-eligible and uninsured, forming the traditional health care safety net. Safety nets can be divided into three categories: those in which uncompensated hospital and emergency care is dominated by state, county, or other publicly owned systems (e.g., Alameda County, Denver, El Paso, Houston, Jackson, Los Angeles, Miami, and Minneapolis); those in which public and privately owned systems or providers are more evenly mixed (e.g., Birmingham, New York City, Seattle, San Diego, and Tampa); and those in which the safety net comprises only private providers (e.g., Detroit and, recently, Boston and Milwaukee). These hospitals not only provide inpatient and emergency care services but also generally run outpatient clinics. Another important category of outpatient providers is nonprofit community health centers, which receive federal grant support and cost-based Medicaid and Medicare reimbursement as "federally qualified health centers." Finally, county and city health departments, in addition to carrying out public health responsibilities, deliver maternal and child health services and, in some jurisdictions, a more comprehensive set of primary care services.

The safety net can also be divided between institutions providing general care to the uninsured and institutions providing care to very vulnerable, hard-to-reach populations, including the homeless, substance abusers, illegal immigrants, and people with HIV/AIDS. Despite some overlap between the two types of institutions, particularly in the highly integrated public models, often the health department or other local clinics serve these vulnerable, hard-to-reach populations. This distinction between "average" uninsured and "special or harder-to-serve" uninsured also seems to apply to the Medicaid population, with private institutions seeking to serve the AFDC/Medicaid clientele, leaving public institutions to serve the SSI, AIDS, homeless, and substance abuser groups.

Sources of Pressure on Safety Net Providers

Local safety net providers are under stress from a variety of sources in ways that may compromise their ability to provide services to Medicaid and uninsured individuals.

Level of uninsured
The proportion of a community that is uninsured is one of the most important determinants of a safety net system’s health, or the level of pressure it encounters. In communities with higher levels of uninsured, as in California, Florida, and Texas, the safety net is under greater pressure than in those with lower uninsured rates, such as Minnesota, Washington, and Wisconsin. Safety net systems in communities with high uninsurance rates tend to be more reliant on federal, state, and local support, while those in communities with lower uninsurance rates can support their uncompensated care burdens more easily through internal cross-subsidies.

Marketplace competition
The growth in managed care can potentially threaten safety net institutions. In some of the study states, such as Alabama and Michigan, fee-for-service indemnity insurance remains strong; in others, such as California and Minnesota, managed care has been an active force in the marketplace for many years. In states where indemnity insurance retains a strong presence, safety net hospitals are often able to finance charity care through cross-subsidies from third-party payments. In contrast, where managed care is mature and marketplace competition is intense, the ability of safety net providers to earn surplus revenues from privately insured patients has diminished. Managed care plans generally reduce payment rates, admissions, and length of stay, and move private patients away from expensive safety net hospitals to lower-cost community hospitals. Increased competition can indirectly affect safety net hospitals as well. If competitive pressures limit the ability of many hospitals to provide even small amounts of indigent care, more of the burden is placed on the safety net institutions.

In states such as California, Massachusetts, and Minnesota, where the managed care market is more mature, mergers and consolidations of hospitals and alliances between hospitals and physicians have proliferated. Hospital mergers and consolidations can have conflicting impacts. Mergers can reduce duplication and increase efficiency, thus lowering costs to managed care plans and making health coverage more affordable. But mergers will also reduce the number of competitors in a marketplace, making it difficult for managed care plans to drive hard bargains, and can result in higher health care prices. Mergers can increase market power of key hospitals within a marketplace, enabling them to extract surplus revenues to continue providing charity care. But in some cases, mergers can increase the intensity of competition in a market, making it less likely that hospitals can afford the luxury of providing charity care.

For-profit competition
In the case-study sites in California, Texas, and Florida, safety net providers indicated that the presence of for-profit competitors increased the level of pressure on hospital rates and thus on revenues. If conversions to for-profit status continue at their previous pace, some observers argued, the ability to support charity care will be increasingly threatened. Competitive pressures seem greater in sites with a higher proportion of for-profit hospitals such as those in California, Florida, and Texas, than in the remaining sites, where the presence of for-profit hospitals was smaller or nonexistent.

Competition for Medicaid patients
Across virtually all the locales visited, safety net providers reported that Medicaid revenues, generated through both patient care services and DSH programs, are vital to their ability to provide services to the uninsured and special enhanced services to vulnerable populations. Observers believe that nontraditional Medicaid providers are expanding into the Medicaid market as a result of competitive pressures in both the private fee-for-service and managed care markets. The degree of competition for Medicaid beneficiaries has also grown because of expansions in Medicaid eligibility in the early 1990s and, more recently, the increasing penetration of managed care. Competition for Medicaid enrollees appears to focus on subsets of the Medicaid population, such as families with children and newly eligible pregnant women and children, but not on the disabled or children with special health care needs.

