Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
3 Eliminating Uninsurance: Lessons from the Past and Present Despite nearly a century of efforts and incremental reforms to extend cover- age, the nationâs multiple sources of coverage leave 15 percent of the total popu- lation uninsured. This chapter develops the historical context of national reform efforts to reduce or eliminate uninsurance in the United States. It then looks at relatively recent federal initiatives to broaden coverage substantially and extension of coverage by some states and counties that have taken leadership roles. Past efforts offer useful lessons for reforming health care financing today. Federal, state, and county reforms have not eliminated uninsurance, although some initiatives have improved access to health services or resulted in better health outcomes for populations who had lacked coverage. Some reforms have affected the basic structure of health care finance, while others have had a more limited focus, building on existing public programs or private insurance. The Committeeâs principles for assessing coverage proposals derive from the historical record as well as from its examination of the consequences of uninsurance. NATIONAL EFFORTS TO BROADEN COVERAGE, 1916â1984 The lack of universal health insurance in the United States is in part a legacy of early twentieth-century precedents in the organization and financing of health services in the United States. It also reflects the absence of political leadership strong, broad, and sustained enough to forge a consensus in favor of universal coverage, despite public support, in the face of opposition from overlapping yet at times incompatible economic interests forged within the constraints of American political institutions and processes (Oberlander, 2003). Our governmentâs federal 66
LESSONS FROM THE PAST AND PRESENT 67 structure and the independence of the legislative and executive branches place a relatively great burden on proponents of change. Coverage reform first became a national issue in the early 1900s, when relatively few people had health insurance and most health care was purchased out of pocket or provided charitably. Over the next 70 years, a series of campaigns attempted to bring about greater coverage nationally. Early campaigns to create mandatory coverage were opposed by the medical profession, commercial insur- ers, and the business community. From the 1930s through the 1970s, the House Ways and Means Committee of the U.S. Congress determined the fate of most federal legislation to extend coverage, and most proposed reforms were prevented from coming before the full House for a vote. In the context of the lack of federal legislation for more widespread public coverage, consumer demand fueled the rapid growth of private-sector coverage, starting in the mid-1930s. The nonprofit Blue Cross and Blue Shield plans, and subsequent plans from commercial insurers, enrolled millions of subscribers and by the early 1960s most Americans were insured through employment-based cover- age. The enactment of the Medicare and Medicaid amendments to the Social Security Act in 1965 filled some of the gaps left by the emerging employment- based approach to financing care. Medicare extended coverage to most of the population over age 65 as well as to smaller groups of eligible persons (the perma- nently disabled). State implementation of the Medicaid program increased cover- age significantly among categories of the low-income population. By the 1970s, the growth in total health care spending facilitated by the creation of Medicare made the issue of controlling health care spending and inflation central to universal coverage reform proposals. The early 1980s marked a high point for coverage levels nationally. Since that time, there has been a gradual increase in the number, and in most years the proportion, of uninsured persons under age 65. See Box 3.1 for a timeline of these efforts. The sections that follow briefly review this history in order to illustrate important issues and basic tensions in the political sphere that have shaped more recent reform efforts. It is organized roughly chronologically, with diversions from the timeline to allow focus on specific topics. Early Efforts: From Protecting the Income of Industrial Workers to Social Insurance for Improving Access to Care Social insurance programs in late nineteenth-century Europe (for example, Germany, 1883) and Great Britain (1911) that included health insurance, and experience with the limited prepaid medical services available to fraternal or mutual benefit society members, spurred interest in universal coverage in the United States (Numbers, 1978). Initial organized efforts to extend coverage broadly occurred during the years around World War I (Starr, 1982). Early twentieth- century America was in the throes of rapid industrial and urban growth, with a booming population of low-income working families. With an eye toward allevi-
68 INSURING AMERICAâS HEALTH BOX 3.1 Landmarks in the History of Coverage 1916â1920 American Association for Labor Legislation campaigns for publicly administered, private-sector sickness insurance to protect the lost income of workers and their families 1932 Committee on the Costs of Medical Care final report calls for group organization and payment of health services on voluntary basis 1935 Social Security Act 1939 First of series of Wagner-Murray-Dingell bills in U.S. Congress that propose universal health insurance as social insurance 1942 War Labor Board ruling permits employers to exclude employment- based coverage from taxable income 1945 President Trumanâs âFair Dealâ proposal includes publicly financed and administered universal coverage 1948 U.S. Supreme Court ruling (Inland Steel) permits collective bar- gaining for employment-based coverage 1960 Social Security Act Amendments including Kerr-Mills program cre- ating federal grants to the states to finance health services for poor persons at least 65 years of age (seniors) 1965 Social Security Act Amendments that create Medicare and Medic- aid, nearly universal publicly and privately financed coverage for seniors and federally guaranteed eligibility for public coverage for specific categories of the poor 1971â1974 President Nixon and U.S. Congress introduce and debate propos- als for universal coverage through a mix of public and private sourc- es, fail to reach a vote in Congress ating the economic burden of illness, a relatively small group of elite reformers, organized as the American Association for Labor Legislation (AALL), campaigned for mandatory workplace-based âsickness insurance.â Sickness insurance, often called health insurance after the English precedent, was modeled after recently created state workmenâs compensation programs (Numbers, 1978).1 It targeted 1For example, former President Theodore Roosevelt made sickness insurance one of the planks in his ultimately unsuccessful campaign against Woodrow Wilson in 1912 on the Progressive ticket (Numbers, 1985).
LESSONS FROM THE PAST AND PRESENT 69 1974 Employee Retirement Income Security Act, prohibiting state regu- lation of self-insured employer health plans. 1981 President Reaganâs proposals to turn Medicaid into block grants to the states, in the Omnibus Budget Reconciliation Act of 1981 1984â1990 Expansions of Medicaid to pregnant women, infants, and children at higher income levels and delinking from eligibility for (state) in- come support 1985 Consolidated Omnibus Budget Reconciliation Act of 1985, with pro- vision to improve continuity of employment-based coverage. 1993â1994 President Clintonâs Health Security Act proposal for universal cov- erage through publicly administered, publicly and privately financed regional purchasing alliances created and debated, fails to reach a vote in Congress 1996 Federal welfare reform delinks Medicaid from income support pro- grams, bars legal immigrants from Medicaid eligibility for first five years of residency 1996 Health Insurance Portabilitiy and Accountability Act, with provisions to improve continuity of coverage 1997 State Childrenâs Health Insurance Program, extending public insur- ance eligibility to children in families earning between 100 and 200 percent of federal poverty level, expansions to their parents 2002 Trade Act, with provision for health insurance premium tax credit for groups of workers displaced by international commerce and retirees industrial workers and their families and included a cash benefit to replace lost income, access to free health care, and a death benefit; premiums would be paid by employers, workers, and the state (Starr, 1982; Hoffman, 2001). The plan was not universal, for it excluded most African Americans and other ethnic minorities by not including agricultural, domestic, and temporary workers. In addition, it re- stricted eligibility to employed men and women earning less than $1,200 annually (low income but not poor), assuming that the poor could rely on charity and that the middle class could afford to purchase their own care (Hoffman, 2001). Between 1916 and 1920, the AALLâs model sickness insurance bill was de-
70 INSURING AMERICAâS HEALTH bated in a number of state legislatures in more urban and industrial parts of the United States, most notably California and New York, with support from many governors and state-level fact-finding commissions (Numbers, 1978; Hoffman, 2001). Ultimately, these bills were defeated. Key opponents included: â¢ the American Medical Association (AMA), an early supporter that became an outspoken opponent as rank-and-file members (county medical societies), gaining experience with contract practice under the new workmenâs compensa- tion programs, grew concerned that insurance would interfere with the practice of medicine and threaten the economic viability of solo practice; â¢ employers who did not welcome broadened financial responsibility for the health of their workforce and who predicted higher costs passed along to consum- ers and the loss of jobs; â¢ commercial insurers who stood to lose their lucrative business selling death benefits (industrial insurance) to low-income workers and who were excluded from the proposed state-level administration of sickness insurance; and â¢ organized labor, principally Samuel Gompers of the American Federation of Labor, to whom mandatory coverage was a threat to the organizing ability of unions (Numbers, 1978; Starr, 1982; Hoffman, 2001). Proposals for state sickness insurance programs were abandoned by 1920, felled by the lack of political leadership, the difficulties of enacting mandatory policies on a state-by-state basis, and the AALLâs inability to build coalitions and fashion workable political compromises with the economic interests that opposed its model for expanding health insur- ance (Numbers, 1985). In addition, a harmful legacy of this first political battle was the framing of mandated coverage as counter to American values, with coverage proposals attacked in newspapers and speeches as fundamentally Germanic, and, after the close of World War One and the onset of the Red Scare, as expressions of Bolshevism (Hoffman, 2001). Starting in the 1920s, demand for health care services by the middle class, fed by the improved effectiveness of hospital-based care and the increasing risk of high-cost medical expenses, reinvigorated public debate about extending coverage (Starr, 1982; Stevens, 1989). Unlike the sickness insurance proposals of the 1910s, the goals of reform were to increase utilization, and health care spending, by making care and coverage more affordable to all Americans universally, not only low-income workers and their dependents (Starr, 1982; Derickson, 2002). During the 1930s and into the 1940s, proposals to extend public coverage significantly took the form of social insurance for all Americans, resembling the system of federal pensions created by the Social Security Act of 1935. Proposed reforms drew on the work of an independent group of scholars and physicians, the Committee on the Costs of Medical Care, that advocated both group practice and financing for care, to rationalize health services delivery and make health care
LESSONS FROM THE PAST AND PRESENT 71 more affordable (Starr, 1982). Because much domestic policy making for the nation now took place at the federal rather than the state level, health finance reformers faced a number of new challenges, including the separation of powers (so that Presidential support would not necessarily be matched by congressional support); a lack of party discipline within the two houses of Congress; powerful regional voting blocks and interest group lobbies; and the ideological constraints of single-party reform efforts (Marmor, 1973; Maioni, 1998). In the 1930s, organized labor supported a publicly mandated extension of coverage, but the medical profession, insurance industry, and business interests continued to resist such proposals. The AMA in particular was a strong political presence in Washington, and its vigorous lobbying of Congress and the public in favor of voluntary rather than mandatory coverage threatened the passage of New Deal legislation, and the Presidentâs bid for reelection. In response, President Franklin Rooseveltâs Administration cut a provision for health insurance from what became the Social Security Act of 1935 (Marmor, 1973; Starr, 1982; Maioni, 1998). In an effort to revive universal coverage proposals, which many perceived as the âmissing pieceâ of the New Deal, in 1939 reform-minded members of Congress began introducing universal coverage bills each year (labeled Wagner- Murray-Dingell bills, after their key sponsors) that framed coverage as a means to lower financial barriers, made eligibility universal (not only for persons in the Social Security system), and included grants-in-aid to the states to support indigent care (Maioni, 1998). None of these bills was reported out of the Ways and Means Committee, undone by the lack of leadership by President Roosevelt and the slim voting majority of the Democrats in Congress that was insufficient to pass contro- versial reform (Marmor, 1973; Maioni, 1998). Universal coverage bills stalled in Congress, but consumer demand for health insurance grew. While commercial insurers were slow to enter the market for group policies organized through the workplace, nonprofit and independent orga- nizations created prepayment plans for hospital services, indemnity and service benefit plans for physician care, and sites for the direct delivery of services that gave fundamental shape to the organization and financing of health services in subsequent decades. The locally organized and directed Blue Cross hospitalization plans and Blue Shield physician plans expanded rapidly from their origins in the early 1930s to become the single largest source of coverage by the 1950s (Cunningham and Cunningham, 1997). Community organizations such as Group Health Cooperative of Puget Sound, consumer cooperatives and private clinics, and health plans organized around industries, such as Kaiser Permanente, extended both coverage and services (Somers and Somers, 1961; Starr, 1982). Enrollment of workers and their dependents in private coverage soared: between 1940 and 1960, the proportion of the general population with private coverage grew from 9.1 percent (about 12 million people) to 67.8 percent (about 122 million people) (Etheredge, 1990; Bovbjerg et al., 1993). Offering health insurance as a benefit was attractive to employers, given the exemption from federal taxes of the em-
72 INSURING AMERICAâS HEALTH ployersâ contributions to health insurance premiums (codified in the Internal Review Code of 1954) (Starr, 1982).2 Although private-sector coverage grew quickly, there was continued public and legislative interest in universal health insurance. After World War II, the federal governmentâs role in many aspects of health policy expanded, for example, in the funding of biomedical research and hospital construction, but congressional resistance, bolstered by the lobbying of reform opponents, blocked public man- dates for coverage (Marmor, 1973; Fox, 1986). The first president to champion universal coverage, Harry Truman revisited this issue in the late 1940s, fighting a contentious political battle against the AMA, commercial insurers, and the busi- ness community in a failed bid to convince the 80th and 81st Congresses to enact Wagner-Murray-Dingell legislation (Starr, 1982).3 President Truman added the key ingredient of political leadership but lacked sufficient votes in Congress, facing a hostile reception in Congress from Southern Democrats committed to reversing New Deal policies and opposed to his Administrationâs civil rights policies (Poen, 1979; Maioni, 1998). The only public coverage to be expanded would be federal grants-in-aid to the states to support indigent care, in the 1950 Amendments to the Social Security Act (Stevens, 1989). In the 1950s enrollment in commercial policies outgrew that of nonprofits (Somers and Somers, 1961; Starr, 1982; Cunningham and Cunningham 1997). One advantage that commercial insurers could offer purchasers of group policies was a less expensive alternative to nonprofit âcommunity ratingâ (where actuarial risk was spread broadly through uniform premiums) in the form of âexperience- rating,â or charging premiums according to the claims experience of the employerâs workforce. Experience rating made policies less expensive for healthier employee groups and more expensive or unavailable for individuals, particularly ill or dis- abled persons, who tried to purchase coverage outside the workplace (Starr, 1982). The growing centrality of health insurance revenue to the fiscal health of the health care system created interest on the part of providers (physicians, hospitals) in seeing greater numbers of the general population covered by insurance, albeit on a voluntary basis (IOM, 2003a). For persons without coverage, the price of being uninsured was growing, as the increasingly widespread coverage fueled higher costs for health care and doubled hospital prices during the 1950s (Starr, 1982). Following the defeat of President Trumanâs campaign for universal coverage 2In 1942, a War Labor Board ruling made health coverage an attractive fringe benefit for employ- ers to offer workers, whose salaries were restricted due to wartime wage controls; between 1942 and 1945, there was more than a fourfold increase in group hospitalization enrollment (Starr, 1982). A 1948 U.S. Supreme Court decision allowed unions to collectively bargain for health insurance, and enrollment jumped (Starr, 1982; Gabel, 1999). 3The Truman Administrationâs proposal included medical and hospital benefits and federal grants- in-aid to the states to finance premiums for the poor, with national administration and financing through a 3 percent payroll tax split between firms and workers (Marmor, 1973).
