If you read the newspapers or professional journals, listen to political speeches or the “man on the street,” a consensus with seeming contradictions becomes apparent: Very few people think Medicaid works well—it costs too much, it does not buy good care, it is out of control. Yet there is a strong base of support among the public and within government and academic sectors to maintain and enhance access to health care, particularly for the poor.The essential design of Medicaid has remained largely unchanged through its 25-year history despite criticism from all quarters. Medicaid is an important feature of the American social welfare system—a system that has survived numerous economic and political cycles. In part, this survival reflects the preference of our political culture for incremental change, a preference that has rescued Medicaid and other safety net programs from dismantling. We are regularly confronted with urgent crises that forestall consideration of long-range problems and solutions. However, the failure to take on broad, systemic issues has also stymied legitimate efforts to reform or create anew the program.Substantive policy changes have occurred—from the retrenchment of budget reconciliation measures in 1981 to the expansions of the late eighties. The best and worst properties of Medicaid have been highlighted through selective, innovative measures pursued at both Federal and State levels in the past decade, from the unrealized potential of cost containment through recent efforts focused on infant mortality, child health, and acquired immunodeficiency syndrome.Today it is clear that these changes are not sufficient. The pent-up forces of new demands and growing costs have overrun our problem-focused, gradual approaches. The basic structural flaws relating to eligibility, benefits, and payment policy, which have characterized Medicaid from its inception, largely remain. States react to emerging crises and new requirements with little structure to reward positive performance or support long-term goals; many opportunities for innovation are thus thwarted.To remain viable through the nineties, Medicaid will require major changes. Either the program's distinct identity should be retained—with significant expansions and modifications—or it should be subsumed within a broader national health financing strategy. An evaluation of these options should consider Medicaid's historical purpose and its current realities, the consequences of (and alternatives to) budgetary controls inherent in Medicaid's structure, and how Medicaid or its successor could better serve the health care needs of the poor in the context of the overall health financing system.
Unrealized intention, unpredicted results
Medicaid was intended to improve health care access for the poor. It has yielded substantial benefits. There were approximately 23.5 million Medicaid recipients in 1989, about 16 million adult family heads and dependent children, and 8 million aged, blind, and disabled. (Health Care Financing Administration, 1990). Those who have participated in the program have certainly benefited, as demonstrated by dramatic increases in utilization and improved health status in the period immediately following Medicaid's enactment, particularly when compared with the poor without coverage (Rogers, Blendon, and Moloney, 1982).However, Medicaid was never designed to serve everyone who is poor, nor to meet all health care needs of those eligible. Generally, Medicaid eligibility is linked—and limited—to the categorical groups and income levels defined for the cash assistance programs; with limited exceptions, the standard of “poverty” for Medicaid is not the Federal poverty level. And, beyond a core set of mandatory, essential services, each State has, and has exercised, considerable discretion in offering additional types of health services to its Medicaid-eligible population.The subset of the poor who are eligible for public assistance is composed primarily of children, the elderly, the disabled, and pregnant women or women who are single parents. Recent expansions have also targeted these groups, to the exclusion of other segments of the poor (e.g., single and married adults without dependent children). This reflects the historical emphasis of social welfare programs, but it does not satisfy contemporary perceptions of the scope of unmet health care access needs.
New initiatives, old structure
Recent incremental expansions in Medicaid address some of these above-mentioned problems. The expansions of eligibility for pregnant women, children, and the elderly adopted during 1986-90 expressly severed the link between public assistance and Medicaid by mandating an income standard at or above the poverty level. Recent statutory reforms also emphasized Medicaid's value as a work incentive, by mandating extended, transitional coverage when the AFDC cash assistance income threshold is exceeded as a result of earnings from employment.However, these innovations were effected within the existing structure. Congress did not secure a broad health care entitlement for the poor outside the welfare system— rather, it made adjustments at the margins of that system. Neither did these measures reallocate or alter Medicaid's responsibilities for other categories of eligibility and services.Moreover, the mandated changes were layered onto the existing Federal-State administrative and fiscal structure. Thus, while Federal matching funds were applied, the new provisions have also added $2.5 billion to States' Medicaid budgets in the past 3 years (National Association of State Budget Officers, 1989). These costs, and the added administrative requirements associated with the new mandates, were also superimposed onto the pre-existing factors driving Medicaid cost growth, primarily long-term care services.The elderly were certainly intended to benefit from Medicaid's enactment. However, the current proportion of Medicaid costs now associated with expenditures for the elderly and the physically or mentally disabled was not acknowledged at this inception. These categories of expenditure (particularly for the disabled) comprise the majority of cost growth since 1975 excluding the new mandates described previously and in the article in this issue by Thomas Reilly, Steven Clauser, and David Baugh. Although the elderly and disabled categories account for only 28 percent of all Medicaid recipients, they account for 73.5 percent of total expenditures (Ruther and Reilly, 1989).