For inpatient safety net providers, changes in the distribution of Medicaid patients across providers as a result of competition can have significant implications for care for the uninsured. To begin with, Medicaid revenues have been used to cross-subsidize care for the uninsured. At the same time, the redistribution of Medicaid patients has potentially severe implications for DSH revenues. In those states, such as Texas and California, that distribute DSH funds on the basis of Medicaid patient loads, safety net providers losing Medicaid patients also can expect reductions in their DSH funds.

Public health departments have also been affected by escalating competition for Medicaid patients. Many health departments have experienced considerable declines in their Medicaid caseload because, increasingly, beneficiaries are enrolling in managed care organizations that link them with a private physician. In some cases, Medicaid beneficiaries have continued to seek care at health departments; however, unless the health departments are in a managed care plan’s network of providers, they are not reimbursed for these services. With less demand from Medicaid patients for services and fewer Medicaid revenues available to cross-subsidize care for the uninsured, some local health departments may have to close their clinics. This move could have adverse consequences for the uninsured, who have limited options for health care services.

While some state and local health departments are concerned about the impact of Medicaid managed care, others welcome it as an opportunity to return the focus of public health to core population-oriented services (such as disease control and environmental health). Still others plan to assume a broader oversight role of managed care delivery or to provide "enabling" services, such as case management and transportation, under contract with Medicaid managed care plans.

State and federal support of the safety net
As the primary method for states to directly subsidize hospitals that provide a disproportionate share of care to the low-income population, DSH funds are extremely important to safety net hospitals. States’ commitments to these programs, however, vary considerably. In New Jersey, New York, and Massachusetts, state governments have attempted to redistribute the costs of uncompensated care across providers using broad-based taxes on insurers and providers and to supply more funds per uninsured individual than most other states. DSH payments also provide substantial support for the safety net in Alabama, California, Colorado, and Mississippi. Florida, Minnesota, Washington, and Wisconsin have relatively low DSH payments, but in the last three states, uninsured rates are also low. In Texas, DSH payments are large, but a relatively high share goes to state mental hospitals, reducing the amount available for general hospitals. As noted above, the Balanced Budget Act of 1997 will reduce the amount of federal DSH payments in the future. While safety net institutions may not have benefited as much as is commonly believed from federal DSH payments (because they often merely replaced state and local monies), they could be adversely affected by cuts in federal DSH payments if state and local governments do not move to replace this funding.

Local support
Local revenues are important to all safety net providers—particularly county- or city-owned hospitals—and are especially important in states with high uninsurance rates, high levels of private sector competition, and low DSH payments. In some states, local support has been particularly significant. In both Dade and Hillsborough counties in Florida, local support for indigent care has risen considerably as a result of increases in local sales taxes. Despite Florida’s experience, local taxes as a primary source of revenue seem a somewhat tenuous source of support because the suburbs do not necessarily want to support care for the low-income uninsured in the inner cities, and because of the current anti-tax climate. For example, in Houston, the Harris County Hospital District experienced a reduction of almost 70 percent in its local revenues between 1992 and 1995, a result of significant reductions in local property taxes. In addition, New York City has eliminated the tax levy subsidy of the public hospital system.

The Safety Net: Strong or Vulnerable?

Although the pressure on the safety net is increasing, safety net systems in the selected sites appeared to be relatively secure. Individual safety net providers may be in severe financial straits, and some have even shut down, but the systems that ensure access to care for the uninsured remain firmly in place. How have they survived the pressures that are building? The answer is threefold: (1) very few communities have all factors working against them at the same time, (2) the majority of providers have responded quickly and effectively to the pressures as they have arisen, and (3) states and localities have also responded quickly and effectively to ensure that safety net providers will survive.

None of the communities had high levels of pressure in all factors influencing safety net strength or vulnerability. The safety net did face varying degrees of pressure among the communities. Table 9 lists the 16 communities and the factors exerting pressure on local safety nets, splitting the communities into three groups: safety nets that are (1) at significant risk—California, Colorado, Florida, and Texas—with high demand, high competition, and high reliance on local funding for support; (2) at some risk—Alabama, Mississippi, and New York—with high demand but relatively low levels of competition; and (3) the least vulnerable—Massachusetts, Michigan, Minnesota, Washington, and Wisconsin—with low demand and low levels of competition (although high or medium levels of private and Medicaid managed care).

Safety net providers have responded to changing circumstances swiftly and, at least so far, productively. In fact, safety net providers’ response to changes in the market differs little from that of their non-safety-net counterparts. Virtually all providers are making concerted attempts to sign up with managed care organizations, develop commercial or Medicaid managed care products, reduce costs and increase efficiency, improve the quality of care or service to patients, better manage care for the uninsured, and lobby for financial support at the federal, state, and local level.