LESSONS FROM THE PAST AND PRESENT 73 in the late 1940s, proposals to extend health insurance by means of public policies focused on incremental change. In 1951, the handful of public officials who had worked for universal coverage shifted tactics, proposing a new 60-day hospitaliza- tion benefit for Social Security beneficiaries. This benefit would be funded off- budget, using Social Security trust dollars to lessen the economic burden of health services already used by the over-65 population (rather than to increase access to care) who were less likely to carry private insurance (Maioni, 1998). Their pro- posal aided the perception that the middle class (i.e., those who had paid Social Security taxes) had earned such coverage. They sidestepped AMA criticism by not including physician services (Marmor, 1973).4 The proposed reform brought out traditional allies and opponents of mandatory coverage extensions, pitting the AMA against organized labor (the AFL-CIO) (Starr, 1982). Annual hearings on hospitalization proposals failed to advance reform, although the 1960 Kerr-Mills Act amending Social Security did acknowledge the interest in financing care for low-income seniors, increasing grants-in-aid to the states that raised sufficient matching funds (Marmor, 1973). An Incremental Compromise for Universal Coverage: Medicare and Medicaid In the late 1950s, the level of health insurance coverage was at an all-time high, although only about 8 percent of the population had coverage that could be called comprehensive (i.e., insurance that covered hospital stays and physician services) (Somers and Somers, 1961). Employment-based coverage was becoming the norm, with gaps of uninsured populations growing among those who did not or could not receive an employerâs offer of group coverage. A crescendo of political and legislative activity on hospital insurance in the early 1960s led to the first significant extension of publicly mandated coverage (Medicare and Medicaid) after the 1964 elections returned President Lyndon B. Johnson to the White House and brought large Democratic majorities favoring passage to Congress. In the years immediately before, President John F. Kennedy had campaigned for a Social Security hospitalization benefit, a scaled-down ap- proach compared with President Trumanâs universal coverage proposals. Despite annual hearings, the garnering of public support by the Administra- tion, and the strategic building of votes on the Hill, House Ways and Means Chair Wilbur Mills released a coverage bill only after legislative developments in 1964 and the elections threatened his continued leadership on health insurance (Marmor, 1973; Maioni, 1998). When the 89th Congress convened in 1965, Medicare was 4Proposals tied to the Social Security system were not for universal coverage; in 1952, only 7 million of the roughly 12.5 million persons over age 65 would have been covered by the proposed reform (Marmor, 1973).
74 INSURING AMERICAâS HEALTH at the top of the agenda and was signed into law later that year, together with Medicaid, as amendments to the Social Security Act (Fox, 1986; Hacker, 1997). Medicare and Medicaid passed because of the unique legislative alignments of 1965 and also because of Chairman Millsâ leadership in crafting a political com- promise acceptable to interest groups, especially the AMA, which continued to lobby against the Administrationâs plans (Marmor, 1973; Maioni, 1998). He de- scribed the amendments as a âthree-layer cakeâ with something for everyone: â¢ Medicareâconsisting of Part A, hospitalization, for which there is auto- matic and permanent enrollment and financing through Social Security, and Part B, coverage for physician visits, with voluntary one-time enrollment and a monthly premium paymentâprovided nearly universal coverage for persons ages 65 and older.5 Part A achieved the goal of social insurance advocates and the Johnson Administration, while Part B accommodated the AMAâs objections to a manda- tory program. â¢ Medicaid, the third layer of the cake, covers certain categories of low- income persons and makes existing grants-in-aid to state programs more uniform in terms of eligibility and benefits (Marmor, 1973).6 Medicaid addressed the interest of the hospital industry in alleviating the burden of uncompensated care. To win support for passage from provider groups, the new law essentially adopted the reimbursement approaches of Blue Shield (for physicians) and Blue Cross (for hospitals), including few limits on reimbursement for physicians, and added favor- able provisions to pay hospitals for capital depreciation. However, these two aspects of the 1965 statute laid the groundwork for significant growth in health care inflation and spending (Starr, 1982).7 The Medicare and Medicaid Amendments represented a limited extension of health insurance coverage. Medicare also augured for improved access to care for African Americans, as certification for a hospital to participate in the program was conditioned on its desegregated status (Reynolds, 1997). While social insurance advocates saw the amendments as an opening wedge for future growth in coverage levels, more pragmatic observers interpreted their design as a way to fend off 5Amendments have added Medicare coverage for persons certified to be permanently and totally disabled and persons with end-stage renal disease (IOM, 2001a). 6Eligibility to enroll in Medicaid involves fulfilling requirements related to income and assets (making it a so-called means-tested program) and being a member of a specific group or category that is eligible for benefits, for example, a minor child. Those who meet economic and categorical criteria must also meet immigration status and state residency requirements. Eligibility standards vary by state and have changed from year to year, with general oversight provided by the federal government (IOM, 2001a). 7The Hill Burton Act of 1946 substantially expanded hospital capacity across the nation in the years before implementation of Medicare, contributing to the availability and utilization of health services nationally.
LESSONS FROM THE PAST AND PRESENT 75 further broadening of the public mandate. For example, financing Medicare phy- sician services with general tax revenues and beneficiary contributions, and using Medicaid to fill coverage gaps for persons under age 65, extended coverage while reinforcing existing approaches (i.e., voluntary coverage for the services most often used, and means-testing for benefits for the poor) (Marmor, 1973). Implementation of Medicare and Medicaid introduced tens of millions of newly insured persons, and billions of new public dollars, into the health care system. By 1970, 20.5 million people were enrolled in Medicare and 18 million people in Medicaid (Bovbjerg et al., 1993). Between 1965 and 1975, annual federal spending on health care jumped from less than $10 billion to more than $40 billion (Hacker, 1997). There was a dramatic increase in utilization; within a year of implementation, 20 percent of all persons 65 years and older had used Medicare for hospital services and 12 million had used Part B coverage for physi- cian services (Marmor, 1973). Nevertheless, gaps in coverage and financial protec- tion remained. Medicaid covered only an estimated one-third of those considered poor, and Medicare reimbursed less than half of seniorsâ spending on health and long-term care services (Starr, 1982). Paying for Reform: Cost Containment Joins Access as a Focus for Reform in the Years Since 1965 With federal taxes supporting coverage for a large segment of the population, and with the country in an economic recession, the rapidly increasing costs of health care to society overall and interest in improving the efficiency (including financing) of the health care system motivated reform (Starr, 1982; Lewis, 1983; Steinmo and Watts, 1995; Hacker, 1997). Public officials, insurers, and employers were united in the widely shared belief that a poorly organized and inefficient health care system fueled health care inflation and that universally mandated coverage could bring cost savings (Starr, 1982).8 A combination of factors defeated efforts to extend coverage further. Some were unique to the political circum- stances of the day (i.e., the impeachment proceedings against President Nixon and Wilbur Millsâ fall from power in 1974). Others, such as the inability to fashion a workable consensus on a congressional bill and opposition from commercial insur- ers and employers, were familiar from earlier efforts to achieve broad-based re- forms. As analyst Stuart Altman has observed of the era, which leads up to the present, many policy actors who previously opposed reform altogether have joined the call for change, but they have often been unwilling to compromise their own vision of reform and the absence of change has been the second-best or fallback option (Kahn and Pollack, 2001). During the early 1970s, members of Congress and the Nixon Administration 8Such interest in improving the design and efficiency of the health care system was reflected in the Health Maintenance Organization Act of 1973, for example.
76 INSURING AMERICAâS HEALTH generated a number of bills to extend coverage and constrain costs, either by filling gaps in the existing mix of public and private coverage (the Nixon Administrationâs proposed employer mandate funded by a payroll tax, paired with expanded public coverage and regulation of payment rates) or by replacing privately purchased policies with a federally administered system under a national budget (e.g., Senator Edward Kennedyâs Health Security Act of 1970) (Starr, 1982; Etheredge, 1990; Hacker, 1997; Maioni, 1998). Despite the resistance of providers, employers (especially small businesses), insurers, and the states to the regulatory aspects of proposed reform, passage of a universal coverage bill came close in 1974 (Starr, 1982; Etheredge, 1990; Steinmo and Watts, 1995; Hacker, 1997). President Nixonâs February message to Congress laid out a Comprehensive Health Insur- ance Plan that combined Medicare, Medicaid, and an employer mandate (Davis, 2001). The lack of political leadership to forge a consensus on one of the many proposed plans in Congress by the summer of 1974, however, combined with the turmoil of President Nixonâs August resignation and the decision of organized labor to withhold its support from reform until after the November elections, resulted in the disintegration of what bipartisan agreement existed (Starr, 1982; Maioni, 1998). After 1974, ongoing economic recession and price inflation, par- ticularly in health care, sank any further serious debate about reforming health care financing to extend coverage. Cost controls became the key emphasis during the Ford and Carter Administrations.9 By the late 1970s, health insurance coverage reached its highest level yet; in 1980, roughly 15 percent of the general population under age 65 was uninsured, about 29.6 million people (Bovbjerg et al., 1993). From 1970 to 1980, aggregate spending on health care had continued to grow, from $69 billion to $230 billion and from 7.2 percent to 9.4 percent of the gross domestic product (Starr, 1982). For employers, health insurance premiums were becoming more expensive, lead- ing them to increase deductibles, decrease dependent coverage, and, as the 1980s wore on, turn to commercial managed care contracting to restrain health care cost increases (Hacker, 1997). In an effort to restrain the continued growth of health care spending under Medicare, federal officials replaced fee-for-service reimburse- ment with capitation. During the 1980s, there would be incremental public expansions, particularly through the Medicaid program, which will be discussed later in this chapter. Absent major reforms of health care financing to extend coverage in the future, the stage was set for the major increases in the number of uninsured seen during the 1980s (Lewis, 1983; Hacker, 1997). Interest in comprehensive national reform of health care financing surfaced again in the early 1990s. Economic recession and continuing health care inflation renewed interest among middle-class voters (Hacker, 1997). During President 9The Carter Administration did develop a universal coverage proposal, never introduced in Con- gress, called the National Health Plan, which divided coverage between private sources (employer mandate) and public sources (a federal program to replace Medicare and Medicaid and expand eligi- bility to all lower income persons) (Congressional Quarterly, 1977; Davis, 2001).