Policy implications of budgetary control
Given the inexorable growth in services and dollars under Medicaid, a variety of cost-control strategies have been initiated at both Federal and State levels. Public and private sector efforts to encourage more rational, efficient utilization of services have not yielded big savings thus far. In fact, most of Medicaid's control and reduction strategies have targeted the groups that represent the least cost to the system.Throughout the eighties, States have tested administrative reforms, utilization controls, and managed-care arrangements in order to promote greater efficiency and cost savings. Many of these problems were modeled after similar innovations in the private sector.The managed-care experiments and demonstration projects have focused primarily on enrollment of non-elderly Medicaid recipients in health maintenance organizations (HMOs) and other managed-care arrangements. Subsequent evaluation has demonstrated that costs could be controlled and quality maintained or improved, but little net savings may result from these managed-care programs. Reduction of emergency room and inpatient use has been offset by more appropriate, less costly increased primary care in many instances. But it has been difficult to launch and sustain managed care under Medicaid: Program design has been complicated and time-consuming, and administrative costs are higher, at least in the initial stages (Freund et. al., 1989; Spitz and Abramson, 1987). The Federal waiver process has been cumbersome for many States. Finally, many conventional HMOs have been reluctant to enter into Medicaid managed-care arrangements. Thus, although managed care under Medicaid has met with some success, the evidence suggests it remains a marginal intervention in most States.The principal means of fiscal restraint under Medicaid is cost avoidance through restrictive eligibility, benefits, and payment policies. The lack of uniform Federal standards beyond the minimum prescribed levels permits States to exercise broad discretion over who and what they cover. This flexibility has been used to control the fiscal dynamics of State Medicaid programs. Given the link to public assistance, Medicaid is also hostage to the budget politics of each State's welfare programs. Recent Federal mandates have curtailed States' flexibility in some areas. Nevertheless, the Medicaid patchwork is still defined primarily by annual State budget and legislative battles, or imposed on States without complementary changes in regulation of the broader health care financing structure.
What Medicaid is, is not
The complexity of the Medicaid program often defies rational discussion and choices. The fact is that Medicaid is not one program, but an umbrella name for numerous unconnected pieces. This presents an enormous challenge in the critical areas of strategic policymaking and budgeting as well as program and fiscal management.Another consequence is that very few people really know the program as a whole. Neither administrators, elected officials, or the public has enough coherent information on which to base rational decisions. In this environment, it is easy to make sweeping generalizations or to rely on symbolic rhetoric.A critical distinction that may be obvious nevertheless merits emphasis—Medicaid is not health insurance. Eligibility, benefit, and payment variations among States, and from 1 year to the next, as well as Medicaid's continued link to the welfare system, defy any meaningful comparison to the employment-based or Medicare systems. This does not mean that insurance approaches are not viable substitutes for portions of who and what is covered under Medicaid. Rather, Medicaid in its current form cannot be expected to operate the way insurance does. Policymakers and the public must frankly evaluate whether marginal reforms that borrow selected characteristics from insurance are likely to be appropriate or effective.Finally, it is tempting to think that innovative measures implemented through Medicaid will have a demonstrable impact. Every waiver, demonstration, and experiment was expected to yield results that would translate into significant systemic change. More often than not, we have been disappointed and perplexed when such change does not happen. Our expectations have been disproportionate to the actual scale and integration of reforms that can realistically be attempted within Medicaid or in the health care system generally. The reasons are varied, but three major factors can be highlighted.First, Medicaid is often assumed to be the salve for multiple social ills, many of which have origins and require interventions outside the health care system. Burgeoning long-term care demands, for example, signify a much broader failure of social supports—Medicaid acts as a financial buffer after the fact. More recently, the phenomenon of infants exposed to drugs in utero signals a devastating deterioration of the social fabric in certain communities. Medicaid financing alone will be wholly inadequate to meet the needs of these children, yet the lack of alternatives has shifted enormous responsibilities onto the health care sector.For the non-elderly population, Medicaid is simply not comprehensive enough to substantially influence the broader health care environment. States that have achieved purchasing leverage through their Medicaid programs have generally tied in other major payers, for example, through all-payer hospital reimbursement systems. Managed-care programs designed specifically for the Medicaid population have also not had an impact on the system as a whole, in part because HMOs have limited their participation.Finally, there are few programmatic or fiscal alternatives to substantially change the amount and distribution of institutional care expenditures for the elderly and the physically or mentally disabled. Medicaid is but one factor in a broad array of resources that must be coordinated and channeled to effectively develop wide-scale community-based alternatives. If successful, these solutions will not necessarily be less costly than current institutional arrangements.