Governments at all levels have shown themselves willing to support the safety net. Although there have been well-publicized crises, the safety net systems (and nearly all of the providers) survived mainly because of government intervention. In Los Angeles, county officials persuaded the federal government to agree to a $364 million bailout, allowing the Los Angeles Department of Human Services to restructure the hospital system on a more gradual basis. Over a five-year demonstration period, the federal government is expected to spend more than $1.1 billion in Los Angeles County to help stabilize the system’s finances, match local funds to serve indigent patients in outpatient settings, and defray costs of the restructuring. In Milwaukee, where John Doyne Hospital (the city’s public hospital) failed to survive, state and local policymakers decided to continue financial support for a safety net system, despite closing down the major safety net provider. In Tampa, an innovative program to cover the majority of the uninsured developed by the County Commissioners helped Tampa General Hospital weather its financial storms. These experiences suggest that, when push comes to shove, federal, state, and local governments have been unwilling to let the safety net completely collapse.

While safety net providers have thus far proven resilient, the pressure to increase efficiency and reduce excess is likely to grow. State coverage programs, state and local subsidies, and Medicaid patient care revenues will become increasingly important. Although the site visits provided evidence that safety net providers can survive a competitive market, changes in Medicaid patient care revenues, and reductions in state and local subsidies, they did not provide evidence that the safety net can survive in the presence of all these changes at the same time.

Long-Term Care for the Elderly¹³

Long-term care services for older adults represent a substantial share of total health care spending in the United States and are an area of major concern for state policymakers. Nursing home and home health care accounted for almost 12 percent of personal health expenditures in 1995 and approximately 14 percent of all state and local health care spending.¹⁴ Neither private insurance nor Medicare covers long-term care to any significant extent, and few older adults have private long-term care insurance. The disabled elderly must rely on their own resources or, when these are depleted, turn to Medicaid or state-funded programs to pay for their long-term care. Because of the high cost of long-term care (a year in a nursing home cost an average of $46,000 in 1995), Medicaid coverage for long-term care provides a safety net for the middle class as well as the poor.¹⁵ In 1997, 68 percent of nursing home residents were dependent on Medicaid to finance at least some of their care.¹⁶ Medicaid long-term care expenditures for the elderly are projected to more than double in inflation-adjusted dollars between 1993 and 2018 because of the aging of the population and price increases in excess of general inflation.¹⁷

Almost $56 billion was spent on long-term care for people of all ages by the Medicaid program in 1996, 36 percent of total Medicaid expenditures. Long-term care spending on older beneficiaries accounted for the majority ($31 billion) of this spending. In that same year, older persons accounted for 9.9 percent of all Medicaid enrollees and 27.3 percent of total Medicaid expenditures. Long-term care services accounted for three-fourths of Medicaid expenditures for the elderly.

Table 10 shows Medicaid long-term care spending on the elderly for the 13 states and the United States, spending on these services as a percentage of total Medicaid spending, spending per elderly enrollee and resident, and the proportion of expenditures by type of service. There is considerable variation across states. While long-term care expenditures for the elderly accounted for 20.1 percent of all Medicaid spending in the United States in 1996, this proportion ranged from 12.1 percent in California to 30.1 percent in Minnesota among the 13 states. Per-elderly-resident spending for long-term care varied from a low of $340 in Florida to a high of $2,673 in New York.

Nationally, more than 82 percent of these long-term care expenditures for the elderly were for nursing home care, about 12 percent were for home care services, and the remaining 6 percent were for intermediate care facilities for the mentally retarded and mental health services. Medicaid long-term care spending for the elderly is more institutionally based than it is for younger people with disabilities. The proportion of long-term care spending for the elderly for nursing facilities ranged from 65.8 percent in New York to 98.5 percent in Mississippi. These same states were also the extremes in the proportion of Medicaid spending for home care: 0.2 percent in Mississippi and 25.3 percent in New York. In 1996, New York accounted for more than 40 percent of all Medicaid home care expenditures for the elderly. In addition, some states, such as California, Florida, Massachusetts, and Wisconsin, have significant state-funded long-term care programs that do not appear in these data.

Like the rest of the Medicaid program, states have considerable flexibility in the provision of long-term care services, and reform efforts differ across the states. In fact, the strategies used by states to control long-term care expenditures are much more varied than for acute care, where there is a single-minded focus on increasing managed care enrollment. Overall, states use three broad strategies to control state spending on long-term care: bring more outside resources (such as private resources, federal Medicaid funds, and Medicare) into the long-term care system to offset state expenditures; reform the delivery system to provide care more efficiently; and reduce Medicaid eligibility, reimbursement, and services. Not surprisingly, the 13 states differ in the extent to which they focus on each of these strategies and how far each state has progressed in implementing substantial long-term care reform.

Increasing Private and Federal Funding

One strategy to reduce state expenditures is to substitute private and Medicare financing for state funding. While some have heralded private long-term care insurance as a potential fix for rising Medicaid long-term care expenditures, little progress has been made in this area and only two case-study states seem seriously committed to this strategy. The "public-private partnerships" in California and New York that have generated so much controversy at the national level have failed to attract many participants and are simply not important sources of financing at this time. These programs allow individuals who purchase state-approved private long-term care insurance policies to qualify for Medicaid nursing home benefits while keeping far more in assets than is usually allowed. The California and New York partnerships have spurred the purchase of fewer than 17,000 policies in the two states combined, even though more than 6 million older persons reside in the two states.