LESSONS FROM THE PAST AND PRESENT 77 Clintonâs first term, his Administration convened a task force of hundreds of experts to craft a legislative proposal for universal coverage (Hoffman, 2001). The Clinton plan was built around an employer mandate, and within the confines of a large federal deficit, with the federal government both restraining health care costs and extending coverage through regional purchasing pools (Skocpol, 1996; Hacker, 1997). Private managed care health plans would compete for contracts with the regional pools, introducing the idea of managed competition (Enthoven, 1978a,b; Starr, 1992; Hacker, 1997). Although there is no single definitive expla- nation for the failure to enact universal coverage in the mid-1990s, many analytic interpretations stress the Administrationâs difficulties in mobilizing public and political support for the complex proposal, the uncertainties involved in the pro- posed overhaul of the nationâs health care financing and delivery arrangements, and the crystallization of Republican opposition to the Clinton proposal as con- tributing to the defeat of the plan before it reached a vote in Congress (Skocpol, 1996; Hacker, 1997; Marmor and Barer, 1997). There was vigorous political opposition of provider, insurer, and business community groups, many of whom favored reform but perceived the terms of the Clinton proposal to be economi- cally threatening. This opposition played out in media campaigns that lowered public support for proposed reform. Despite multiple reform campaigns over the past century, enduring characteristics have shaped the current fragmented approach to financ- ing health care services, including: â¢ the presence of well-organized and well-financed provider, in- surer, and business groups with economic stakes potentially adversely affected by proposed coverage expansions (although many have agreed with the need for reform); â¢ the obstacles posed to comprehensive policy change by the Ameri- can political process; and â¢ insufficient political leadership to fashion a workable consensus and then shepherd that consensus through the legislative process. In the wake of the failure to implement universal coverage, the federal, state, and some local governments in recent years have experimented with incremental coverage expansions with impacts much more modest than that of Medicare and Medicaid in the 1960s. In the sections that follow, the Committee will first explore federal expansions of Medicaid, the State Childrenâs Health Insurance Program (SCHIP), and targeted federal programs related to employment-based coverage. Following that is a discussion of five states (Hawaii, Massachusetts, Minnesota, Oregon, and Tennessee) that have led the way in covering more residents. Lastly, the experiences of three counties that have adopted innovative approaches to extending coverage are discussed to illustrate local reform efforts and their limits.
78 INSURING AMERICAâS HEALTH FEDERAL INITIATIVES TO EXTEND COVERAGE SINCE 1984 The level of employment-based coverage has a fundamental effect on uninsurance. Public extensions of coverage at the federal, state, and local level, as well as public policies affecting the availability of individually purchased (nongroup) policies, have filled some of the gaps remaining from the employment-based approach (IOM, 2001a). Almost two-thirds of those under age 65 obtain health insurance through their workplace or that of a family member, with public pro- grams covering an additional 15 percent and individually purchased policies al- most 7 percent (Fronstin, 2002). This section reviews major federal extensions of coverage since the mid- 1980s, both public (Medicaid expansions to cover pregnant women, infants, and young children, and the State Childrenâs Health Insurance Program) and private (federal regulation of the market for private insurance). None has approached even the partial success of the initial Medicare and Medicaid programs in boosting coverage in the general population. As Figure 3.1 depicts, there has been little overall change in the proportions of privately and publicly insured persons and in uninsured rate overall for the nation over the period from 1987 to 2002. Since the early 1980s, the number of uninsured persons under age 65 has grown in step with general population growth, while employment-based coverage has alternately grown, declined (in the late 1980s and early 1990s), grown again through the late 1990s, and has declined once again since the turn of the century. Tempered by public expansions of coverage, the uninsured rate as a result has varied within a few percentage points, between 14.9 percent and 17.2 percent of the noninstitu- tionalized population under age 65 (Levit et al., 1992; Fronstin, 2002; Mills and Bhandari, 2003). Federal extensions of public coverage since the mid-1980s have often fol- lowed the lead of successful state insurance reforms, for example, in the case of SCHIP and the regulation of private-sector plans through the Health Insurance Portability and Accountability Act of 1996 (HIPAA) and the Consolidated Omni- bus Budget Reconciliation Act of 1985 (COBRA). They have lowered uninsured rates among lower income children (in households earning less than 200 percent of poverty) and boosted the numbers of lower income persons with public cover- age. At the same time, the proportion of adults with public coverage has declined, reflecting shifts in the demographic composition of the general population as well as public policy changes such as welfare reform. Concurrently with broadened eligibility for public coverage, new federal statutes (i.e., COBRA, HIPAA, and the Trade Act) have been enacted to protect and support the retention of work- place health benefits. These have established important precedents in principle for uninterrupted health insurance coverage but overall have had a limited impact on uninsurance, for reasons to be discussed. These extensions have succeeded in improving coverage among selected groups when the programs receive adequate and stable public funding (particularly
300 Uninsured Public Insurance Individually Purchased Insurance Employment-Based Coverage 250 200 150 Estimated Number (Millions) 100 50 29.5 31.1 31.7 32.9 33.6 35.4 36.4 36.5 37.3 38.3 39.9 40.7 39.0 39.4 40.9 43.3 0 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 FIGURE 3.1 Sources of health insurance coverage and number of uninsured persons for population under age 65 years, 1987â2002. NOTE: Coverage reported as military (i.e., Tricare, CHAMPVA) is counted as employment-based. Estimates for 2000, 2001, and 2002 use Census 2000-based weights. Precise estimates do not add to 100 percent due to rounding and because respondents may report more than one source of coverage within a year. SOURCES: Fronstin, 2002; Mills and Bhandari, 2003. 79
80 INSURING AMERICAâS HEALTH federal dollars), when they foster continuity of coverage within families or among health plans, and when they make insurance coverage affordable both to individu- als and families and to society over time. In addition to specific Committee findings for each of the expansions, to be discussed, the Committee draws three more general findings about federal initiatives. Finding: Federal incremental reforms over the past 20 years have made little progress in reducing overall uninsured rates nationally, although public program expansions have improved coverage for targeted, previously uninsured groups. Federal reforms of employ- ment-based insurance have not included provisions for assuring affordability and, thus, have had limited effect. Finding: Extensions of program eligibility for one group of unin- sured often affect the coverage status of other population groups indirectly, for example, when SCHIP enrollment efforts identify children who are eligible for but not enrolled in the Medicaid pro- gram. Finding: Public programs fall short of their coverage goals when not all eligible persons enroll. When outreach and enrollment are made a priority, coverage levels rise. Public coverage programs some- times employ administrative barriers to enrollment to contend with inadequate or unstable funding during periods of economic stress within the state. Medicaid Expansions for Pregnant Women, Infants, and Children, 1984â1990 Finding: Incremental coverage extensions through Medicaid have been less effective than initially anticipated. A sizable population is eligible but not enrolled or does not maintain enrollment over time, diminishing the effectiveness of the coverage. When Medicaid was created in 1965, some envisioned that Medicaid and Medicare together would fill the coverage gaps left by the employment-based health insurance system (Marmor, 1973). However, with categorical structures of eligibility tied to the rules of the cash assistance programs, the income eligibility levels then in existence, and funding shortfalls, state Medicaid programs did not cover all those with incomes too low to afford private health insurance coverage (Starr, 1982; Stevens, 1989).10 Medicaid significantly raised the level of federal 10Most Medicaid enrollees have been low-income children and their mothers. A sizable group of adults are eligible for but not enrolled in Medicaid, roughly 16 percent of the 19 million uninsured
LESSONS FROM THE PAST AND PRESENT 81 support for the states, with matching dollars pegged to per capita income in each state, and made eligibility for public coverage an entitlement. Such entitlement was limited to certain categories of persons (see Box 3.2 for current federal eligibility requirements); then, as now, single, childless adults under age 65 often are the group least eligible for Medicaid. The scope of each stateâs program dictates how far dollars go to pay for the federally mandated minimum benefits package for Medicaid recipients. Expansion of the Medicaid program was prompted by public concern about the worsening health status of poor children, indicated by a reversal in the long- term decline in infant mortality rates, and the diminishing percentage of low- income children and mothers enrolled (due to more stringent eligibility standards for welfare) (Schlesinger and Kronebusch, 1990). Between 1984 and 1990, the Congress altered categorical eligibility requirements and gradually increased Med- icaid income eligibility levels for pregnant women, infants, and children through provisions in its annual spending bills; Box 3.3 summarizes the legislative provi- sions. These changes delinked Medicaid coverage for pregnant women, infants, and children from the requirement that they meet their stateâs welfare eligibility requirements, a process continued with welfare reform in 1996 (KCMU, 2002a).11 Some states folded these changes into Section 1115 Medicaid waivers granted by the federal government for research and demonstration projects that are de- signed to be cost neutral. These and other waivers from required minimum benefits and restrictions on categories of eligibility allow states to offer coverage more widely within the limits of their federal matching funds (Lambrew, 2001). Under waivers, some states have reallocated Medicaid Disproportionate Share Hospital (DSH) funds to expand eligibility (e.g., Tennessee), while others have attempted to stretch Medicaid dollars by shifting portions of their Medicaid popu- lation from fee-for-service to managed care contracting (e.g., Arizona, Maryland, New York, and Rhode Island) or by dampening utilization by charging premiums adults with incomes under twice the poverty level (Schneider, 2002). As of 2002, there were 47.2 million persons enrolled. More than half (51 percent) were children, about one-fifth (22 percent) were nondisabled adults under age 65, 17 percent were certified as blind or disabled, and another 10 percent were low-income persons at least 65 years old (KCMU, 2003b). The bulk of Medicaidâs $212 billion spending goes for its blind and disabled recipients (44 percent of spending) and seniors (27 percent of spending). Children and nondisabled adults under age 65 together account for less than one-third of program spending (29 percent), though they represent 73 percent of enrollees. 11Through section 1931, part of the Personal Responsibility and Work Opportunity Reconcilia- tion Act of 1996, Medicaid eligibility for parents was broadened, delinking Medicaid enrollment from welfare participation and basing minimum eligibility on a stateâs welfare eligibility criteria as of July 1996, even if parents were not enrolled in welfare. It also permitted states to loosen income and asset tests, as well as adjust these tests for inflation (so-called income and asset disregards) (Birnbaum, 2000; Broaddus et al., 2002).
82 INSURING AMERICAâS HEALTH BOX 3.2 Eligibility for Medicaid The federal government requires that the following groups be entitled to enroll in Medicaid: â¢ federal required minimum: children under age 6 and pregnant women whose family income is below 133 percent of the federal poverty level (FPL) ($20,296 for a family of three in 2003); â¢ federal required minimum: children ages 6â18 with family incomes at or below 100 percent of FPL ($15,260 for a family of three in 2003); â¢ no federal minimum: states set income standards for adults without chil- dren; parents of children are categorically eligible if they meet income and asset tests. On average, the statesâ income eligibility level for parents is 41 percent of FPL ($6,257 for a family of three in 2003), varying from a low of 21 percent ($3,205 for a family of three) in Alabama to a high of 275 percent ($41,965 for a family of three) in Minnesota; â¢ Supplementary Security Income (SSI) recipients or those aged, blind, and disabled individuals who qualify in states that apply more restrictive eligibility re- quirements; â¢ recipients of adoption assistance and foster care who are under Title IV-E of the Social Security Act; â¢ special protected groups (typically individuals who lose their cash assis- tance from SSI due to earnings from work or increased Social Security benefits but may keep Medicaid for a period of time); and â¢ qualified Medicare beneficiaries, specified low-income Medicare beneficia- ries, and disabled-and-working individuals who previously qualified for Medicare but lost their coverage because of their return to work. and imposing cost sharing for new categories of enrollees (e.g., Arkansas, New Mexico, and Oregon) (Mann, 2002; Schneider, 2002).12 Expanding Medicaid eligibility levels brought concern that increased enroll- ment reflected the substitution of public insurance for employment-based cover- age for workers and dependents at the lower end of the wage scale. Studies give a wide range of estimates of the amount of substitution (often referred to as âcrowd outâ), although none is fully comparable in terms of methods, data sources, or criteria measured (Cutler and Gruber, 1997; Dubay, 1999; Alteras, 2001). Early estimates of Medicaid substitution approaching 50 percent (Cutler and Gruber, 1996a, b) have since been reevaluated. Recent estimates have been as low as 4 percent, with upper bounds between 17 percent and 23 percent, depending on how substitution is defined and measured (Cutler and Gruber, 1997; Alteras, 2001). This concern about substitution of Medicaid coverage for employment- 12In 1997, federal law was revised so that implementation of Medicaid mandatory managed care contracting no longer required a Section 1115 waiver (Mann, 2002).