Future questions and strategies
For all of its shortcomings and mythology, Medicaid has withstood the test of time and remains an essential part of both the social welfare and health care financing systems. Recent Federal expansions reinforce the importance of Medicaid in securing health care access for many living in or near poverty.But the limitations of Medicaid and the encompassing health care system have moved to the front of Federal and State policy agendas. The inevitable questions are what role Medicaid should have as part of a broader national health care plan, and how best to target improvements in that context.A comprehensive reform of the Medicaid program was recommended by an ad hoc committee, which I chaired, of the Health Policy Agenda for the American People. The committee cited wide variations in eligibility, benefits, and payment levels as the principal deficiency in Medicaid. The link to public assistance eligibility and stringent asset tests also create substantial and unjustified barriers to Medicaid access.Based on this critique, the committee recommended that national standards and goals govern the Medicaid program. Eligibility levels would be set at no less than the poverty level nationally. States would be mandated to establish a medically needy program with spend-down provisions; asset testing would be permitted only for this component. A standard benefit package would be mandated, including a basic, comprehensive primary and acute care schedule of benefits. Finally, Medicaid expansion would have to be coupled with delivery system improvements. These would include measures to promote cost effectiveness (e.g., managed care) and incentives for increased provider participation (Health Policy Agenda for the American People, 1989).An alternative, proposed in other recent reports, would be establishment of a new program for the poor and others outside the mainstream health care system. For example, the widely reported proposals of the National Leadership Commission (National Leadership Commission on Health Care, 1989) and, most recently, the Pepper Commission (U.S. Bipartisan Commission on Comprehensive Health Care, 1990) recommend a fundamental restructuring of Medicaid as part of broader reforms. Although differing in their detailed prescriptions, each of these proposals addresses problems central to Medicaid's design: Coverage for the poor must be integrated with, not separate from, the financing system for those covered through employer-based plans; and benefits and payment policies must be reasonably uniform throughout the system.The importance of integrating coverage strategies involving Medicaid and the employer-based system is starkly illustrated by a recent fiscal analysis of proposed Medicaid expansions and various congressional proposals to extend coverage beyond Medicaid.If Medicaid coverage were severed from public assistance eligibility and extended to all uninsured persons with incomes up to the poverty level, 13 million people would be directly affected. The gross public sector cost for this expansion would be approximately $13.5 billion.In the absence of private sector coverage expansion, 3.7 to 7.4 million additional poor people with group or individually purchased coverage could shift over to Medicaid. This would result in additional Medicaid costs of $4.5 to $9.7 billion, for a total additional public sector cost of $23.2 billion.However, if Medicaid expansion for the poor uninsured were adopted along with a requirement for employers to cover all full-time employees and their dependents, only 7.2 million people would become Medicaid eligible. The public sector cost would be reduced to $9.9 billion, less than one-half the amount for a Medicaid-only strategy (Thorpe and Siegel, 1989).This analysis highlights some consequences and interactions of various reform strategies. However, a more fundamental choice faces policymakers and their multiple constituencies: Should coverage for the poor remain separate from the larger systems of employment-based coverage available to most working families and many retirees and from Medicare coverage for the elderly?A separate Medicaid system operates according to a unique set of rules for providers and recipients. The fiscal pressures on States will continue to mount as Medicaid expenditures impinge upon other high-priority social investments. State initiatives to improve Medicaid will be limited, with expansions increasingly dictated at the Federal level. These will, in turn, be limited by the mounting Federal budget deficit.Cost-containment innovations will target reductions or reallocations built into the existing system. For example, Oregon has enacted legislation that would deliberately limit Medicaid expenditures by prioritizing services: Benefits will be curtailed and managed care will be the rule rather than the exception.Oregon's explicit choices are mirrored by thousands of implicit choices at the Federal, State, and local levels that will shape Medicaid under the existing structure. The poor who rely on Medicaid for their health care coverage will remain part of a separate and unequal system.Demographic trends and technological advances suggest that health care costs for the elderly and disabled will continue to escalate. Under the present configuration, long-term care financing is part of Medicaid's means-tested safety net. States' choices under Medicaid will thus be further limited to the extent that long-term care claims a growing proportion of expenditures.Conversely, long-term care financing could be integrated with the broader social insurance protection afforded by Medicare and income security programs that serve most long-term care users. However, this reallocation would certainly accelerate Medicare's impending financial crisis and the need for reforms in the coming decade.
Conclusion
Each of these policy choices suggests very different national strategies for health and social welfare financing. Fiscal realities in both public and private sectors will require a gradual response—the question is whether the increments of that response will be part of an overall design that is comprehensive and national in scope.Yet within this enormous challenge lies an equally important opportunity. Unifying the public and private sector financing responsibilities would permit the development of broad-based coverage and purchasing strategies. Resource allocation would be debated and decided as appropriate for the population as a whole. A strategy that integrates the various elements of health care financing will reinforce our substantial commitment to the common good and promote an affirmative American social ethic governing health care access for all (Reinhardt, 1985).