Over the last decade, policymakers and the media have focused attention on middle-class and wealthy elderly persons who transfer, shelter, and underreport assets to appear poor enough to qualify for Medicaid-financed nursing home care. While almost all of the selected states believe that so-called "Medicaid estate planning" is somewhat of a problem, it is a major concern only in Massachusetts, New Jersey, and New York, and few states are engaged in major efforts to combat it. On a related issue, the Omnibus Budget Reconciliation Act of 1993 has achieved its goal of states’ establishing programs to recover the costs of long-term care from the estates of deceased Medicaid beneficiaries, except in Texas and Michigan, where there is strong political opposition to these programs. But few observers expect major savings.

States have long sought to shift Medicaid long-term care expenditures to Medicare, but have been frustrated by narrow Medicare coverage of nursing home and home health care. That situation has changed dramatically since 1989 when Medicare post-acute-care coverage rules were liberalized, making the benefits much more long-term-care-oriented. Although increasing federal contributions through Medicare maximization is a strategy being used effectively by some of the case-study states (including Massachusetts, New York, and Wisconsin), this strategy simply shifts costs to the federal government. In some states, the low Medicaid reimbursement rate (e.g., Alabama for home health) gives providers a logical incentive to bill Medicare rather than Medicaid if at all possible.

System Reform

A second, more ambitious approach calls for states to develop more effective and efficient financing and delivery systems by encouraging developments in the integration of acute and long-term care and expanding home and community-based service programs. Almost all of the selected states see managed long-term care and the integration of acute and long-term care services as a potential way to lower the rate of increase in expenditures, improve quality of care, reduce the number of providers with which state officials must deal, and shift much of the financial risk from the state government to providers. However, most of these efforts are only in the planning stage and are limited in scope. Despite substantial interest among state policymakers, progress on these initiatives has been slow, in part because Medicaid (and often Medicare) waivers are needed for their implementation.

While acknowledging that the current financing and delivery system is fragmented and does not serve persons with disabilities well, long-term care advocates and providers in the states were not always convinced that integration would be good for elderly with disabilities. First, they argued that managed care organizations had little experience or skill with the elderly or long-term care. Second, opponents worried that fiscal pressures within an integrated system could shortchange long-term care if managed care organizations do not view long-term care as a priority or if acute care overruns its budget. Third, long-term care could become overmedicalized and services less consumer-directed if the balance of power shifts from individual clients and their chosen provider to HMOs. Fourth, long-term care providers were also concerned about their relative bargaining strength and expertise in their negotiations with managed care organizations.

While the recent expansion of Medicaid home and community-based care has focused mostly on younger people with disabilities, efforts also are being made to expand home and community-based services for older persons. All of the 13 states express a policy commitment to the expansion of home and community-based long-term care services, although the extent of this commitment varies by state. In all but two states (New York and Texas), only a small percentage of Medicaid funds for the elderly is spent on home and community-based services.

Several case-study states have shifted state-funded home care programs into Medicaid, especially through Medicaid home and community-based services waivers, taking advantage of the flexibility these waivers offer in terms of services and the ability to limit enrollment and expenditures. Although some states complained about the paperwork relating to the waiver, none found that the current system prevented them from doing what they wanted. Some of the recent emphasis on community-based care is based on consumer preferences, but the primary impetus for these reforms is the promise of cost-savings—an outcome about which research has been equivocal at best.¹⁸ To achieve these cost-savings, states will have to be effective in keeping per-person costs down and limiting the increase in utilization that typically occurs when home care services are offered.

Most of the 13 states are increasingly debating the use of nonmedical residential care for the elderly as an alternative to nursing home care. A sizable stock of residential facilities exists in several of these states. These states (and others) face difficult issues as they consider expanding residential care options, most notably how to allow people with substantial disabilities to age in place without making these facilities into substandard nursing homes. In addition, concerned that most assisted-living facilities are expensive and geared to upper-income elderly, states are exploring how to make these new residential options available to moderate- and lower-income elderly.

Traditional Cost-Containment Mechanisms

In the short run, especially if faced with an economic downturn, states are likely to rely on more traditional strategies to reduce spending, such as controlling nursing home supply, cutting reimbursement rates, and tightening eligibility. Many states have responded to growing Medicaid long-term care expenditures by limiting the number of long-term care providers. These efforts have focused largely on nursing home beds, following the general premise that new beds are likely to be filled with Medicaid residents. Six of the 13 states (Colorado, Massachusetts, Minnesota, Mississippi, Texas, and Wisconsin—and, until 1996, Alabama) had moratoriums on new nursing home construction or certification for participation in Medicaid in 1997. While limiting supply will save money over the short to medium term, it does not address the underlying demographics of an aging population.