LESSONS FROM THE PAST AND PRESENT 83 In addition, federal law permits the states to extend eligibility beyond the min- imum requirements listed. Most but not all state expansions may be supported by federal matching funds. The most common optional expansions by the states in- cluding eligibility for the following groups: â¢ infants up to age 1 and pregnant women not covered under the mandatory rules whose family income is up to 185 percent of FPL; â¢ recipients of state supplementary income payments; â¢ certain aged, blind, or disabled adults who have incomes above those re- quiring mandatory coverage but below the FPL; â¢ persons receiving care under home and community-based waivers; â¢ persons infected with tuberculosis (TB) who would be financially eligible for Medicaid at the SSI income level (eligibility is only for TB-related ambulatory ser- vices and for TB drugs); â¢ institutionalized individuals with income and resources below specified lim- its; â¢ medically needyâpersons who meet categorical requirements and have significant health care expenses with incomes in excess of the mandatory or op- tional levels; these individuals may âspend downâ to Medicaid eligibility by incur- ring medical and/or remedial care expenses to offset their excess income, thus reducing it to a level below the maximum income allowed by that stateâs Medicaid plan; and â¢ legal resident aliens and other qualified aliens who entered the United States on or after August 22, 1996, made ineligible for Medicaid for five years because of the 1996 welfare reform law. SOURCES: Broaddus et al., 2002; IOM, 2002b; KCMU, 2002b; DHHS, 2003. sponsored health insurance, which influenced the design for SCHIP, will be discussed. Infant mortality rates improved in the latter half of the 1980s, reflecting the development of neonatal intensive care units made possible by Medicaid financing even before the eligibility expansions (Cutler and Gruber, 1996b; Howell, 2001). One study estimates an 8.5 percent decrease in the infant mortality rate associated with the 30 percentage point increase in Medicaid eligibility for women of repro- ductive age (15 to 44 years old) between 1979 and 1992, with the expansions targeted to women eligible for welfare having more of an effect on health (Currie and Gruber, 1996b). Millions of children and adults remain eligible for but not enrolled in Medic- aid. Difficulties in enrolling and maintaining enrollment, confusion about eligibil- ity, the decision by some who are eligible not to participate in public coverage, and, in the case of children, the diminished likelihood of enrollment when their parents are ineligible for the same program as their child, are uninsured or have weak connections to the health care system, all contribute to low take-up rates for Medicaid (IOM, 2002b; Schneider, 2002). Particularly for those at the upper end
84 INSURING AMERICAâS HEALTH BOX 3.3 Medicaid Expansions, 1984â1990 â¢ Deficit Reduction Act of 1984. Medicaid eligibility required for children born after September 30, 1983, up to their fifth birthday, and for women either pregnant for the first time or pregnant in an unemployed family with two parents, in families eligible for Aid to Families with Dependent Children (AFDC) (welfare), roughly 40 percent of the FPL. Required eligibility for one year for infants born to Medicaid-eligible women. â¢ Consolidated Omnibus Budget Reconciliation Act of 1985 (COBRA). Medicaid eligibility required for children up to age 5 in families eligible for AFDC and for pregnant women in families both eligible for AFDC and with two parents (no longer required to be unemployed). â¢ Omnibus Budget Reconciliation Act of 1986. Option to expand Medicaid eligibility to pregnant women and children under age 5 in families earning no more than 100 percent FPL. Note that at the time, eligibility for state welfare programs averaged about 45 percent (KCMU, 2002a). Option to use presumptive and con- tinuous eligibility to expand Medicaid to pregnant women. â¢ Omnibus Budget Reconciliation Act of 1987. Medicaid eligibility required for children under age 8 in families eligible for AFDC on basis of income. Option to expand Medicaid eligibility to pregnant women and infants in families earning no more than 185 percent of FPL and children under age 8 in families earning no more than 100 percent of FPL. â¢ Medicare Catastrophic Coverage Act of 1988. Medicaid eligibility re- quired for pregnant women and infants in families earning less than 100 percent of FPL, to be implemented over time. Although most provisions of the Act were re- pealed in 1989, this provision remains in effect. â¢ Family Support Act of 1988. Option to expand Medicaid eligibility to preg- nant women and children above required levels. â¢ Omnibus Budget Reconciliation Act of 1989. Medicaid eligibility required for pregnant women and children under age 6 in families earning no more than 133 percent of FPL. â¢ Omnibus Budget Reconciliation Act of 1990. Medicaid eligibility required, to be phased in by September 2002, for children ages 6 through 8 in families earning up to 100 percent of FPL. SOURCE: Mann et al., 2003.
LESSONS FROM THE PAST AND PRESENT 85 of the income eligibility scale, take-up rates for persons newly eligible under the Medicaid expansions were relatively low (LoSasso and Buchmueller, 2002). The Omnibus Budget Reconciliation Act of 1990 mandated that eligibility for chil- dren through age 18 in families earning less than the federal poverty level (FPL) be phased in by September 2002, so that most U.S.-born children in this category are now eligible for public coverage.13 Yet the latest Census Bureau data indicate that nearly a quarter of children in this income bracket remain uninsured; foreign-born children who are ineligible for Medicaid comprise a part but not all of this uninsured group (Ku and Blaney, 2000; IOM, 2001a; Fronstin, 2002). The State Childrenâs Health Insurance Program and Medicaid, 1997â2002 Finding: SCHIP has extended coverage among children to a signifi- cant degree and, to a much lesser extent to date, their parents. Finding: The reduction in uninsurance achieved through SCHIP is jeopardized by the programâs financing structure as a time-limited program of grants-in-aid rather than an individual entitlement. Despite the Medicaid expansions of the late 1980s, nearly 15 percent of children under age 18 were uninsured in 1997 (Fronstin, 2000). Many of these children were ineligible for Medicaid because their families earned too much for them to qualify for Medicaid yet too little to afford private coverage. In addition, welfare reform in 1996 had led many low-income children and their parents to drop out of the Medicaid program because they did not realize that eligibility no longer depended on being a welfare recipient (Mann et al., 2003). In 1997, with a political consensus on the need for children to be insured, Congress enacted SCHIP as part of the Balanced Budget Act.14 SCHIP created a new option for states to use matching funds from a capped, ten-year grants-in-aid program providing roughly $40 billion to the states (fiscal years 1998 through 2007). Funds are unevenly allocated over the decade, with a âdipâ in federal dollars scheduled for fiscal years 2003â2005 likely to result in program cutbacks (Dubay et al., 2002b). Most states quickly established their own SCHIP programs, structuring them as Medicaid expansions, as a new separate program, or as a combination of the two. As of February 2003, 19 states have 13Since 1996, welfare and immigration reform legislation has barred legal immigrants who arrive after August 1996 from eligibility for Medicaid, SCHIP, and other federal means-tested benefits programs for their first five years in the United States, except for the financing of emergency care, with exceptions made for specific categories of persons, including refugees (Rosenbaum, 2000). See discussion in the Committeeâs first report, Coverage Matters (IOM, 2001a). 14Title XXI of the Social Security Act.
86 INSURING AMERICAâS HEALTH BOX 3.4 How the State Childrenâs Health Insurance Program (SCHIP) Differs from Medicaid Programs Eligibility Medicaid creates an entitlement for those who meet the eligibility criteria and SCHIP does not.1 This fundamental distinction between the two programs means that states can reduce or eliminate their SCHIP programs at any time, jeopardizing progress in reducing uninsurance (Rosenbaum and Smith, 2001). States using the SCHIP approach can limit total enrollment even for those otherwise eligible for the program. While Medicaid recipients may also have access to private coverage, SCHIP enrollees must be eligible for no other type of public coverage and not be covered under a private health plan. While Medicaid eligibility focuses on pregnant women, infants, and younger children (children through age 18 are eligible if they live in families earning less than poverty), SCHIP makes age breaks in eligibility more uniform among the states and broadens eligibility up to twice the poverty level for all children (Ullman and Hill, 2001; Dubay et al., 2002a). Under Medicaid waivers, states have used SCHIP dollars that they are not using to cover eligible children to raise eligibility levels for low-income parents, either directly or by means of premium assistance to purchase employment-based coverage.2 Financing Both programs are financed by federal matching grants to the states, but SCHIP provides a more generous federal match, although it is coupled with a cap on the total federal amount available to each state (Wooldridge et al., 2003). States have access to their annual allotment over a three-year period, with unspent funds being returned to the general pool of SCHIP dollars after that point and available to SCHIP programs separate from Medicaid, 21 states have expanded Medicaid as their SCHIP program, and 16 states have combined a separate SCHIP with expanded Medicaid (CMS, 2003b).15 See Box 3.4 for a description of SCHIPâs key features. The program was originally intended to reach 40 percent of children unin- sured at the time, targeting children and families earning between 100 and 200 percent of poverty and allowing states to raise eligibility above 200 percent if their existing Medicaid program already covered children at twice poverty (Wooldridge et al., 2003). State SCHIP programs vary in their maximum income eligibility thresholds and in eligibility levels for children at different ages; on average, the proportion of children between 100 and 200 percent of poverty eligible for public coverage rose from 22 percent to 82 percent (LoSasso and Buchmueller, 2002). 15Note that the count includes the 50 states and the District of Columbia, plus 5 territories.
LESSONS FROM THE PAST AND PRESENT 87 be allocated to states that have spent their entire allotment, unless restricted by federal rules. Unlike Medicaid, where premiums and copayments for children are prohibited, under SCHIP modest out-of-pocket expenditures are permitted within federal constraints (e.g., copayments, premiums, enrollment fees, deductibles) (Wooldridge et al., 2003). Benefits SCHIP programs created as part of an existing Medicaid program must offer Medicaidâs comprehensive benefits package, while programs established sepa- rately are held to a lower standard, with federal law establishing the minimum benefits package required. However, most separate SCHIP programs include ben- efits more generous than the federally required minimums, with one-third of these programs offering the same benefits as Medicaid, and a national evaluation of SCHIP finds a consensus among respondents that SCHIP benefits packages gen- erally meet the needs of most children enrolled (Weil and Hill, 2003; Wooldridge et al., 2003). 1Populations to whom states choose to extend Medicaid eligibility (so-called optional groups, rather than groups for whom the federal government mandates eligibility) may not be entitled to coverage, which may be restricted or withdrawn by the state at any time. 2Eight states have received waivers to expand eligibility to low-income parents and 7 have obtained waivers to use SCHIP funds for premium subsidies (Wooldridge et al., 2003). The new Health Insurance Flexibility and Accountability waivers, a revamped version of the Sec- tion 1115 waiver, are also being used by 4 states to expand eligibility to low-income parents using SCHIP, starting in January 2001 (Schneider, 2002; Wooldridge et al., 2003). During 2002, an estimated 5.3 million children were enrolled in SCHIP at one time or another, about one-seventh as many as are enrolled in the Medicaid program (CMS, 2003a; Wooldridge et al., 2003). In the process of actively engaging in outreach to build and maintain SCHIP enrollment, states have increased Medicaid enrollments (Dubay et al., 2002b; Wooldridge et al., 2003). A study of SCHIP at 12 urban sites across the country (representative and nonrandom sample) in 2000 and 2001 found that outreach strategies allowed states to overcome their initial difficulties in reaching enroll- ment targets and in some cases exceeding their goals for SCHIP, as well as raising Medicaid enrollments (Felland and Benoit, 2001; Mann, 2002). The states stream- lined applications and application processes, used materials translated into lan- guages other than English, and involved local community-based groups and orga- nizations (e.g., health clinics, schools, hospitals, employers) in outreach campaigns tailored to reach groups with high uninsured rates (e.g., ethnic and racial minori- ties, immigrants) (Felland and Benoit, 2001). A follow-up study of enrollment between 1997 and 2001 observed a large increase during 2000â2001 and the
88 INSURING AMERICAâS HEALTH biggest increases in coverage in areas with the highest uninsured rates for children (Cunningham et al., 2002). As noted earlier, the design of SCHIP programs has reflected the concern that public coverage not substitute for existing employment-based health insurance (Lutzky and Hill, 2001). This concern is greater with SCHIP than with Medicaid because SCHIP has higher income eligibility levels, and extensions to parents and premium assistance for employment-based coverage are planned. Federal regula- tions require that states explicitly attempt to minimize substitution, for example, by requiring that enrollees be uninsured for a specified period of time before enrolling, as do two-thirds of the states (Alteras, 2001; Lutzky and Hill, 2001). Because of differences in how substitution is defined and in analytic methods, no consensus exists about the extent of its impact. Some researchers conclude that the lack of affordable private coverage for families earning less than 150 percent of FPL makes substitution less of a concern. An interim evaluation of SCHIP in New York finds that only 4 to 6 percent of children enrolled in Child Health Plus had parents who reported dropping private coverage within the previous 6 months (Lutzky and Hill, 2001; Wooldridge et al., 2003). Alternatively, two recent studies of changes in coverage between 1997 and 2001 for children in families earning less than twice the poverty level report estimated levels of substitution comparable to early estimates under the Medicaid expansions of the late 1980s and early 1990s of between 18 and 50 percent (Cunningham et al., 2002; LoSasso and Buchmueller, 2002). A national interim evaluation of SCHIP notes its successes in â¢ extending broad and affordable coverage, â¢ providing greater access to health care for the millions of children newly enrolled, â¢ securing ongoing federal financial support, and â¢ improving outreach and streamlining program administration compared with Medicaid (Woodridge et al., 2003). Between 1997 and 2001, the proportion of uninsured children in families earning less than twice the poverty level declined nearly 4 percentage points, and there was an even larger drop for children between 100 and 200 percent of poverty (Cunningham et al., 2002). For children in families with income below the poverty line, however, the uninsured rate increased during the 1990s (Holahan et al., 2003a). For all children nationally between 1994 and 2000, there was a decline in public coverage, from 18.5 percent to 16.4 percent, and it was growth in employment-based coverage, reflecting economic prosperity and welfare re- form, that lowered uninsured rates (Holahan et al., 2003a). Despite outreach efforts and measures to simplify enrollment and reenroll- ment, SCHIP resembles the Medicaid expansions that preceded it in that only two-thirds of eligible children are enrolled (Dubay et al., 2002b). In 2002, there were 7.8 million uninsured children, and more than 4 million of these were eligible for but not enrolled in either Medicaid or SCHIP (Kenney et al., 2003).