Because the impact of rate changes on state budgets is predictable, immediate, and potentially large, Medicaid payment rates for nursing facility care are a logical target for states trying to reduce the rate of growth in long-term care expenditures for the elderly. The "Boren amendment," included in the Omnibus Reconciliation Act of 1980, governed how states reimbursed nursing homes, requiring them to pay rates adequate to cover the costs of an economically and efficiently operated facility that met quality and safety standards. All of the 13 study states thought it impossible to operationalize the Boren amendment’s requirements without providing overly generous reimbursement to nursing homes. The Boren amendment was repealed in the Balanced Budget Act of 1997, with the effect that states have much greater legal freedom to impose rate cuts on nursing homes. However, doing so may still be very difficult for the states. In all of the 13 states, the for-profit nursing home industry is one of the most powerful health care interest groups and will resist these cuts. Moreover, to the extent that these cuts are believed to affect nursing home residents adversely, elderly advocacy groups will oppose them as well.

The New Federalism and State Health Policy for the Low-Income Population¹⁹

Perceptions of how much freedom states have to design and operate their Medicaid programs vary greatly. Federal policymakers tend to believe that Medicaid is primarily a federal program over which they have little control because of the enormous amounts of state flexibility. On the other hand, state policymakers believe that Medicaid is essentially a state program over which they have little control because of extensive federal requirements. Whatever the actual balance, federal rules and mandates clearly shifted power toward the national government in the 1980s and early 1990s, and that power has shifted back toward the states since then. Congressional proposals in 1995 and 1996 for a Medicaid block grant have triggered a vigorous debate on what Alice Rivlin called "dividing up the job" in health care.²⁰

Those who favor reduced federal regulation contend that restrictive and uniform national rules do not work well, given the wide variation in economic circumstances and voter preferences across states. As Sparer and Brown note, "America is an extraordinarily heterogeneous society, and Americans have long believed that public policy should, wherever possible, reflect disparate local needs and preferences. This perception is particularly strong in the health field because health care institutions are thought to "belong’ to the community."²¹ Opponents of uniform standards argue that restrictive national rules place federal lawmakers and regulators, who lack thorough understanding of the situation in each state, in the position of "micromanaging" Medicaid. Many have suggested that allowing states broad discretion to experiment with Medicaid might lead to better policy designs by allowing innovations to be tested in a limited number of states without running the risks of implementing an untried initiative on a nationwide basis. Thus, in Justice Louis Brandeis’ famous phrase, states can function as "laboratories of democracy."²²

Almost without exception, high-level state officials interviewed for the case studies wanted more freedom to design and administer the Medicaid program in their states. Not surprisingly, then, officials generally liked the flexibility in the proposed Medicaid block grant, even when (as in Florida) they expressed concerns over the plan’s fiscal adequacy.²³ States complained most about federal rules governing the Early and Periodic Screening, Diagnosis, and Treatment program (EPSDT), hospital and nursing home reimbursement, and managed care. The Balanced Budget Act of 1997 addressed two of these issues, repealing the Boren amendment and allowing states to implement mandatory managed care enrollment without a federal waiver.²⁴ Some states (e.g., Washington and Wisconsin) were also interested in implementing a sliding scale premium or copayment schedule, especially for Medicaid beneficiaries above the federal poverty level, that was inconsistent with federal law or regulation. Few states were looking for flexibility to reduce eligibility, although one conceded that they might not expand Medicaid coverage to older children if not required by federal law to do so. Indeed, if forced to make cuts as a result of a Medicaid block grant, several states (including Minnesota and Alabama) had explicit policies of reducing eligibility only as a last resort.

State administrators often complained that applying for Medicaid home and community-based services, managed care, and research and demonstration waivers was time-consuming and a drain on limited staff. In addition, these policymakers generally believed that the Health Care Financing Administration (HCFA) was slow to respond to waiver requests and made it too difficult for the state to "try things out" or implement locale-specific programs that made sense in one part of the state but would not in another. States generally had little sympathy for HCFA’s contention that federal law limited what they could approve and that "research and demonstration" waivers actually should have a research and demonstration focus.

Proponents of a strong federal role believe that because Washington provides the lion’s share of funds, it is only reasonable to expect federal legislators and administrators to retain a major voice in how the programs are run. Because funds are raised nationally, their spending should reflect "national" interests. At the state level, advocates for the poor and most provider groups opposed block grants because they did not trust their states to "do the right thing" and feared the loss of legal leverage to take the state to court to enforce federal rules.