LESSONS FROM THE PAST AND PRESENT 89 Federal Regulation of Private (Employment-Based) Coverage Finding: Federal initiatives to regulate portability and renewability of employment-based coverage have failed to reduce overall rates of uninsurance significantly because they have been too narrowly tar- geted and have not addressed the affordability of insurance premi- ums. Because most Americans receive their health insurance through their employ- ers or as dependents on anotherâs employment-based policy, there has been much interest in reducing uninsurance by extending such coverage, through insurance market reforms or regulation of premiums, benefit packages, and eligibility. The federal government influences the degree of private coverage generally through its favorable tax treatment of premiums (the employerâs contribution is tax exempt for both the employer and the employee). Although much of the authority to regulate insurance products is reserved to the states, the U.S. Department of Labor may regulate qualified employer benefit plans (including health insurance) under the Employee Retirement Income Security Act of 1974 (ERISA). ERISA, administered by the U.S. Department of Labor, constrains the ability of states to regulate health care financing in the private sector. It was enacted as a reform of private-sector employer benefit plans, including health insurance (But- ler, 2002). Large multistate employers supported preemption of state laws to avoid having to comply with often-conflicting state regulations (Fox and Schaefer, 1989). ERISA preempts the states from directly regulating health coverage plans offered by private employers (âERISA plansâ), although states may regulate the state-licensed commercial insurance products that some employers purchase on behalf of their workers.16 The federal government, on the other hand, may regulate employer health plans under ERISA, imposing standards for services, benefits, and eligibility that can expand private coverage. To date, federal regulation aimed at extending private coverage has func- tioned through three statutes whose intentions are to improve the portability and continuity of employment-based coverage. They include: â¢ the Consolidated Omnibus Budget Reconciliation Act of 1985 (CO- BRA), which guarantees that workers (and their covered spouses and dependents) enrolled in their employerâs group health insurance be permitted to continue enrollment for up to 18 months after leaving their job (up to 36 months in special cases) (Meyer and Stepnick, 2002); â¢ the Health Insurance Portability and Accountability Act of 1996 (HIPAA), which guarantees that certain categories of workers, previously enrolled in quali- fied employment-based health insurance, be allowed to purchase coverage after they have exhausted their COBRA eligibility and restricts the imposition of 16The market of firms that insure their own health coverage risks is sizable; in 2003, nearly half of workers with health insurance were in partially or fully self-funded plans (Kaiser/HRET, 2003).
90 INSURING AMERICAâS HEALTH waiting periods for persons with preexisting conditions who switch health plans (Meyer and Stepnick, 2002); and â¢ the Trade Act of 2002 (TA), which provides a fully refundable, advanceable federal income tax credit for 65 percent of the cost of coverage for certain groups of workers displaced by international commerce and retirees and establishes new grants to the states to support administrative costs and state high-risk insurance pools (Dorn, 2003). This provision extends subsidized eligibility for a relatively small number of unemployed workers and their dependents, roughly 260,000 people (Dorn, 2003). Because of federal preemption, state laws do not reach certain firms or types of coverage. But many states have similar or stronger regulations that preceded COBRA and HIPAA, such as guaranteed issue and renewal; COBRA-like provi- sions for small firms (employers with fewer than 20 workers are exempt from COBRA); restrictions on premiums; requirements that coverage be convertible to a nongroup policy; and limitations on insurersâ use of preexisting conditions to deny or restrict coverage (Nichols and Blumberg, 1998; Gruber, 2001; GAO, 2002). All three federal laws attempt to preserve coverage for unemployed workers and their families, particularly those unemployed temporarily. HIPAA also targets anyone changing from one employment-based plan to another. One recent study puts the number of unemployed at 8.3 million and those both unemployed and uninsured at 3.9 million, with an additional 1.5 million persons per month losing their jobs in 2003 (Etheredge and Dorn, 2003). Unemployed adults have an uninsured rate threefold higher than the uninsured rate for adults in the general population (Lambrew, 2001). About 58 percent of uninsured adults (point-in- time estimate of coverage status) report having changed or lost jobs in the previous year (Gruber, 2001). Over the course of a year, workers earning less than twice the poverty level and their dependents are both more likely to be uninsured and more likely to lose what coverage they may have, whether private or public. They are more likely to experience changes in family structure that affect eligibility, more likely to change or lose jobs, less likely to be offered employment-based coverage when they are employed, and more likely to find COBRA unaffordable (Ku and Ross, 2002). Improving the level of coverage within this group would close a significant gap, as more than 80 percent of the uninsured are members of working families and nearly two-thirds are in families that earn less than 200 percent of FPL (IOM, 2001a). If all persons in households with incomes below twice the poverty line were able to keep their private coverage continuously over one year, there would be more than a 25 percent decrease in the number of uninsured adults and nearly a 40 percent decrease in the number of uninsured children (Ku and Ross, 2002). However, workers earning less than twice poverty are less likely to be eligible for COBRA than are workers in higher income brackets, reflecting the fact that many
LESSONS FROM THE PAST AND PRESENT 91 lower income employees do not obtain coverage through their employers or work for small firms (Doty and Schoen, 2001; Lambrew, 2001; Zuckerman et al., 2001). There is little information with which to assess the success of COBRA and HIPAA in improving coverage levels among the unemployed. The lack of regu- lation to make premiums affordable to unemployed workers has seriously limited their impact (GAO, 2002; Meyer and Stepnick, 2002; Pollitz et al., 2000). The average premium bill under COBRA in 2002 is 102 percent of the full cost of employment-based coverage, or $288 per month for an individual and $771 monthly for family coverage (Kaiser/HRET, 2003). While three out of four workers are eligible for COBRA, only about 20 percent of those eligible take up enrollment under COBRA (Lambrew, 2001). A survey of unemployment ben- efits in six states finds that the monthly premium for coverage under COBRA would consume much of, and in some cases the entire, unemployment benefit (GAO, 2002). At the same time, in four of the six states, adults receiving unem- ployment benefits had too high an income to qualify for public insurance, al- though their dependent children might have been eligible (GAO, 2002). In 2000, more than 50 percent of the children of unemployed adults obtained coverage through Medicaid or SCHIP, compared with 9 percent of unemployed adults (Lambrew, 2001). It remains to be seen whether the 65 percent subsidy given to displaced workers under the TAâs authority will make premiums affordable enough to increase insurance coverage among this group (Etheredge and Dorn, 2003). STATE INITIATIVES TO EXTEND COVERAGE In the 1990s, particularly in the years surrounding and following the failure of the Clinton Administrationâs proposed comprehensive health care reform, the states took leadership in efforts to reduce uninsurance within their boundaries (Oliver and Paul-Shaheen, 1997; Paul-Shaheen, 1998; Brown and Sparer, 2001; Holahan and Pohl, 2003). This section describes a key constraint on state options, the ERISA, and summarizes the experiences of five statesâHawaii, Massachu- setts, Minnesota, Oregon, and Tennesseeâthat invested both funds and political capital during the 1980s and 1990s in programs that markedly extended public coverage and lowered their uninsured rates.17 While many more states have established innovative coverage expansions in the 1980s and 1990s, the Commit- tee has chosen to look briefly at five leading states whose efforts were both intended to and did have a dramatic impact and whose experiences provide a 17Lower uninsured rates are primarily influenced by a stateâs level of employment-based coverage and also reflect economic characteristics of the state or region (including the propensity of employers to offer coverage) and specific demographic and socioeconomic characteristics of their populations; limited evidence allows for the sorting out of these different influences on coverage status of state populations over time. See discussion in Appendix B of the Committeeâs fourth report, A Shared Destiny (IOM, 2003a).
92 INSURING AMERICAâS HEALTH comparison of different paths to reform.18 These states have regulated private- sector coverage (employment-based coverage and individually purchased cover- age in the small and nongroup markets) and revamped public coverage (both programs that receive federal matching funds and programs funded wholly by the state and localities). Fiscal pressures on state governments and a growing unmet need for health services have catalyzed reform efforts to extend coverage. The need for states to stay within fixed budgets that cannot have deficits, to convince legislators to allocate new funds for public coverage, and to forestall the substitution of new public coverage for existing private coverage have stemmed the early ambitions of state programs to achieve universal coverage (Gold et al., 2001). In addition, state governmentâs capacity to finance health care and coverage extensions tends to be weakest at times when demands for such support are likely to be highest, for example, during an economic recession (IOM, 2003a). Administrative barriers to enrollment, including those that discourage crowd out, protect a stateâs program from running out of money. However, these provisions can defeat the purpose of maximizing public coverage, for example, requiring that a child be uninsured for six months before becoming eligible for SCHIP. When states do lower adminis- trative barriers to enrollment, the response can be positive and rapid. Box 3.5 illustrates one example, that of New Yorkâs Disaster Relief Medicaid program. Starting in 2001, state efforts to extend coverage have been complicated by a period of financial stress on state budgets that has intensified with the subsequent economic downturn (Guyer, 2003; Jenny and Ferradino, 2003; McNichol, 2003). For some states, including the ones to be discussed, economic pressures have translated into cutbacks in Medicaid and other public coverage programs, with cuts in eligibility for parents and other low-income adults, the trimming of ben- efits, the addition of administrative procedures likely to slow or interrupt cover- age, and greater cost-sharing (Fossett and Burke, 2003; Ross and Cox, 2003; Smith et al., 2003).19 However, because state dollars spent on Medicaid and SCHIP bring in federal matching funds, states have made efforts to maintain their investment in public coverage (Howell et al., 2002; Boyd, 2003; Holahan et al., 2003d). State budget cuts to programs that receive matching funds result not only 18It should be noted that the existing limits to reform have motivated recent planning activities to guide new approaches that build on current programs. Starting in 1999, the federal Health Resources and Services Administrationâs State Planning Grant program has awarded three separate cycles of one- year grants to fund data gathering about state uninsured problems and planning for reform of health services organization, financing, and delivery to close the âgapsâ left by existing public and private coverage in the state (Sacks et al., 2002). 19Recent passage of the Jobs and Growth Tax Relief Reconciliation Act of 2003 is expected to give states a new infusion of funds by temporarily raising the Federal Medical Assistance Percentage (FMAP), or matching rate for the Medicaid programs, lessening the likelihood of further program cuts (Ku, 2003).
LESSONS FROM THE PAST AND PRESENT 93 BOX 3.5 Administrative Simplification: New Yorkâs Disaster Relief Medicaid Program When the twin towers of the World Trade Center collapsed on September 11, 2001, New York state lost its computer system for the Medicaid program. The dramatic economic and public health effects of the terrorist attacks on local resi- dents, particularly low-income persons who lost their jobs as a result, were ampli- fied by the threatened loss of public health insurance coverage. Within weeks, the state of New York obtained a waiver from the federal gov- ernment for a four-month emergency expansion of full Medicaid benefits called the Disaster Relief Medicaid program. Income eligibility levels were raised to the level planned for the stateâs Family Health Plus expansion (not yet implemented at the time of the attacks), assets tests were removed, and emphasis was placed on simplifying enrollment. Eligible persons could visit one of many outreach centers located around the city, be assisted in filling out the one-page application, and receive coverage immediately. Community-based groups and local philanthropies were engaged to publicize the program and participate in the enrollment process. Disaster Relief Medicaid succeeded dramatically in raising public coverage levels locally. During the four-month period of enrollment, which ended on January 31, 2002, nearly 380,000 persons enrolled in Disaster Relief Medicaid, which is approximately eight times the number that would have been predicted to have enrolled during that time period in conventional Medicaid (Szalavitz, 2002). After the closing of the temporary program, the state began a more complicated process of reenrolling some Disaster Relief Medicaid recipients in the conventional pro- gram and dropping others from public coverage, returning to previous enrollment protocols for Medicaid that present many administrative barriers (Adcox, 2002). Vigorous, community-based outreach and a streamlined, simplified approach to enrollment have been credited with Disaster Relief Medicaidâs success in enroll- ing eligible persons where previously eligible persons encountered administrative and other barriers to enrollment (Russakoff, 2001). Latino and Chinese-speaking focus group participants in a recent study of the program cited their enthusiasm for the program, which they often learned about through word of mouth from family members, friends, and neighbors, and praised the relatively simple enrollment pro- cess, especially in contrast with their experiences and perceptions about the Med- icaid program (Perry, 2002). Respondents noted their difficulties in obtaining need- ed care before enrolling in Disaster Relief Medicaid and reported increased utilization of services, particularly to obtain prescription drugs, preventive and screening services, and care for chronic conditions (Perry, 2002). SOURCES: Russakoff, 2001; Adcox, 2002; Perry, 2002; Szalavitz, 2002; Doyle, 2003. in lost coverage and access to care for formerly insured residents but also the loss of federal funds. Despite budget pressures, over the past year reform advocates in a number of states, other than the five discussed below, have put together sweeping coverage extensions that have garnered broad legislative support (Associated Press, 2003; Orenstein and Fox, 2003). California, for example, has enacted a mandate, to be
94 INSURING AMERICAâS HEALTH phased in over 4 years, that firms with 50 or more employees either offer coverage or support a state pool that provides coverage (a âpay or playâ statute) (Freudenheim, 2003; Ingram et al., 2003; Rundle, 2003; Waldholz, 2003). If fully funded, the legislation is anticipated to cut the number of uninsured by one million persons. However, the law raises several ERISA issues and will likely be challenged in court. A second example, the state of Maineâs recently enacted Dirigo Health plan, is intended to bring about universal coverage for the stateâs roughly 180,000 uninsured persons by the year 2009, relying on a mix of public and private coverage sources and financing by means of a premium tax levied on employers as well as federal Medicaid dollars (Associated Press, 2003; Carrier, 2003; Haskell, 2003). The States and the Employment Retirement Income Security Act of 1974 (ERISA) Finding: The federal Employee Retirement Income Security Act of 1974 (ERISA) constrains the ability of states to mandate employ- ment-based coverage, one strategy to extend private coverage within their boundaries. Because most Americans receive their health insurance through their employ- ers or as dependents on anotherâs employment-based policy, there has been much interest in extending private-sector coverage, through insurance market reforms or regulation of premiums, benefit packages, and eligibility. However, as discussed earlier in the chapter, ERISA effectively precludes states from trying to expand workplace coverage by directly regulating employersâ health plans (Butler, 2002). ERISA does permit states to regulate the insurance companies and the com- pany products that the state licenses, but not self-insured employer plans (Butler, 2000). The statute has implications for the options for extending coverage, either publicly or privately. In the area of public financing, states can levy a tax on employers and insurers to finance a public coverage program (i.e., through a payroll, income, or other transaction tax), but the tax may not be imposed directly on an employerâs health plan, and state assistance to help purchase employment- based coverage for low-waged workers must depend on voluntary participation by employers and cannot be mandated (Butler, 2000). Neither state nor federal benefits mandates (which also apply to self-insured plans exempt from state regu- lation under ERISA) appear to influence the decision of a firm to self-insure (Jensen and Morrisey, forthcoming). Although court rulings since the mid-1990s have given states more leeway to regulate in ways that may affect ERISA health plans, ERISA remains a significant influence. In the shadow of ERISA, the states have largely been limited in their ap- proaches to extending coverage to regulating their small-group and nongroup insurance markets and extending public coverage. The division of responsibilities between the private and public sectors and, within the public sector, among
LESSONS FROM THE PAST AND PRESENT 95 100 90 80 70 60 Rate (Percent) 50 40 30 20 10 121 637 397 511 606 43,316 0 Hawaii Mass. Minn. Oregon Tenn. United States Employment-Based Individually Purchased Coverage Insurance Public Insurance Uninsured FIGURE 3.2 Coverage and uninsured rates and number of uninsured (in thousands) for population under age 65 in selected states and national averages, 2002. NOTE: Coverage reported as military (i.e., Tricare, CHAMPVA) is counted as employ- ment-based. Precise estimates do not add to 100 percent due to rounding and because respondents may report more than one source of coverage within a year. SOURCE: U.S. Census Bureau, 2003d. federal, state, and local jurisdictions, has made it difficult for states to fund public extensions for persons with incomes above the poverty level.20 As a result, across the states discussed in this section, there is little variation in sources of coverage, other than the percentage of employment-based coverage; see Figure 3.2. 20See Gold et al. (2001) for a discussion of this âstructural pluralismâ and the inherent tensions and ambiguities that this pluralism engenders.
96 INSURING AMERICAâS HEALTH Five Leading States for Coverage Extensions Finding: At the state level, significant reduction of uninsurance is more likely with incrementalism that is integrated into a compre- hensive reform strategy over time than it is with ad hoc or disjointed changes implemented without a plan to achieve universal coverage. Finding: Although some states have made significant progress in reducing uninsurance, even the states that have led major coverage reforms have large and persisting uninsured populations. Finding: Despite the use of federal Medicaid waivers to leverage the dollars spent by entirely state-funded public insurance programs, states do not have the fiscal resources to implement fully their existing public coverage programs. Reliance on income or sales taxes to raise revenues makes the funding base for public coverage programs unstable. States are further constrained from eliminating uninsurance within their boundaries by categorical limits on eligibil- ity for federally supported public coverage programs. In the sections that follow, the Committee highlights the differing approaches and experiences of five states that led reform in the 1990s. Since the mid-1980s, each of these states has taken advantage of Medicaid Section 1115 waivers to expand their programs beyond the mandatory populations and eligibility levels established by the federal government. Hawaii, Massachusetts, Minnesota, and Oregon folded in their own separate coverage programs for persons ineligible for Medicaid. It is important to note, however, that Section 1115 waivers do not bring new federal dollars into a state; all waivers are required to be cost neutral to the federal government. Following Hawaiiâs 1974 employer health coverage man- date, in the early 1990s Massachusetts, Minnesota, and Oregon each enacted and attempted to implement extensions of coverage geared toward universality, and Tennessee made its goal to dramatically increase public coverage. The experiences of each of these states points out the limits of reform without new federal funds. Despite the fact that these states started with broad employment-based coverage and public programs and have extended public coverage significantly, none has achieved universal coverage within their state. The progress they have made is now jeopardized by ongoing fiscal and economic difficulties facing states. Not- withstanding the recent passage of legislation in Maine that would move the state toward universal coverage, it is important to recognize that few states have taken on significant extensions of public coverage. Most states have higher uninsured rates than the states selected here for discussion and, thus, much further to go to achieve universal coverage.
LESSONS FROM THE PAST AND PRESENT 97 Hawaii Hawaii is home to nearly 1.2 million people of all ages, more than one-third of whom live in households earning less than 200 percent of FPL (KCMU, 2003d). The state has attempted to boost health insurance coverage through its employer mandate (the only one in the United States), called the Prepaid Health Care Act of 1974 (PHCA), and through public coverage. PHCA was enacted prior to ERISA and received an exemption from Congress in 1983 (The Hawaii Uninsured Project Leadership Group, 2002). PHCA requires most employers (with the exception of government) to offer coverage to employees working at least part-time (19.5 hours per week on a continuing basis) and makes a premium subsidy available to small firms (Law, 2000). Hawaii has addressed some of the coverage gaps left by PHCA through public coverage. A Medicaid Section 1115 waiver awarded in 1994 created the Health QUEST program, a capitated Medic- aid managed care strategy that combines Medicaid, SCHIP, and state-only cover- age programs. While Hawaiiâs employer mandate has been interpreted as an important precedent for other states and a positive step toward universal coverage within the state, it is unclear how much implementation of PHCA has reduced uninsurance, compared to other factors including a relatively lower rate of increase in health care inflation, the particulars of the stateâs market for insurance (i.e., existence of modified community rating, small number of insurers), and health care utilization patterns (Neubauer, 1993; GAO, 1994). An 8 percent drop in the number of uninsured has been attributed to PHCA (about 8,000 newly insured persons), with socioeconomic and demographic characteristics such as the overrepresentation of Asian ethnic populations with higher-than-average coverage rates better ex- plaining shifts in private coverage and the uninsured rate observed after enactment of the employer mandate (Dick, 1994). Even with its unique exemption from ERISA, serious constraints on Hawaiiâs employer mandate have kept it from reaching its full potential. Criticisms of PHCA have focused on the stateâs lack of enforcement or oversight of employer compliance, the lack of a mandate that coverage be offered to a workerâs depen- dents, and the fact that workers whose employers are excluded from the Act are most likely to be uninsured (e.g., temporary, contract, and self-employed workers) (Dick, 1994; Law, 2000); some of the difficulties with PHCA stem from the fact that the one-time exemption granted from ERISA did not permit the state to update or modernize its program once the exemption was granted, leading to administrative inefficiencies and in some ways a mismatch between subsequent needs and program capabilities. However, more than 90 percent of employers do offer their workers the option of purchasing a health insurance plan that includes dependents (The Hawaii Uninsured Project Leadership Group, 2002). Because Hawaiiâs ERISA exemption has preserved PHCA basically as it was in 1974, the stateâs ability to modernize the program has been hindered and it has established other programs to cover the uninsured. Public coverage through
98 INSURING AMERICAâS HEALTH Health QUEST extends eligibility to pregnant women at 185 percent of the poverty line, children through 18 years at twice the poverty line, and other adults who earn less than poverty (The Hawaii Uninsured Project Leadership Group, 2002). Two other public programs extend coverage to those who would likely be uninsured otherwise: legal immigrant children ineligible for Medicaid and SCHIP and short-term coverage to persons earning less than 300 percent of poverty who no longer qualify for other public insurance (The Hawaii Uninsured Project Leadership Group, 2002). Funds to support the QUEST programs come from state general revenues, federal matching funds from Medicaid and SCHIP, and enrollee premiums. The most recent Census Bureau estimates put Hawaiiâs uninsured rate for persons under age 65 at 11.4 percent (2002), lower than the national average though not the lowest in the nation, and about 121,000 people remain uninsured (U.S. Census Bureau, 2003d).21 Of the stateâs uninsured, most are adults under age 65, although there are 24,000 uninsured children under age 18 (U.S. Census Bureau, 2003d). Three-quarters of the uninsured are in households earning less than twice the poverty line, for whom the uninsured rate is 22 percent (KCMU, 2003d). Massachusetts Nearly 6.3 million people of all ages live in Massachusetts, and almost a third live in households earning less than twice the poverty line (KCMU, 2003e). The state first moved toward universal coverage in the late 1980s and subsequently has significantly extended public coverage. Legislation passed in 1988 promised uni- versal coverage, expanding current programs, filling in selected gaps and integrat- ing certain aspects of the programs, and including an employer âpay or playâ provision that was never implemented.22 Voters came close to passing a second universal coverage referendum during 2000. In 1997, Massachusetts implemented a Section 1115 Medicaid waiver that broadened eligibility for existing public coverage programs, created two new programs, and merged the entirety into a unified MassHealth program (Bovbjerg and Ullman, 2002). 21The stateâs annual telephone Health Survey provides a much smaller estimate of 5.5 percent uninsured, or approximately 64,440 uninsured people, for the most recent year (2001) for which data are available (State of Hawaii, 2002; The Hawaii Uninsured Project Leadership Group, 2002). The state and Census Bureau estimates are not comparable, given differences in data collection, definitions of coverage status, and methods. For the states discussed in this section, the estimated uninsured rate is based on data from the Census Bureauâs March Current Population Survey. 22The term âpay or playâ typically refers to a requirement that employers either offer a health insurance plan to their workers and dependents or pay (often a payroll tax) into a public fund to support health insurance coverage for uninsured employees and families.