In recent years, a common argument in favor of minimum national standards in social programs for the poor has been that they are necessary to ward off a downward spiral in eligibility and benefits.²⁵ Without federal minimum standards, the argument is that states will become more concerned about the relative generosity of their programs. A state that provides more benefits than its neighbors may fear that it will attract poor residents from these states. Paying benefits to a larger number of poor people costs money and requires higher taxes, which will make it harder to attract the job-creating businesses and affluent taxpayers that every state wants. To avoid becoming a "welfare magnet," some states may cut their benefits to levels less generous than they feel are appropriate. This could touch off a negative competition in which neighboring states repeatedly cut back their benefits, creating a "race to the bottom."²⁶

Federal involvement in safety net programs can restrain this interstate competition through two mechanisms. First, open-ended federal cost-sharing of the type that characterizes Medicaid and the former AFDC program reduces the costs to a state that has relatively generous benefits, wants to raise its benefits, or experiences an influx of low-income families. This is because at least half, and for most states more than half, of the marginal cost for new residents or expanded benefits is assumed by the federal government. The federal Medicaid matching rate is also more generous for lower- than higher-income states, reducing the "net price" that poorer states must pay for services, thus encouraging them to spend more money then they would if they had to pay the full amount.²⁷ Second, federal standards reduce interstate variation by establishing minimum benefits and eligibility conditions for receiving federal aid. For example, because all states must provide Medicaid to children under six and pregnant women with incomes below 133 percent of the federal poverty level, states cannot compete on whether they will incur the costs of such coverage.

Our case studies revealed very little interstate competition in health care for the low-income population. Although officials are interested in the policy initiatives of other states, "homegrown" solutions are strongly preferred, in part because proposals that are not developed within a state are unlikely to be enacted.²⁸ Moreover, few states consciously look to other states to determine their benefit and eligibility levels (although in Alabama, Medicaid officials did compare their nursing home reimbursement levels to those of other states). Especially in lower-benefit states, such as Alabama and Texas, it simply was inconceivable to state officials or advocates that these states could ever be welfare magnets.²⁹ There was somewhat more concern about being a welfare magnet in states with more extensive benefits and eligibility, but in no state was it a major determinant of health policy. Indeed, in a few states (e.g., Minnesota), policymakers felt pride that their health care system was so good that some people would move to the state to use it.³⁰

One reason a race to the bottom has not occurred in Medicaid is that state officials viewed cash assistance and health care very differently. Although recognized as an unpleasant necessity of life in some states, cash assistance is held in low regard. In contrast, receiving Medicaid is not necessarily thought to be bad. While some policymakers argued that individuals should not necessarily be able to obtain cash benefits, no state official argued that low-income sick people should not receive health care. Reflecting this differentiation, some Section 1115 Medicaid research and demonstration waivers proposed increasing the number of people eligible for Medicaid (e.g., in Massachusetts, Texas, and Wisconsin); and in some states (e.g., Wisconsin) falling Medicaid caseloads have caused considerable concern. In contrast, declining cash welfare enrollment levels have generally been viewed as a positive development.

Medicaid and, to a lesser extent, other health care programs for the poor have broader political support than cash assistance for three reasons. First, morally, most people believe that people have a right to health care regardless of their financial status. This, however, does not mean that they are willing to pay taxes to provide everyone with comprehensive health care. As a result, proposals for Medicaid eligibility expansions and Section 1115 research and demonstration waivers have depended on intergovernmental transfers (e.g., in Texas), managed care savings (e.g., in New York and Massachusetts), or federal Medicaid matches for existing state-funded programs for the uninsured (e.g., in New York and Massachusetts) so that higher general revenue spending by the state would be minimal. The resistance to increased state spending is such that some states, such as Alabama and Texas, may not use all of the federal State Children’s Health Insurance Program (S-CHIP) funds allocated to them because the states have not authorized the required matching funds (even though S-CHIP has a substantially higher federal match rate than Medicaid).

Second, unlike cash welfare, large, well-organized health care provider associations depend (to a greater or lesser extent) on Medicaid for their financial survival and fight to support it. With more than two-thirds of residents dependent on Medicaid, nursing homes are particularly active in supporting the program.³¹ Third, and closely related, the inclusion of long-term care for the elderly (with its more middle-class beneficiaries) in Medicaid engenders much more political support for the program than would exist if it were solely an acute care program for low-income children and nonelderly adults.³²

In sum, the case studies showed little evidence of a "race to the bottom" in health care based on interstate competition for business. However, neither was there evidence of a "race to the top" in which states increase spending in order to be a leader in providing benefits to the poor. Nor has the current excellent fiscal condition of most states been an opportunity to greatly expand eligibility and coverage under Medicaid.³³ Instead of being a function of interstate competition, state decisions on health care seem to be the function of the political culture, values, and fiscal situation of individual states.

Summary

This paper has addressed a wide range of state health policy issues affecting the low-income population, including third-party coverage, financing and delivery, safety net providers, long-term care, and federal-state relations. Of the many important themes, eight findings dominate the study’s portrait of state health policy.

First, while states historically have had primary responsibility for insurance regulation, most states have had to make major changes to strengthen their regulation of the small-group and individual health insurance markets in order to comply with the Health Insurance Portability and Accountability Act of 1996. The passage of new federal insurance standards is an exception to the devolution of authority to the states that has characterized recent federal policy in health care.

Second, public coverage through Medicaid and other programs significantly lowers the uninsurance rate for those under 200 percent of the FPL, especially children, but the level of employer-sponsored coverage is the main determinant of the uninsurance rate for the overall population. For the 13 selected states, employer-sponsored health insurance coverage ranged from 75 percent or more in Michigan, Minnesota, and Wisconsin to about 57 percent in California, Texas, and Mississippi. States with the highest rates of employer-sponsored coverage had uninsurance rates of about 10 percent, while states with the lowest rates of employer-sponsored coverage had uninsurance rates of 20 percent or higher. A high rate of uninsurance does not necessarily indicate a lack of state effort to expand coverage through Medicaid or other public programs or a lack of insurance reform. Rather, it reflects the fact that state governments face much larger problems in financing care for the uninsured where employer-sponsored coverage is low (and vice versa).

Third, despite state efforts to maintain Medicaid enrollment for those beneficiaries leaving the cash welfare rolls without obtaining employer-sponsored health insurance, Medicaid enrollment is falling. Yet Medicaid enrollment has not declined as much as cash welfare rolls because of the many ways to remain eligible for Medicaid, especially for children and pregnant women. Moreover, S-CHIP will increase coverage and thus offset some of the decline in coverage of children resulting from welfare reform. However, rates of uninsurance for adults, especially women, are likely to increase. These changes imply that the composition of the uninsured will change, with fewer children and more adults. Falling enrollment will depress the rate of growth in Medicaid expenditures.

Fourth, the Medicaid managed care revolution has been more of a skirmish than a revolution. The goals of Medicaid managed care were to expand access to mainstream providers and to save money, but success on both fronts has been limited. Medicaid managed care is predominantly limited to children and younger adults; few states have extended enrollment to more expensive elderly and disabled enrollees, limiting potential savings. States are also finding that managed care savings are modest because traditionally low Medicaid fee-for-service payment rates make it difficult for states to substantially slash capitation levels or for HMOs to negotiate further price discounts. In addition, safety net providers that need Medicaid revenues to survive have received special protections from the states, which has both reduced potential savings and steered Medicaid beneficiaries to traditional providers of charity care. The combination of low capitation rates and protections for safety net providers have limited the willingness of commercial HMOs in several states to contract with states, thus restricting the expansion of access to mainstream providers.

Fifth, state administrators have been extremely adept in maximizing federal funds under Medicaid, which is reflected in the explosion of disproportionate share hospital payments between 1988 and 1992, and in the Medicaid coverage of various health and social services that previously were financed solely with state and local funds. However, recent policy changes at the federal and state level are at odds with this strategy. The Balanced Budget Act of 1997 substantially reduced federal DSH spending, reflecting a desire for federal Medicaid savings and the belief that many states have abused the DSH program. As a result, states that used DSH payments to reduce the state share of Medicaid expenditures will have to spend more of their own money or find a way to cut costs. Furthermore, safety net providers that have received a substantial increase in revenues through the DSH program will have to find additional funding from state and local governments at a time when the number of uninsured is increasing.

In addition, the shift of state- and locally funded health and social services into Medicaid is complicated by Medicaid’s movement to managed care. In many cases, these services are included in the capitation rate, but HMOs either do not contract with government providers or do not authorize the services to the extent that was previously the case in the fee-for-service system. Consequently, state and local governments must find ways to ensure that managed care organizations use government providers, finance services wholly with their own revenues, or reduce service levels.

Sixth, local safety net providers—public hospitals, community health centers, and some voluntary hospitals—are experiencing financial stress, but there have been no widespread failures as some feared. The development of several problems at the same time, such as increased hospital competition, Medicaid managed care, growth in the number of uninsured, and reductions in DSH payments, could create serious problems in many cities.

So far, safety net providers have successfully responded to these problems in several ways. To begin with, few providers face situations in which everything is working against them. For example, some states had aggressive Medicaid managed care but generous DSH payments. In addition, safety net providers have not sat idly by while the world collapsed around them. Rather, they have responded to these pressures by making organizational changes, becoming more efficient and customer oriented, and joining or creating managed care plans. Moreover, when safety net providers have faced certain collapse, state and local governments have intervened (very occasionally with the help of the federal government) to make sure that critical services continued. More generally, some level of government inevitably will have to provide direct subsidies to the safety net to replace the indirect subsidies that are evaporating. Historically, responsibility for these subsidies fell to local governments, but their ability and willingness to support safety net providers is not clear.

Seventh, while long-term care is often downplayed at the national level, it is much too large a part of Medicaid spending (34 percent) to be ignored. Given the aging of the population, the potential for spending growth is enormous. Long-term care spending varies greatly among states, much more than acute care does. For example, in 1996, New York spent $2,643 per elderly resident while Florida spent only $340 per elderly resident.

State policy initiatives include promoting private long-term care insurance and reducing the transfer of assets to artificially qualify for Medicaid nursing home benefits, but these efforts are not providing major savings to the states. Another approach is improving the efficiency of the system, primarily through expanding managed care to include nursing home and home care and increasing the amount of home and community-based services. In particular, states have sought to change the balance between institutional and noninstitutional care for the elderly. While the policy rhetoric has favored home and community-based services, only 10 percent of Medicaid long-term care spending for the elderly is for noninstitutional services.