LESSONS FROM THE PAST AND PRESENT 99 In the first two years of the waiver, there was a 14 percentage point increase in public coverage and greater than a 7 percentage point drop in the uninsured rate for low-income children; for low-income adults, there was an 8 percentage point increase in public coverage and an 11 percentage point drop in the unin- sured rate (Bovbjerg and Ullman, 2002). Three years into the waiver, enrollment in public coverage had grown from about 700,000 to 926,000 (Bovbjerg and Ullman, 2002). MassHealth consists of two programs that offer eligibility to the disabled and Medicare-eligible persons and four programs that extend coverage to the noninstitutionalized, nondisabled population (Bovbjerg and Ullman, 2002). These programs extend eligibility to â¢ pregnant women and infants in families earning up to twice the poverty level; â¢ children up to 150 percent of FPL, and parents up to 133 percent of FPL (Standard); â¢ unemployed low-income adults and their families earning up to four times the poverty level (Basic); â¢ children in families earning between 150 and 200 percent of poverty, adults earning up to 200 percent of poverty through an employer premium subsidy, and persons living with HIV (but not with active AIDS) earning less than 200 percent of poverty (Family Assistance); and â¢ emergency services (through MassHealth Limited) for low-income per- sons ineligible for other programs (e.g., undocumented immigrants) (Rosenbaum et al., 2002). MassHealth is funded by state revenues and from federal matching funds from both Medicaid and SCHIP. Massachusettsâ uninsured rate of 11.3 percent (2002) is lower than the na- tional average for the population under age 65. Nonetheless, there are approxi- mately 637,000 uninsured under age 65 in the state (U.S. Census Bureau, 2003d). High levels of both employment-based and public coverage contribute to the relatively low uninsured rate (Bovbjerg and Ullman, 2002). Six percent of the stateâs children under age 18 are uninsured (approximately 88,000 uninsured children), and the rest of the uninsured are adults under age 65 (U.S. Census Bureau, 2003d). Minnesota A total of about 4.9 million people live in Minnesota, and 22 percent of them live in households that earn less than 200 percent of FPL (KCMU, 2003f). Min- nesota has extended public coverage gradually, through a step-by-step or phased- in incremental approach of filling in private coverage gaps using the availability of federal matching funds through Medicaid to maximize the eligibility levels it can
100 INSURING AMERICAâS HEALTH offer state residents (Chollet and Achman, 2003). In 1994, Minnesota inaugurated its own public coverage program, MinnesotaCare, intended to achieve universal coverage in ten years by expanding eligibility to all members of low- and moder- ate-income families ineligible for Medicaid (Gold et al., 2001). A year later, the state folded MinnesotaCare into a Section 1115 Medicaid waiver, together with the stateâs Medicaid program and other state-supported public coverage (Gold et al., 2001). Minnesotaâs public coverage programs extend eligibility through a small group purchasing pool for county, town, and school district employees and their families; to childless adults earning less than 70 percent of FPL who are ineligible for Medicaid (e.g., noncitizen immigrants, persons in mental institutions) (General Assistance Medical Care); pregnant women and infants under two years at nearly three times the poverty level, families with children earning less than 275 percent of FPL, and childless adults earning less than 175 percent of FPL (Medical Assis- tance, MinnesotaCare); and a high-risk pool (Chollet and Achman, 2003). Be- cause the stateâs public programs made children eligible well above 200 percent of FPL before the enactment of SCHIP and SCHIP funds can be used only to extend coverage to newly eligible populations, Minnesotaâs SCHIP program covers only a few hundred children (Chollet and Achman, 2003). Public coverage is financed with a mix of federal, state, county, and private dollars, including enrollee premi- ums and a tax on health care providers; about half of the stateâs spending is matched by federal dollars, for the most part Medicaid (Chollet and Achman, 2003). The state is at a relative disadvantage financially compared with other states because it cannot reap the full benefit of new federal funds for SCHIP that are matched at a higher rate than is Medicaid spending. Changes in enrollment have reflected welfare reform and growth in earnings among the poor, for example, a dip in public programs (Medical Assistance and General Assistance Medical Care) after the delinking of welfare from Medicaid (Chollet and Achman, 2003). Over the decade from 1991 to 2001, both of these programs have declined somewhat in enrollment overall (although the numbers of enrollees outside of the welfare system grew), while there has been steady growth for MinnesotaCare (Chollet and Achman, 2003). The stateâs low uninsured rate of 8.8 percent (2002) for the population under age 65 reflects high levels of both employment-based and public coverage (U.S. Census Bureau, 2003d). In 2000, public programs enrolled nearly all uninsured low-income persons under age 65 in the state (Chollet and Achman, 2003). Yet there are still an estimated 397,000 uninsured persons under age 65 (U.S. Census Bureau, 2003d). Most (325,000) are adults. About two-thirds of this uninsured population (including 91 percent of uninsured children) are believed to be eligible for but not enrolled in some type of coverage. If all eligible persons were enrolled, the stateâs uninsured rate, it is estimated, would stand at 2.7 percent (Chollet and Achman, 2003).
LESSONS FROM THE PAST AND PRESENT 101 Oregon There are approximately 3.4 million persons of all ages in Oregon, one-third of whom live in households earning less than twice the poverty line (KCMU, 2003g). Since the 1980s, health reformers have attempted to implement universal coverage in the state in coordinated phases. The most widely known of these reforms is the Oregon Health Plan, enacted in 1994, which expanded Medicaid (through a Section 1115 Medicaid waiver), imposed an employer mandate (never implemented because the state did not receive the exemption it sought from ERISA), provided a public subsidy for workers to purchase employment-based coverage, and created a state high-risk pool (Gold et al., 2001). The Medicaid expansion received national attention for its innovative benefits package (initially intended to be applicable to the employer mandate and other parts of the Oregon Health Plan), which ranked conditions by priority for coverage, given the budget- ary constraints imposed by the Medicaid waiver obtained to implement the ex- pansion (Conviser, n.d.; Skeels, 1994; Jacobs et al., 1999). Conditions for which treatment would be covered under the expansion have been ranked in order of priority (reflecting cost effectiveness, the number of people potentially affected, and other factors) and funding decisions made on the basis of the prioritized list. A series of working groups and meetings across the state engaged, and continue to engage in, public and legislative discussion about the scope of health insurance benefits. Currently, the Oregon Health Plan extends Medicaid eligibility to pregnant women and children under age 12 with incomes up to 170 percent of FPL and other residents earning up to the poverty line (Gold et al., 2001). Through the Family Health Insurance Assistance Program, subsidized coverage is available for persons ineligible for Medicaid who earn up to 170 percent of FPL (Gold et al., 2001). The Health Plan is financed by cost savings achieved through mandatory managed care participation by enrollees and state revenues including income taxes and a sales tax on cigarettes, a funding base vulnerable to change with the changing economic fortunes of the state. As a result, enrollment barriers have been raised to slow growth of the public expansion, through income and assets tests and premi- ums (Gold et al., 2001). In 1993, before the Oregon Health Plan was implemented, the stateâs unin- sured rate was 14.7 percent, or about 453,000 persons under age 65 (Gold et al., 2001). In the first year, there was an unanticipated groundswell of participation, with approximately 100,000 persons newly enrolled (the initial goal was to extend coverage to an additional 130,000 persons), of whom roughly 75,000 were new to public coverage (Leichter, 1999; Gold et al., 2001). Since the first year, growth in public coverage has been more modest, covering in total an estimated additional 130,000 low-income persons who would otherwise be uninsured (Leichter, 1999). Four years into the Health Plan, roughly two-thirds of the 1993 low-income uninsured population was enrolled (Gold et al., 2001).
102 INSURING AMERICAâS HEALTH Current coverage gaps include uninsured adults earning more than 100 per- cent of FPL, including those with incomes below 170 percent of FPL who, while eligible for the public subsidy, either may not be able to enroll because of fiscal limits on the coverage programs or may be unable to find affordable coverage (Gold et al., 2001). The stateâs uninsured rate of 16.5 percent (2002) for the population under age 65 is barely lower than the national average, and an esti- mated 511,000 persons under age 65 lack coverage (U.S. Census Bureau, 2003d). More than eleven percent of the stateâs children under age 18 are uninsured (roughly 95,000 uninsured children) (U.S. Census Bureau, 2003d). Tennessee Tennessee is home to about 5.6 million people of all ages, nearly 40 percent of whom live in households earning less than twice the poverty level (KCMU, 2003h). Although the stateâs insurance expansion was not expected to bring about universal coverage, it did broaden public coverage dramatically and significantly in 1994, through reform of its Medicaid program (Gold et al., 2001). A few years earlier, Congress had restricted the use of provider taxes to raise state matching funds for Medicaid, throwing Tennesseeâs publicly financed health care into fiscal crisis. State officials responded by obtaining a Section 1115 waiver that allowed the state to extend public coverage to greater numbers of uninsured persons and recapture federal Medicaid dollars that would no longer be available through the DSH program. The waiver created a new program, TennCare, that implemented mandatory managed care for all enrollees and doubled the number of enrollees within its first year. TennCare extends eligibility for coverage to 400 percent of FPL. However, since January 1995, enrollment has been capped at 1.3 million persons because of limited public dollars to support further enrollment. Although TennCare is funded by federal, state, and local funds, including Medicaid DSH payments to hospitals and annual insurer assessments, support has been insufficient to cover all who are eligible to enroll and has also constrained provider reimbursements. One study estimates that, during TennCareâs first five years, the federal and state governments spent about $700 million less than would have been predicted for the Medicaid program without the waiver. At the same time, TennCareâs expansion of eligibil- ity cost approximately $3.8 billion more (net new costs of $3.1 billion) from all payers than would have been predicted when anticipated changes in charity care, patientsâ cost sharing, and local government spending were considered (Conover and Davies, 2000). At present, no new eligible persons may enroll unless they are members of groups required to be covered by Medicaid, dislocated workers, children in fami- lies earning less than 200 percent of FPL, and children in families earning between twice and four times the poverty limit who do not have access to employment- based coverage (Conover and Davies, 2000; Gold et al., 2001). In the year before TennCare (1993), Tennesseeâs uninsured rate was 13.2 percent (about 673,000 persons under age 65) (Gold et al., 2001). In the first year after TennCare began,
LESSONS FROM THE PAST AND PRESENT 103 the stateâs uninsured rate shrank by one-third to one-half, putting the state well below the national average uninsured rate. Four years into the program, nearly 400,000 persons have left the ranks of the uninsured and enrolled (Gold et al., 2001). Currently about 12.0 percent (606,000 persons) of the stateâs residents under age 65 are uninsured (2002) (U.S. Census Bureau, 2003d). Most are adults, al- though nearly one in six are children under age 18 (U.S. Census Bureau, 2003d). LOCAL INITIATIVES TO EXTEND COVERAGE In many states across the country, counties are the providers of last resort for the underserved and uninsured (IOM, 2003a). Counties have responded to this charge in a variety of ways. The three counties reviewed in this section are among the few jurisdictions that have programs devoted explicitly to increasing the level of insurance coverage significantly and reducing uninsurance. Each takes a differ- ent approach, tailored to the characteristics of the local population, the resources available to deliver and pay for care that would otherwise go uncompensated or not be received at all, and local leadership. At each site, state or local conversion foundations, created as part of changes in the ownership status of health plans or hospitals, have contributed financing for coverage initiatives. From the following discussion of local extensions, the Committee draws the following observation. Finding: Extensions of public or private coverage at the county level have focused on increasing coverage among targeted populations rather than the entire uninsured population locally. Despite the potential of local programs to address targeted gaps, the lack of a reliable funding source limits their scope and effectiveness. San Diego, CA In 2001, there were nearly 365,000 uninsured children and adults under age 65 in San Diego County, or about 15.1 percent of the countyâs 2.4 million residents (Brown et al., 2002).23 There is a sizable coverage gap among low- income workers. In 1997, two local organizationsâthe Sharp Health Plan (a nonprofit insurer) and the Alliance Healthcare Foundation (a conversion founda- tion)âcreated a small-scale demonstration program to reduce uninsured rates 23This is a point-in-time estimate (e.g., the survey respondent reported his or her coverage status at the time of participation in the survey). Analysts evaluating the countyâs uninsured problem devel- oped a much higher estimate of the number of uninsured in the county, 537,000 persons, based on a three-year moving average of national Census data (the March Current Population Survey 1998â 2000) that estimates uninsured status over a one-year period of time (the calendar year preceding the year of the survey) rather than a point-in-time estimate (Kronick, 2002). This higher estimate of the number of uninsured persons yields a higher county uninsured rate of 21.7 percent.