Limitations in the effectiveness of other strategies means that traditional cost-containment methods, such as controlling Medicaid payment rates and limiting the supply of nursing home beds, remain the most potent tools available to state policymakers. The repeal, by the Balanced Budget Act, of federal rules regarding Medicaid nursing home payment rates raises questions of how states will respond to their new flexibility and what impact that will have on quality of, and access to, nursing home care.

Finally, development of state health policy for the low-income population takes place within the confines of a financial and regulatory relationship between the states and federal government, principally through the Medicaid program. States have long sought additional flexibility to run the Medicaid program as they choose, especially in the areas of managed care and nursing home and hospital reimbursement. States would also like additional freedom to "try things out" without having to obtain federal permission to do so and without having to construct a "research" rationale.

In recent years, opponents of additional state flexibility have argued that greatly reducing federal rules could result in a "race to the bottom" as states compete to attract businesses with lower tax rates, which may necessitate minimizing benefits to the low-income population. The contention is that, without federal minimum standards, state fears of becoming "welfare magnets" will result in benefit levels below what states might provide if it were not for the interstate competition. These case studies found little evidence of this type of interstate competition, at least as it relates to health care. Instead, Medicaid coverage and benefits are determined mostly by intrastate political culture and availability of tax revenues. Unlike cash assistance welfare, health care is viewed as a positive good that everybody should receive. Although neither the money nor the mechanism to implement that conception is available in any state, declining Medicaid caseloads are matters of concern for state officials while falling cash welfare caseloads are matters of pride. The race to the bottom is further impeded by large, well-organized, well-financed provider interest groups that depend on Medicaid for their survival and lobby to make sure that the program is adequately financed. Finally, Medicaid long-term care services benefit more than the poor, giving the middle class a reason to support the program. In many cases, the elderly parents of middle-class adults are the ones who receive Medicaid nursing home benefits.

But if there is no "race to the bottom," then the rationale for federal rules must be that some states will not do the "right thing" unless forced to do so by the national government. The key problem for policymakers is to decide when and where state variation and accommodation of local conditions makes sense and where there is one standard way of doing things with which all states should comply.

Notes

Tables

Table 1. Characteristics of States Selected for Intensive Study

Table 2. Medicaid Expenditures in ANF States, 1996

Table 3. Medicaid Enrollees and Expenditures per Entrollee in ANF States, 1996

Table 4. Medicaid Benefit Expenditures in ANF States, 1996

Table 5. Health Insurance Coverage of Nonelderly under 200 Percent of FPL, 1994-95, by Scope of Publicly Supported Health Insurance Programs (Percent Distribution)

Table 6. Health Insurance Coverage of Nonelderly Adults and Children under 200 Percent of FPL, 1994-95, by Scope of Publicly Supported Health Insurance Programs (Percent Distribution)

Table 7. Recent Changes in Participation Levels for Medicaid and AFDC/TANF

Table 8. Medicaid Disproportionate Share Hospital (DSH) Payments in ANF States, 1996

Table 9. Factors Threatening Safety Net Systems

Table 10. Medicaid Long-Term Care (LTC) Expenditures for the Elderly by State, 1996

About the Authors

John Holahan is the director of the Urban Institute’s Health Policy Center. He has authored many publications on the Medicaid program. He has also published research on the effect of expanding Medicaid on the number of uninsured and the cost to federal and state governments. Other research interests include expanding health insurance for children, health system reform, changes in health insurance coverage, physician payment, and hospital cost containment.

Joshua Wiener is a principal research associate in the Urban Institute’s Health Policy Center, where he specializes in research on Medicaid, long-term care, and health care for the elderly. He is the coauthor or editor of seven books and more than 70 articles. Before joining the Urban Institute, Dr. Wiener did health care research and policy analysis for the Brookings Institution, the Health Care Financing Administration, the Massachusetts Department of Public Health, and the New York City Department of Health.

Susan Wallin is a research associate in the Urban Institute’s Health Policy Center. Her research has centered on access to care for low-income populations, including issues of health professions maldistribution, Medicaid managed care, and public health departments. Previously, she served as an analyst for the Physician Payment Review Commission.

The authors would like to thank the many state and local officials and others who participated in the interviews and provided the information contained in this report. We also acknowledge the Urban Institute and Child Trends staff who conducted the site visits and prepared the background papers that served as the source for the state- specific information. We are particularly grateful for the many contributions of Sandra Clark.

Usage and reprints: Most publications may be downloaded free of charge from the web site and may be used and copies made for research, academic, policy or other non-commercial purposes. Proper attribution is required. Posting UI research papers on other websites is permitted subject to prior approval from the Urban Institute—contact [email protected].

If you are unable to access or print the PDF document please contact us or call the Publications Office at (202) 261-5687.

Disclaimer: 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. Copyright of the written materials contained within the Urban Institute website is owned or controlled by the Urban Institute.