104 INSURING AMERICAâS HEALTH among low-income workers by offering employer premium assistance to small firms that do not offer employment-based coverage (Silow-Carroll et al., 2001). The program, abbreviated FOCUS (for Financially Obtainable Coverage for Un- insured San Diegans), targeted firms with 50 or fewer workers and formally began in April 1999. Firms are eligible to participate if they have not offered coverage in the previous year, and they are given a two-year commitment of support (Silow- Carroll et al., 2000). Employees may enroll if they work full-time, earn up to 300 percent of FPL, and have been uninsured for the past year, with the requirement that all dependents be enrolled as well; however, dependents may also be eligible for public coverage (Silow-Carroll et al., 2000). Both employers and employees contribute to the premium, which is subsidized by private dollars from Sharp Health and two foundations (the California Endowment and the California Health Care Foundation) established when Blue Cross of California converted to for- profit ownership status (personal communication, Jeffrey Lazenby, Sharp Health Care, April 29, 2003). Providers are paid on a fee-for-service basis, and the benefit package is comparable to that of other local commercial benefits, with copayments but no deductibles (Silow-Carroll et al., 2001; personal communication, Jeffrey Lazenby, Sharp Health Care, April 29, 2003). Although outreach has been an important part of FOCUS and the business community has offered much interest and support, enrollment has been limited by the availability of funding. To date, FOCUS has obtained roughly $3 million in private support (personal communication, Jeffrey Lazenby, Sharp Health Care, April 29, 2003). The target population initially identified was the approximately 49,000 adult workers employed by firms that did not offer coverage and the initial goal of FOCUS was to enroll 1,000 workers. As of mid-2000, nearly 2,000 workers and dependents were estimated to be enrolled, representing 232 busi- nesses; participating employers are more likely than average to have uninsured owners who are also more likely to be foreign born and to have a very low-waged workforce (Kronick, 2002). An estimated 55,000 to 80,000 uninsured workers and dependents would be eligible to enroll if the program were expanded (per- sonal communication, Jeffrey Lazenby, Sharp Health Care, April 29, 2003). At present, enrollment is closed to new firms but open to new employees and their dependents that join firms already participating in FOCUS. Alameda County, CA About 1.3 million people live in Alameda County, situated in the Bay Area and including the cities of Oakland, Berkeley, and Hayward. The county has an uninsured rate of about 8.4 percent, or roughly 109,000 uninsured under age 65 (Brown et al., 2002).24 More than half of the uninsured adults are in the workforce, 24In 2000, the county supported its own survey of sources of coverage status over a 3-month
LESSONS FROM THE PAST AND PRESENT 105 with 28 percent earning less than the poverty level and another 37 percent earning between 100 and 250 percent of FPL (Ponce et al., 2001). In 1996, the county health department began a collaboration with local providers, screening low- income people for Medi-Cal (the stateâs Medicaid program) for enrollment in a managed care plan (personal communication, Nina Maruyama, Alameda Alliance, May 29, 2003). The program, known as the Alameda Alliance for Health, has evolved into a private, nonprofit managed care health plan that receives a core of public funding and builds programming around its private-sector and foundation fundraising. Its goal has been to provide or coordinate seamless, continuous cov- erage for all members of families earning up to 300 percent of FPL who are county residents, considered as a unit rather than as individually eligible, regardless of eligibility for other public programs or immigration status (Ibarra, 2002). The Allianceâs strategy has been to gather stakeholders and to participate in coalitions devoted to improving access to care. Alliance coverage programs emphasize primary and preventive services while offering a comprehensive benefit package. Care is provided by a network of providers. Enrollees pay part of their premiums, with the Alliance supporting premiums that are not covered through public programs such as Medi-Cal or SCHIP. Copayments are tied to the type of service, with none for primary and preventive services to encourage utilization and a copay for emergency depart- ment visits to discourage nonurgent use. In addition, the Alliance takes a culturally sensitive approach to outreach among the countyâs diverse population, translating materials into a number of languages and working with community-based groups. More than half of the enrollees in its Family Care program are Spanish speaking and 19 percent speak Cantonese. The Alliance offers subsidized coverage to all members of families earning no more than 300 percent of FPL (Medi-Cal, an SCHIP program called Healthy Families, and Family Care) and unsubsidized coverage (through the First Care program) for those with higher incomes (Ibarra, 2002). For Family Care, which has nearly 5,200 enrollees, family members may qualify under different programs, for example, Family Care program eligibility is extended to family members (parents, siblings) of those who qualify for Medi-Cal or Healthy Families. As of spring 2002, approximately 81,000 persons were enrolled through the Alliance, most of whom (68,000) received coverage through Medi-Cal. In addition, the Alliance Group Care program begun in 2002 offers subsidized coverage to full- time home supportive services workers (about half of whom were previously uninsured). Financing comes from a combination of public and private sources. The period, arriving at an estimate that roughly 16 percent of the countyâs adults, or 140,000 persons, were uninsured (Ponce et al., 2001). While this surveyâs estimated uninsured rate is about double the estimate given by a statewide survey that included a sample of county residents questioned about their coverage status over the course of a year, both surveys arrive at estimated numbers of uninsured persons that are surprisingly similar (Brown et al., 2002).
106 INSURING AMERICAâS HEALTH nonprofitâs endowment is supported by grants from the California Endowment, the California Health Care Foundation, and tobacco settlement dollars from the county. Coverage programs have been added as new funding streams have be- come available, for example, through the federal SCHIP program (1998). Hillsborough County, FL In 2000, roughly 40,000 of the countyâs 1 million residents were uninsured and living below the poverty line (personal communication, Toni Beddingfield, Hillsborough Health Plan, April 30, 2003).25 Like other Florida counties, Hillsborough (which includes the city of Tampa) serves as a provider or payer of last resort for the countyâs medically indigent population. In the early 1990s, rising health care costs, especially uncompensated care costs at the county public hospitalâs emergency department, motivated county officials to devise a health care plan for the portion of its uninsured population below the poverty line. The Hillsborough Health Plan is intended to promote the use of primary and preventive services, targeting low-income families and coordinating the provision of coverage with other public services in the county (personal communication, Toni Beddingfield, Hillsborough Health Plan, April 30, 2003). Eligibility is restricted to persons earning no more than 100 percent of FPL; for persons earning up to 125 percent of FPL, catastrophic coverage is available with an income-based sliding scale premium (Hillsborough County, 2003). En- rollees obtain care on a fee-for-service basis through one of four networks of providers that have contracts with the county (personal communication, Toni Beddingfield, Hillsborough Health Plan, April 30, 2003). Since its inception, the plan has been supported by county revenues from property and a dedicated sales tax, as well as premiums for the catastrophic care plan (Hillsborough County, 2003). The county made a concerted outreach effort through its neighborhood service centers and with assistance from local community-based groups, reaching an initial enrollment of 15,000 out of the 40,000 eligible (personal communica- tion, Toni Beddingfield, Hillsborough Health Plan, April 30, 2003). In 2002, nearly 28,000 persons were enrolled in the Health Plan, divided between indi- vidual members (61 percent) and families (39 percent) (personal communication, Toni Beddingfield, Hillsborough Health Plan, April 30, 2003). In addition to extending coverage, the Health Plan has been estimated to have saved more than $11 million in hospital emergency department costs (personal communication, Toni Beddingfield, Hillsborough Health Plan, April 30, 2003). Like many other states and counties, however, Hillsborough has faced budget shortfalls over the past few years that have led County Commissioners to make difficult decisions. In 25A state-level survey in 2000 estimated the countyâs uninsured rate at 14 percent, with 117,000 uninsured persons under age 65 out of a general population of 839,000 (Lazarus et al., 2000).
LESSONS FROM THE PAST AND PRESENT 107 2003, cost-saving measures have been geared toward decreasing enrollment (by requiring more frequent reenrollment). SUMMARY None of the reform campaigns or public initiatives discussed in this chapter has achieved universal health insurance coverage. Economic and demographic changes over time have influenced the level of private coverage (employment- based coverage), with larger public insurance programs such as Medicaid and SCHIP modestly lowering the proportion of uninsured persons by filling in some of the many gaps in coverage created by the employment-based system. As illus- trated in Figure 3.1, however, given the absence of major federal reform targeting the general population since the mid-1960s, there has been little variation in the sources of coverage and in the uninsured rate over the past 25 years. Those pursuing extended coverage in recent years have grappled with con- cerns and obstacles shared by reformers before them. During the era before Medi- care and Medicaid, reformers sought health care financing arrangements that would apply universally (covering all members of society), be affordable to those seeking coverage, and be adequate in benefits to sustain health and well-being (making accessible the demonstrated and perceived advantages of medical care). The boom in private coverage between the mid-1930s and the 1960s, through Blue Cross and Blue Shield and independent plans initially, then by commercial insurers in the group market, made employment-based coverage the norm for most Americans. Implementation of Medicare and Medicaid in the 1960s filled in some of the gaps left open by employment-based coverage, for persons aged 65 and over and for categories of the poor and medically indigent. The two programs, which today cover nearly a quarter of the U.S. population, grew to be much more expensive than initially anticipated. After the federal government became a major insurer and purchaser of health care, reform campaigns shifted their focus to controlling costs, stressing the need for insurance schemes to be affordable to society, administra- tively efficient, and transparent to political stakeholders. Over the decades, there has been moderate public sympathy for the general idea of universal coverage, yet no groundswell of public interest in a particular strategy to reach this goal (indeed, polls report a drop in support for universal coverage when couched in terms of specific provisions or financing requirements) (Marmor, 1973), and the spillover effects of a large uninsured population on persons other than the uninsured them- selves have gone largely unacknowledged by the public (IOM, 2003a). Even though most uninsured persons are members of families with at least one worker, government efforts to reduce uninsurance since Medicare and Med- icaid continue to focus on public coverage to fill gaps. Since the mid-1980s, major federal initiatives to extend both public and private coverage have lowered unin- sured rates among children and raised the numbers with public coverage, although the number of uninsured overall has continued to grow. Between 1984 and 1990,
108 INSURING AMERICAâS HEALTH Congress gradually expanded Medicaid eligibility for pregnant women, infants, and young children, delinking coverage from welfare eligibility. These Medicaid expansions were followed in 1997 by the creation of SCHIP grants-in-aid to the states. SCHIP appears to have reduced the number of uninsured children recently, but millions of children remain uncovered, more than half of whom are eligible (Broaddus and Ku, 2000; Dubay et al., 2002a,c; Kenney et al., 2003). Federal initiatives to extend employment-based coverage have targeted improved port- ability and continuity of coverage through the COBRA, HIPAA, and TA statutes, yet the lack of authority or resources under COBRA and HIPAA to make insurance premiums affordable has seriously limited their usefulness and impact. With the exception of Tennessee, the states discussed in this chapter have relied on relatively high levels of employment-based coverage, plus generous public coverage for their low-income populations using Medicaid waivers and additional state funds to keep uninsured rates below the national average. Con- straints on federal dollars, for example, due to the budget-neutrality requirement of Medicaid waivers, and the shortcomings of the federal matching formula (FMAP) to compensate adequately for the effects of economic recessions, contrib- ute to the difficulties experienced by the states in extending coverage (Corrigan et al., 2003; IOM, 2003a). The federal ERISA also narrows state options to reform their private insurance markets, through which most of their residents obtain coverage. However, it is mainly the lack of sufficient or sufficiently stable public dollars that has checked broad access reforms, despite the willingness, albeit lim- ited, of taxpayers in these states and others to tax themselves in order to raise funds to extend coverage, for example, through tobacco taxes (Marquis and Long, 1997; IOM, 2003a). Hawaiiâs inability to update provisions of its employer mandate to meet newer needs for coverage and failure to enforce the mandate contribute to an ongoing population of uninsured low-income workers, while limited resources have constrained public programs intended to fill coverage gaps. The breadth of Massachusettsâ reforms is comprehensive, yet implementation of its âpay or playâ law for employment-based coverage was delayed and eventually repealed due to waning political support, limiting the stateâs ability to boost private-sector cover- age. Budget shortfalls have limited public coverage programs (MassHealth). In Minnesota, the uninsured rate is nearly the lowest in the nation, but gaps in coverage remain, jeopardizing its goal of universal coverage. The inability to obtain an ERISA exemption kept Oregon from using its innovative Medicaid expansion to broaden employment-based coverage in the state, and reliance on tax revenues has left the Oregon Health Plan underfunded. Finally, Tennesseeâs am- bitious efforts to use existing Medicaid dollars and managed care contracting to markedly extend the program to its uninsured population with low and moderate incomes resulted in increased enrollment among the poorest segment of this group in the first year, followed by closed enrollments to all but those required under law to be covered. Localities, whose economic resources are more limited than those of the states (which can cross-subsidize among communities), may come close to
LESSONS FROM THE PAST AND PRESENT 109 filling a particular local gap in coverage but remain even more constrained than are the states by a lack of sustainable revenues for public coverage. Despite gradual extensions of public programs at the federal, state, and local levels and isolated efforts around the country to move toward the goal of universal coverage, the lack of national political consensus has hindered a substantial reduc- tion, if not elimination, of the problem of uninsurance in the United States. Laudable state- and county-level efforts to extend coverage have been constrained by a lack of resources. The circumscribed nature of these past and present initia- tives suggests that attempts to provide universal coverage through state or local efforts without a substantial infusion of additional federal funds, and federal lead- ership, are unrealistic. Conclusion: The persistence of uninsurance in the United States requires a national and coherent strategy aimed at covering the entire population. Federal leadership and federal dollars are neces- sary to eliminate uninsurance, although not necessarily federal ad- ministration or a uniform approach throughout the country. Uni- versal health insurance coverage will only be achieved when the principle of universality is embodied in federal public policy. In the chapters that follow, the Committee builds on this base of knowledge about past and present efforts to reduce uninsurance to formulate its set of prin- ciples to guide a universal approach to insuring all Americans. These principles will be used to assess comprehensive models that describe âpureâ or ideal ap- proaches to universal coverage (for example, an employer mandate) in order to reach conclusions and recommendations about how the United States can defi- nitely and successfully eliminate uninsurance among its population.